Handelshistorik

Horisont: 1-2 years Begrundelse: Highest analyst upside in the universe (+44% to $328 target) after a ~6% pullback from last week's highs; long-term AI infrastructure compounder, sized to keep ~$170 cash buffer. === Session-analyse === [main/US] The portfolio is effectively frozen: 558 DKK cash (~$82 spendable) cannot buy any US name in the universe, since the cheapest, NVDA at $225, is far beyond reach. Three straight HOLD cycles confirm the freeze. The way out is a full exit of one position, which executes before any BUY and reopens the window. Candidate exits: MSFT is a +35.6% winner with only 8% analyst upside, the most stretched valuation in the book, and a cross-feed BUY that rests on analyst ratings alone. Selling it crystallizes the gain, which is disciplined profit-taking. AMZN and GOOGL are near breakeven with 28% and 22% analyst upside respectively, so selling them would abandon the stronger remaining theses. MSFT is the correct exit. Proceeds of about $542 plus $82 cash gives roughly $620 spendable. The best redeployment is NVDA: its fundamental upside is the highest in the universe (+44% analyst target) and it has pulled back from about $239 to $225 over the last week, a reasonable entry for a long-term holder. Two shares cost about $451, leaving ~$170 in cash buffer to avoid a re-freeze and cover fees. NVDA is cross-feed MIXED (1 BUY, 1 SELL), so this is a moderate-conviction rotation, not a high-conviction one. Both sides stay small relative to the 40% position cap. Rejected alternatives: META and TSLA report in 7 days, so no fresh positions before the print, and AAPL has a cross-feed SELL signal. AMZN and GOOGL are kept: both are near breakeven, and the 28% and 22% analyst upside plus BUY consensus support the long-term thesis. No new positions in META or TSLA ahead of the 2026-10-21 earnings, and no AAPL given the SELL consensus.

Horisont: exit now Begrundelse: Full exit of a +35.6% winner with only ~8% analyst upside left; frees ~$542 to break the cash freeze and recycle into higher-upside quality. === Session-analyse === [main/US] The portfolio is effectively frozen: 558 DKK cash (~$82 spendable) cannot buy any US name in the universe, since the cheapest, NVDA at $225, is far beyond reach. Three straight HOLD cycles confirm the freeze. The way out is a full exit of one position, which executes before any BUY and reopens the window. Candidate exits: MSFT is a +35.6% winner with only 8% analyst upside, the most stretched valuation in the book, and a cross-feed BUY that rests on analyst ratings alone. Selling it crystallizes the gain, which is disciplined profit-taking. AMZN and GOOGL are near breakeven with 28% and 22% analyst upside respectively, so selling them would abandon the stronger remaining theses. MSFT is the correct exit. Proceeds of about $542 plus $82 cash gives roughly $620 spendable. The best redeployment is NVDA: its fundamental upside is the highest in the universe (+44% analyst target) and it has pulled back from about $239 to $225 over the last week, a reasonable entry for a long-term holder. Two shares cost about $451, leaving ~$170 in cash buffer to avoid a re-freeze and cover fees. NVDA is cross-feed MIXED (1 BUY, 1 SELL), so this is a moderate-conviction rotation, not a high-conviction one. Both sides stay small relative to the 40% position cap. Rejected alternatives: META and TSLA report in 7 days, so no fresh positions before the print, and AAPL has a cross-feed SELL signal. AMZN and GOOGL are kept: both are near breakeven, and the 28% and 22% analyst upside plus BUY consensus support the long-term thesis. No new positions in META or TSLA ahead of the 2026-10-21 earnings, and no AAPL given the SELL consensus.

Horisont: 12-24 months Begrundelse: Consensus BUY with analyst 1.4/5 and ~22% target upside; quality blue-chip adds diversification, exit if thesis erodes or analyst upside collapses. === Session-analyse === [main/US] The portfolio is effectively frozen: 369 DKK cash (~$54 spendable) against a $40 fee floor, with the cheapest name (NVDA) at $227. No legal BUY exists today. The only way to act on any conviction is a full exit in this same basket, since SELLs execute before BUYs. I looked at the two candidates. Fully exiting MSFT (+34.6%, ~$538) would lock in a gain, but it is a quality compounder with 10% analyst upside and only mild cross-feed support (1 BUY, 2 HOLD). Selling it would also realize a taxable gain and pay fees just to rotate into names with similar or weaker conviction. That is not compelling. NVDA is the better rotation candidate: it is the weaker position (-4.4%, drifting down for 7 of the last 10 days, ~$908 of trapped capital) and its cross-feed verdict is MIXED (1 BUY, 1 SELL, 2 HOLD). Analyst target upside is large (+43%), but the screening shows no independent agreement beyond that single feed. The cleaner use of the proceeds is AMZN (analyst 1.3/5, +26% upside, consensus BUY) and GOOGL (+22% upside, consensus BUY). Both are diversifiers away from NVDA's semiconductor concentration, and I can buy several shares of each from the ~$908 plus ~$54 existing cash (~$962 total). Plan: fully exit NVDA (4 shares at ~$227 = ~$909), then buy 2 AMZN (~$525) and 1 GOOGL (~$355), totaling ~$880. That leaves ~$80 headroom for fees, FX and price drift. Position sizes stay well under the 40% cap. This frees trapped capital, tilts toward the names with the best fundamental support, and leaves MSFT, the strongest performer, untouched. Realizing the NVDA loss is modest (~-4%). MSFT is kept: +34.6%, consensus BUY, 10% analyst upside, and selling it would realize a gain without a clearly better replacement. Cash is rebuilt only to a ~$80 buffer after the basket, which is intentional to cover fees and FX drift.

Horisont: 12-24 months Begrundelse: Strongest cross-feed consensus with analyst 1.3/5 and ~26% target upside; diversifies away from semiconductor concentration. === Session-analyse === [main/US] The portfolio is effectively frozen: 369 DKK cash (~$54 spendable) against a $40 fee floor, with the cheapest name (NVDA) at $227. No legal BUY exists today. The only way to act on any conviction is a full exit in this same basket, since SELLs execute before BUYs. I looked at the two candidates. Fully exiting MSFT (+34.6%, ~$538) would lock in a gain, but it is a quality compounder with 10% analyst upside and only mild cross-feed support (1 BUY, 2 HOLD). Selling it would also realize a taxable gain and pay fees just to rotate into names with similar or weaker conviction. That is not compelling. NVDA is the better rotation candidate: it is the weaker position (-4.4%, drifting down for 7 of the last 10 days, ~$908 of trapped capital) and its cross-feed verdict is MIXED (1 BUY, 1 SELL, 2 HOLD). Analyst target upside is large (+43%), but the screening shows no independent agreement beyond that single feed. The cleaner use of the proceeds is AMZN (analyst 1.3/5, +26% upside, consensus BUY) and GOOGL (+22% upside, consensus BUY). Both are diversifiers away from NVDA's semiconductor concentration, and I can buy several shares of each from the ~$908 plus ~$54 existing cash (~$962 total). Plan: fully exit NVDA (4 shares at ~$227 = ~$909), then buy 2 AMZN (~$525) and 1 GOOGL (~$355), totaling ~$880. That leaves ~$80 headroom for fees, FX and price drift. Position sizes stay well under the 40% cap. This frees trapped capital, tilts toward the names with the best fundamental support, and leaves MSFT, the strongest performer, untouched. Realizing the NVDA loss is modest (~-4%). MSFT is kept: +34.6%, consensus BUY, 10% analyst upside, and selling it would realize a gain without a clearly better replacement. Cash is rebuilt only to a ~$80 buffer after the basket, which is intentional to cover fees and FX drift.

Horisont: n/a - full exit to unfreeze capital Begrundelse: Weakest holding with only mixed cross-feed support; full exit frees ~$908 trapped capital so conviction BUYs can execute, and bypasses the fee-guard. === Session-analyse === [main/US] The portfolio is effectively frozen: 369 DKK cash (~$54 spendable) against a $40 fee floor, with the cheapest name (NVDA) at $227. No legal BUY exists today. The only way to act on any conviction is a full exit in this same basket, since SELLs execute before BUYs. I looked at the two candidates. Fully exiting MSFT (+34.6%, ~$538) would lock in a gain, but it is a quality compounder with 10% analyst upside and only mild cross-feed support (1 BUY, 2 HOLD). Selling it would also realize a taxable gain and pay fees just to rotate into names with similar or weaker conviction. That is not compelling. NVDA is the better rotation candidate: it is the weaker position (-4.4%, drifting down for 7 of the last 10 days, ~$908 of trapped capital) and its cross-feed verdict is MIXED (1 BUY, 1 SELL, 2 HOLD). Analyst target upside is large (+43%), but the screening shows no independent agreement beyond that single feed. The cleaner use of the proceeds is AMZN (analyst 1.3/5, +26% upside, consensus BUY) and GOOGL (+22% upside, consensus BUY). Both are diversifiers away from NVDA's semiconductor concentration, and I can buy several shares of each from the ~$908 plus ~$54 existing cash (~$962 total). Plan: fully exit NVDA (4 shares at ~$227 = ~$909), then buy 2 AMZN (~$525) and 1 GOOGL (~$355), totaling ~$880. That leaves ~$80 headroom for fees, FX and price drift. Position sizes stay well under the 40% cap. This frees trapped capital, tilts toward the names with the best fundamental support, and leaves MSFT, the strongest performer, untouched. Realizing the NVDA loss is modest (~-4%). MSFT is kept: +34.6%, consensus BUY, 10% analyst upside, and selling it would realize a gain without a clearly better replacement. Cash is rebuilt only to a ~$80 buffer after the basket, which is intentional to cover fees and FX drift.

Horisont: 12-18 months Begrundelse: Strongest cross-feed consensus (3 BUY/0 SELL) with analyst target implying +38% upside, adding to existing position near recent range lows. === Session-analyse === [main/US] Cash is tight at ~$290 spendable. NVDA has the strongest cross-feed consensus (3 BUY, 0 SELL) with analyst upside of +38% to target, and the current price (234) is near recent lows within its range, offering a reasonable entry to average down on the existing position (currently -2.0%). AMZN also has a clean BUY consensus with +28% upside, but with only ~$290 available I can only afford one meaningful add without breaching the basket budget. Prioritizing NVDA since it adds to a quality AI-compute holding at a technically favorable level and deepens conviction with the strongest cross-feed agreement. Cash is nearly exhausted after this single NVDA add; holding MSFT (strong +32.5% gain, long-term AI/cloud thesis intact) and remaining NVDA shares. Will look to recycle cash from weaker DK positions in the next DK session if more US conviction signals emerge.

Horisont: n/a Begrundelse: Full exit of a modest, thesis-less winner to unlock trapped cash and break the buy-freeze per BUY WINDOW guidance. === Session-analyse === [main/DK] Cash is frozen at 699 DKK, below the fee-floor for any DK buy. The BUY WINDOW explicitly shows that a full exit of NOVO-B.CO would free ~1,241 DKK and reopen several legal buys this cycle, including DSV.CO and ORSTED.CO which both carry strong cross-feed BUY consensus with analyst upside of 70% and 32% respectively. NOVO-B.CO is only a modest +3.5% winner with no fresh catalyst, held briefly, and is the position explicitly flagged by the system as the unlock mechanism, so rotating out of it to deploy into higher-conviction names is reasonable rotation rather than panic-selling. Holding MSFT and NVDA for continued US growth exposure; after the NOVO-B.CO exit and small DSV.CO buy, remaining cash stays available for future DK opportunities like ORSTED.CO or GN.CO on further dips.

Horisont: 12-18 months Begrundelse: Strongest cross-feed BUY consensus (analyst target +37% upside) among AI infra leaders, adding to existing conviction position after the AMZN exit frees cash. === Session-analyse === [main/US] Cash is frozen at ~609 DKK with no legal US BUY possible (cheapest NVDA share is $238, well above the $90 spendable ceiling). The BUY WINDOW explicitly lists that a FULL EXIT of AMZN would unlock legal buys (AAPL or NVDA, 1 share). AMZN is my smallest USD position (1 share, ~253 DKK proceeds), has the weakest standalone conviction here relative to NVDA (strong cross-feed BUY consensus with 3 independent BUY-leaning feeds incl. analyst upside +29% and +37% for NVDA) and MSFT (large unrealized gain, strong thesis intact). Exiting AMZN is the least disruptive way to break the freeze — it's a tiny position, flat P&L, and full-exit bypasses the fee-guard cleanly. Holding MSFT (strong +31.6% gain, solid analyst target upside) and NOVO-B.CO (recent buy, positive momentum). Keeping LLY, AAPL, GOOGL, META, TSLA out of this cycle's trades — no fresh catalyst and limited capital after recycling AMZN into NVDA.

Horisont: n/a Begrundelse: Full exit of smallest, flattest USD position to unlock cash and break the buy-freeze per BUY WINDOW guidance. === Session-analyse === [main/US] Cash is frozen at ~609 DKK with no legal US BUY possible (cheapest NVDA share is $238, well above the $90 spendable ceiling). The BUY WINDOW explicitly lists that a FULL EXIT of AMZN would unlock legal buys (AAPL or NVDA, 1 share). AMZN is my smallest USD position (1 share, ~253 DKK proceeds), has the weakest standalone conviction here relative to NVDA (strong cross-feed BUY consensus with 3 independent BUY-leaning feeds incl. analyst upside +29% and +37% for NVDA) and MSFT (large unrealized gain, strong thesis intact). Exiting AMZN is the least disruptive way to break the freeze — it's a tiny position, flat P&L, and full-exit bypasses the fee-guard cleanly. Holding MSFT (strong +31.6% gain, solid analyst target upside) and NOVO-B.CO (recent buy, positive momentum). Keeping LLY, AAPL, GOOGL, META, TSLA out of this cycle's trades — no fresh catalyst and limited capital after recycling AMZN into NVDA.

Horisont: 1-2 years Begrundelse: Blue-chip pharma at a pulled-back price near recent lows, strong long-term GLP-1 franchise moat. === Session-analyse === [main/DK] The portfolio is cash-frozen: only 24.18 DKK free, and no BUY fits under the fee-guard floor. The BUY WINDOW explicitly identifies that a FULL EXIT of ORSTED.CO (14 shares, ~1,841 DKK) would unlock legal buys this cycle. ORSTED.CO has been flat/slightly negative since purchase (-0.3%, plus -271 DKK realized loss history on it), making it the weakest-conviction holding to recycle despite a nominal analyst BUY consensus — the position has already been round-tripped (sold DSV, bought ORSTED, both roughly flat) without a strong new catalyst, and at 14 shares it's a small, low-conviction allocation relative to the portfolio. Keeping AMZN, MSFT, NVDA as long-term US holdings (not tradeable this session). Exiting ORSTED.CO fully to unlock trapped cash, then redeploying into NOVO-B.CO and COLO-B.CO, both quality Danish blue-chips with reasonable entries, while keeping total new spend within the unlocked cash plus ~10% fee headroom.

Horisont: n/a Begrundelse: Full exit of a flat, low-conviction position to break the cash freeze and redeploy into higher-conviction names this cycle. === Session-analyse === [main/DK] The portfolio is cash-frozen: only 24.18 DKK free, and no BUY fits under the fee-guard floor. The BUY WINDOW explicitly identifies that a FULL EXIT of ORSTED.CO (14 shares, ~1,841 DKK) would unlock legal buys this cycle. ORSTED.CO has been flat/slightly negative since purchase (-0.3%, plus -271 DKK realized loss history on it), making it the weakest-conviction holding to recycle despite a nominal analyst BUY consensus — the position has already been round-tripped (sold DSV, bought ORSTED, both roughly flat) without a strong new catalyst, and at 14 shares it's a small, low-conviction allocation relative to the portfolio. Keeping AMZN, MSFT, NVDA as long-term US holdings (not tradeable this session). Exiting ORSTED.CO fully to unlock trapped cash, then redeploying into NOVO-B.CO and COLO-B.CO, both quality Danish blue-chips with reasonable entries, while keeping total new spend within the unlocked cash plus ~10% fee headroom.

Horisont: 6-12 months Begrundelse: Strong cross-feed consensus (3 BUY/0 SELL) with analyst target implying +38% upside, and it's the only trade size that fits current spendable cash. === Session-analyse === [main/US] Cash is very tight (~$245 spendable) and the BUY WINDOW confirms only 1 share of NVDA fits. NVDA has the strongest cross-feed signal of the session: 3 BUY votes, no sells, and analyst consensus showing +38% upside to target - a clear conviction filter pass. I already hold 1 NVDA share at a small gain, and adding 1 more is within position-size limits and fits the cash constraint exactly. Other BUY-rated names (AMZN, GOOGL, MSFT, LLY) all exceed the $245 spendable ceiling for even a single share, so NVDA is the only actionable legal trade this session. Holding AMZN, MSFT, and ORSTED.CO positions as long-term blue-chip conviction remains intact; MSFT shows strong unrealized gains and no reason to trim given its quality. Cash is nearly exhausted after the NVDA add, so no further buys are feasible this cycle without compromising other positions, and none of the existing holdings show weak enough conviction to warrant a sale for recycling capital right now.

Horisont: 1-2 years Begrundelse: Strong analyst upside (+30% to target) and BUY consensus across feeds, entering after a pullback from recent highs in a beaten-down renewable energy name. === Session-analyse === [main/DK] Portfolio is cash-frozen: only 99.28 DKK free, nowhere near the 1,160 DKK fee-guard floor for any DK name. The BUY WINDOW explicitly flags that a full exit of DSV.CO would free ~3,440 DKK and reopen legal buys across several blue-chip names. DSV.CO is currently my weakest-conviction DK holding here (-3.0% unrealized, small 3-share position, and while cross-feed shows BUY consensus with strong +67% analyst upside, the position is too small to trim partially under fee-guard rules — a partial sell would be rejected). Rather than do a full exit purely to chase GN.CO's mediocre signal (only +7% upside, 2.0/5 analyst rating), I'll use the freed capital to add to DSV.CO's own strong analyst thesis isn't available since I'd be closing it, so instead I'll rotate into the strongest cross-feed consensus name with the best risk/reward: ORSTED.CO (+30% upside, BUY consensus) and add to GN.CO modestly for diversification. Holding AMZN, MSFT, NVDA (US positions untouched this DK session). Exiting DSV fully to break the cash freeze per BUY WINDOW guidance, then redeploying into ORSTED (strongest analyst upside in the cross-feed screen) and GN.CO (cheapest fee-guard-clearing entry with BUY consensus) while keeping a cash buffer.

Horisont: n/a Begrundelse: Full exit of small, underwater DSV position to break the cash freeze and redeploy into names with stronger near-term upside signals. === Session-analyse === [main/DK] Portfolio is cash-frozen: only 99.28 DKK free, nowhere near the 1,160 DKK fee-guard floor for any DK name. The BUY WINDOW explicitly flags that a full exit of DSV.CO would free ~3,440 DKK and reopen legal buys across several blue-chip names. DSV.CO is currently my weakest-conviction DK holding here (-3.0% unrealized, small 3-share position, and while cross-feed shows BUY consensus with strong +67% analyst upside, the position is too small to trim partially under fee-guard rules — a partial sell would be rejected). Rather than do a full exit purely to chase GN.CO's mediocre signal (only +7% upside, 2.0/5 analyst rating), I'll use the freed capital to add to DSV.CO's own strong analyst thesis isn't available since I'd be closing it, so instead I'll rotate into the strongest cross-feed consensus name with the best risk/reward: ORSTED.CO (+30% upside, BUY consensus) and add to GN.CO modestly for diversification. Holding AMZN, MSFT, NVDA (US positions untouched this DK session). Exiting DSV fully to break the cash freeze per BUY WINDOW guidance, then redeploying into ORSTED (strongest analyst upside in the cross-feed screen) and GN.CO (cheapest fee-guard-clearing entry with BUY consensus) while keeping a cash buffer.

Horisont: 1-2 years Begrundelse: Strong cross-feed BUY consensus with analyst target implying +40% upside, entry near the stock's recent 10-day low. === Session-analyse === [main/US] Cash is tight (~$252 spendable). NVDA shows a strong consensus BUY with analyst upside of +40% and a reasonable entry near recent lows (~228-237 range), making it the strongest cross-feed signal this session. AMZN is also BUY-rated with +31% upside, but I already hold a full position at near cost, and buying another would concentrate further without much new diversification benefit. Given the single-share constraint from the BUY WINDOW, I'll deploy into NVDA which has the stronger consensus (2 BUY feeds including the analyst upside leader) and sits at a technical low point in its 10-day range, offering a better entry than chasing AMZN which just hit a 10-day high. Holding AMZN, DSV.CO, and MSFT — all intact long-term theses (MSFT +33% is a clear winner to let run, AMZN and DSV.CO are modest near-cost positions with no new deterioration signal). Cash remains minimal after the NVDA buy, consistent with the conservative capital-preservation mandate; no further US buys fit within the fee-guard floor this cycle.

Horisont: n/a Begrundelse: Full exit of recently-added position in a short-term downtrend (258.80 to 247.95 over 10 days) to unlock trapped cash and break the buy-freeze, keeping higher-conviction DSV.CO intact. === Session-analyse === [main/DK] Cash is frozen at 482 DKK — no legal BUY exists for any DK name at current cash levels, confirmed by the BUY WINDOW message. To break the freeze and redeploy capital, I need a FULL EXIT. DSV.CO is the smaller, newer position (3 shares, ~3,550 DKK) acquired recently with a thin loss (-0.9%), versus NOVO-B.CO which I just added to on conviction two sessions ago. However, DSV.CO carries the strongest cross-feed signal of the group (BUY consensus, analyst target +63% upside) — selling it would mean abandoning my best-rated name. Comparing the two unlock options: exiting DSV.CO frees ~3,518 DKK and opens the widest BUY range (CARL-B.CO, COLO-B.CO, DEMANT.CO, GN.CO, NOVO-B.CO and more), while exiting NOVO-B.CO only frees ~1,211 DKK with a narrower range. Given NOVO-B.CO has weaker recent price action (down from 258.80 to 247.95 over 10 days, a clear downtrend) and no cross-feed coverage shown here, while DSV.CO has strong analyst backing and I'd rather keep the higher-conviction name, I'll exit NOVO-B.CO fully to unlock cash and rotate into GN.CO which carries a BUY consensus with analyst upside. Holding AMZN, MSFT (US, not tradeable this session) and keeping DSV.CO given its strong cross-feed BUY signal (analyst target +63% upside) despite the small unrealized loss — it's the highest-conviction DK holding and shouldn't be sacrificed for cash recycling when a smaller, weaker-signal position (NOVO-B.CO) can serve that purpose instead.

Horisont: 12-18 months Begrundelse: Strong cross-feed BUY consensus with 33% analyst upside and improving price momentum (258 vs 246 ten days ago); deploy freed capital into the highest-conviction blue-chip name available. === Session-analyse === [main/US] Portfolio is cash-frozen at $88.95 spendable with no US name fitting the buy window directly. The BUY WINDOW mechanism indicates a full exit of NVDA or MSFT would unlock capital. NVDA is the better exit: cross-feed consensus is SELL (1 BUY/2 SELL/3 HOLD) despite a bullish analyst price target, it's my smallest US position, and I've only held it one session without a developed long-term thesis. MSFT remains a strong long-term holding with BUY consensus, +29% unrealized gain, and continued analyst support (+13% upside to target) — no reason to sell a quality compounder just to free cash. Holding DSV.CO, NOVO-B.CO, and MSFT unchanged — MSFT remains a high-conviction long-term compounder with BUY consensus and strong gains; the Danish positions are untouched this US-only session.

Horisont: n/a - full exit Begrundelse: Full exit of small NVDA position to break the cash freeze; cross-feed consensus is SELL despite high analyst target, and position is too small to add further conviction. === Session-analyse === [main/US] Portfolio is cash-frozen at $88.95 spendable with no US name fitting the buy window directly. The BUY WINDOW mechanism indicates a full exit of NVDA or MSFT would unlock capital. NVDA is the better exit: cross-feed consensus is SELL (1 BUY/2 SELL/3 HOLD) despite a bullish analyst price target, it's my smallest US position, and I've only held it one session without a developed long-term thesis. MSFT remains a strong long-term holding with BUY consensus, +29% unrealized gain, and continued analyst support (+13% upside to target) — no reason to sell a quality compounder just to free cash. Holding DSV.CO, NOVO-B.CO, and MSFT unchanged — MSFT remains a high-conviction long-term compounder with BUY consensus and strong gains; the Danish positions are untouched this US-only session.

Horisont: 1-2 years Begrundelse: Clean uptrend over last 10 sessions (250 to 260), blue-chip pharma quality at reasonable entry after recent pullback stabilized. === Session-analyse === [main/DK] Cash is frozen at 420 DKK — below the DK fee floor for any BUY, so no new position can be opened without first unlocking capital. The BUY WINDOW explicitly lists DSV.CO and ORSTED.CO full exits as the unlock paths. DSV.CO carries the strongest cross-feed conviction of all four flagged names (analyst upside +68%, BUY consensus) and is only a small 3-share position already at a loss (-2.8%) — not a name I want to abandon on a dip. ORSTED.CO, by contrast, is a larger, newer position (11 shares, bought just yesterday) sitting flat at -0.3%, with decent but not standout analyst upside (+28%), and exiting it frees ~1,456 DKK — enough to redeploy into higher-conviction names like DSV.CO (which I'd keep) or add to NOVO-B.CO which has shown a clean uptrend. Holding DSV.CO (strongest cross-feed conviction, +68% analyst upside target) despite its small loss, and holding MSFT/NVDA positions in the US market unchanged this session. Deploying freed ORSTED.CO cash into NOVO-B.CO which shows the clearest technical momentum in the DK universe.

Horisont: n/a - full exit Begrundelse: Full exit of a newly-opened flat position to unlock trapped cash and break the cash-freeze, redeploying into stronger-conviction names. === Session-analyse === [main/DK] Cash is frozen at 420 DKK — below the DK fee floor for any BUY, so no new position can be opened without first unlocking capital. The BUY WINDOW explicitly lists DSV.CO and ORSTED.CO full exits as the unlock paths. DSV.CO carries the strongest cross-feed conviction of all four flagged names (analyst upside +68%, BUY consensus) and is only a small 3-share position already at a loss (-2.8%) — not a name I want to abandon on a dip. ORSTED.CO, by contrast, is a larger, newer position (11 shares, bought just yesterday) sitting flat at -0.3%, with decent but not standout analyst upside (+28%), and exiting it frees ~1,456 DKK — enough to redeploy into higher-conviction names like DSV.CO (which I'd keep) or add to NOVO-B.CO which has shown a clean uptrend. Holding DSV.CO (strongest cross-feed conviction, +68% analyst upside target) despite its small loss, and holding MSFT/NVDA positions in the US market unchanged this session. Deploying freed ORSTED.CO cash into NOVO-B.CO which shows the clearest technical momentum in the DK universe.

Horisont: 1-2 years Begrundelse: Strongest cross-feed BUY consensus with 43% analyst upside target, entering on recent pullback to ~$225-231 range. === Session-analyse === [main/US] Cash is essentially frozen at $46 spendable, below the fee-floor for any name except the hypothetical post-sale scenarios shown. The BUY WINDOW explicitly notes a FULL EXIT of MSFT would free ~$519, opening legal buys across the board including NVDA, AMZN, GOOGL, and AAPL. However, MSFT is my strongest winner (+29.8%) with a BUY consensus and 13% analyst upside — selling it purely to unlock cash would abandon a quality compounder for no new thesis, which violates the long-term blue-chip mandate. AMZN, by contrast, is a smaller position (+0.7%, barely a day old), and a full exit of AMZN frees ~$248, enough to open exactly 1 share of NVDA per the BUY WINDOW guidance — satisfying the cash-recycling escape hatch while preserving my best long-term holding. Holding MSFT (strong winner, BUY consensus, 13% upside), DSV.CO and ORSTED.CO (DK positions untouched this US session) — no reason to abandon quality names; freed cash redeployed into the highest-conviction cross-feed BUY signal instead.

Horisont: n/a Begrundelse: Full exit of smallest, newest position to break the cash freeze per BUY WINDOW guidance, recycling capital into higher-conviction NVDA. === Session-analyse === [main/US] Cash is essentially frozen at $46 spendable, below the fee-floor for any name except the hypothetical post-sale scenarios shown. The BUY WINDOW explicitly notes a FULL EXIT of MSFT would free ~$519, opening legal buys across the board including NVDA, AMZN, GOOGL, and AAPL. However, MSFT is my strongest winner (+29.8%) with a BUY consensus and 13% analyst upside — selling it purely to unlock cash would abandon a quality compounder for no new thesis, which violates the long-term blue-chip mandate. AMZN, by contrast, is a smaller position (+0.7%, barely a day old), and a full exit of AMZN frees ~$248, enough to open exactly 1 share of NVDA per the BUY WINDOW guidance — satisfying the cash-recycling escape hatch while preserving my best long-term holding. Holding MSFT (strong winner, BUY consensus, 13% upside), DSV.CO and ORSTED.CO (DK positions untouched this US session) — no reason to abandon quality names; freed cash redeployed into the highest-conviction cross-feed BUY signal instead.

Horisont: 1-2 years Begrundelse: Cross-feed BUY consensus with analyst target +27% upside; renewable energy utility trading near multi-month lows offers attractive entry. === Session-analyse === [main/DK] Cash is tight (1,832 DKK) and the BUY WINDOW confirms limited options. Among the legal choices, ORSTED.CO stands out with a cross-feed BUY consensus and a strong analyst-rating upside (+27%), and it fits comfortably within spendable cash at 9-13 shares. DSV.CO also has BUY consensus with impressive +67% target upside, but I already hold 3 shares at a loss and adding only 1 share (the max allowed) wouldn't meaningfully improve the position while using significant cash. CARL-B.CO has mixed signal (1 BUY, 1 HOLD) and only 2 shares fit, which is a small allocation for the capital used. Holding DSV.CO (already own 3 shares, recent realized loss, limited capital to add meaningfully), AMZN and MSFT (US holdings untouched this DK session). Deploying available cash into ORSTED.CO which has the strongest cross-feed conviction and fits the buy window without over-concentrating in any single name.

Horisont: 1-2 years Begrundelse: Strongest cross-feed conviction among affordable names post-exit (BUY consensus, analyst target +34% upside) with reasonable entry near recent range lows. === Session-analyse === [main/US] Cash is essentially frozen at ~1,186 DKK ($179), too small to buy even the cheapest US name (NVDA at $230). The BUY WINDOW analysis shows a full exit of GOOGL (single share, +2.0% P&L, modest analyst upside +26%) would unlock enough cash to open several legal buys this cycle, while MSFT (+28.5% unrealized gain, strongest position, analyst target still +13% upside with BUY consensus) is the stronger long-term holding to keep. GOOGL is a single share with a thin thesis at this point — closing it is the cleanest recycling move since it fully bypasses the fee-guard and unlocks the freeze without disturbing the better-performing MSFT position. Keeping MSFT (strong momentum, +28.5% gain, still BUY-rated with upside) and DSV.CO untouched. After the GOOGL exit and AMZN buy, remaining cash stays as reserve; no other US names fit the fee-optimal window this cycle.

Horisont: N/A - exit Begrundelse: Full exit of single-share GOOGL position to break the cash freeze and recycle capital into a higher-conviction buy this session. === Session-analyse === [main/US] Cash is essentially frozen at ~1,186 DKK ($179), too small to buy even the cheapest US name (NVDA at $230). The BUY WINDOW analysis shows a full exit of GOOGL (single share, +2.0% P&L, modest analyst upside +26%) would unlock enough cash to open several legal buys this cycle, while MSFT (+28.5% unrealized gain, strongest position, analyst target still +13% upside with BUY consensus) is the stronger long-term holding to keep. GOOGL is a single share with a thin thesis at this point — closing it is the cleanest recycling move since it fully bypasses the fee-guard and unlocks the freeze without disturbing the better-performing MSFT position. Keeping MSFT (strong momentum, +28.5% gain, still BUY-rated with upside) and DSV.CO untouched. After the GOOGL exit and AMZN buy, remaining cash stays as reserve; no other US names fit the fee-optimal window this cycle.

[stop-loss triggered at $152.35 — broker-fill $152.42]

[stop-loss triggered at $160.90 — broker-fill $160.88]

Horisont: 1-2 weeks Begrundelse: WSB has 4 active SPCX posts today with bullish skew and disclosed positions ('doubled portfolio, adding SPCX'), cross-feed consensus BUY with analyst +32% upside — exit if the thesis fades or it breaks below $137 support. === Session-analyse === [saxomoon/US] Portfolio is tiny ($538) with RDDT and UBER already at cap, cash at 46%. Looking at today's setups: AMD and NVDA are both UNTRADEABLE at this risk budget per the server (too expensive for min-entry/risk-sizing), despite AMD trending with accelerating mentions and a hot WSB post on 2027 supply (net-sentiment +0.92, cross-feed consensus BUY). SPCX is heavily discussed (4 WSB posts, YOLO culture, cross-feed consensus BUY with analyst upside) but the gulv-stop is 8.3% and max 1 share (~$176) — that clears the floor and fits the risk budget, and SPCX gapped down slightly overnight (not chasing a pump). MSTR is untradeable at this budget (0 shares). APLD earnings in 1 day is a real catalyst but earnings volatility the day before a print is exactly when NOT to open fresh — per earnings-watch guidance, skipping. SPCX stands out: tradeable, cross-feed consensus BUY (analyst target +32% upside), WSB has 4 posts today with bullish skew (3 bull/1 bear) and real skin-in-the-game ('Portfolio doubled, trimming MSTR to add SPCX and TSLA'), RISK-ON regime (QQQ +0.46%), VIX calm at 15.3, Crypto Fear&Greed at 73 (greed but not extreme). This is a 1-share entry at a name priced well above $50 so the min-shares exception applies. I'll size exactly to the gulv-stop floor (1 share, ~$176, 8.3% stop) which clears the fee-guard floor comfortably. Holding RDDT and UBER unchanged — both are CAP-FULL per server params, no room to add. RDDT is down -2.2% but cross-feed still shows consensus BUY (2 BUY, 0 SELL, 3 HOLD) with analyst upside +42%, so rule #4's exception keeps it alive despite WSB flair being ironic/bearish (1 post, net -1.50, puts position) — that's a single ironic post, not real signal (rule #15: thin sample = zero signal). UBER barely moved, thesis intact, ride_hailing/platform/autonomy theme is fine at 26% each, nowhere near the 50% cap. No stop changes needed this cycle — both positions are near flat, ratchet only makes sense once there's real gain to protect (RDDT's resting stop at 6.1% under and UBER's at 5.8% under are both already inside this cycle's noise bands, tightening would just donate to volatility). RDDT and UBER are both CAP-FULL per server constraints — no room to add regardless of conviction. AMD, NVDA, MSTR, MU are all UNTRADEABLE at this risk budget despite WSB chatter (AMD accelerating mentions, NVDA mega-hot meme post) — the server's gulv-stop sizing math simply doesn't clear a whole share at this account size. APLD has earnings tomorrow — skipping a fresh entry into that volatility per the earnings-watch guidance, even though WSB loss-porn posts are active on it. Sitting mostly in cash (46%) is appropriate given the small book size; only SPCX cleared both the WSB-sentiment bar and the cross-feed/risk-sizing floor this cycle.

[stop-loss triggered at $111.37 — broker-fill $111.52]

Horisont: 1-2 weeks Begrundelse: Fresh 318-comment WSB DD post ('follow the smart money') on UBER, risk-on tape (QQQ +0.76%), tight gulv-stop of 4.1% fits the risk budget cleanly. === Session-analyse === [saxomoon/US] MU is the day's clearest WSB catalyst — reports today, JPMorgan calling for beat-and-raise, a $50K earnings YOLO post, and the interpreted signal shows net-sentiment +0.94 with actual disclosed positions (calls+shares), not drive-by hype. But MU itself is untradeable at this account size (>$1000/share, min-entry would blow the $270 budget and the position cap), so I can't act on that thesis directly. UBER has a fresh 318-comment DD post ('follow the smart money') plus a calm gulv-stop (4.1%) that actually fits the risk budget — but it's flagged CAP-FULL... wait, checking again, UBER is NOT in the cap-full list (only HOOD/RDDT are), so it's tradeable. Cross-feed doesn't cover UBER explicitly but the WSB DD is fresh and RISK-ON regime plus VIX calm at 16 supports leaning in on a reasonable size. HOOD and RDDT are both CAP-FULL — no adding. RDDT has a fresh WSB post (net-sentiment +1.00, fundamental thesis, disclosed calls) plus cross-feed consensus BUY (analyst target +47% upside) — this supports holding, not cutting, despite the -3.8% P&L; rule #4's cross-feed exception applies. HOOD is red on both WSB attention (not mentioned) and overnight gap (-1.8%), down -8.3% — but it's a small single-share position where a full exit would only marginally help and no cross-feed BUY line was surfaced for HOOD in this session's block, meaning weaker support than RDDT. I'll tighten stops on both to protect capital given no fresh bullish catalyst. Given cash is thin (~$270 spendable) and both current holdings are cap-full, the only actionable slot is a fresh UBER position sized to the risk budget (~2 shares, gulv-stop 4.1%). MU thesis is real but inaccessible; I'll skip forcing it into an untradeable name. No crypto-mining exposure add given the -6.6% 5d tape-fade and prior -620 DKK theme history (theme-facit discipline holds). HOOD and RDDT are both position-cap-full so no adds are possible regardless of conviction. RDDT keeps its cross-feed BUY consensus (analyst +47% target upside) and a fresh WSB post with disclosed calls-positioning, so I'm holding per rule #4's exception rather than cutting the -3.8% drawdown. HOOD lacks that cross-feed support and is down -8.3% with no fresh WSB mention, but a full exit on a single share isn't worth the churn versus tightening the stop close to current levels to cap further downside. MU is the session's strongest WSB catalyst (earnings today, JPMorgan beat-and-raise call, $50K YOLO post) but is priced entirely out of reach for this account's budget — noted for tracking, not actionable. Sitting mostly in cash given thin spendable capital and two cap-full core positions; will reassess at 14:00 for any post-MU-earnings reaction that opens new opportunities in adjacent semiconductor names.

Horisont: 1-2 weeks Begrundelse: WSB YOLO post ($146K RDDT Jan2028 calls, net-sentiment +0.75) plus cross-feed consensus BUY (analyst target +40% upside) — exit if momentum thesis fades or price breaks below the $140 shelf. === Session-analyse === [saxomoon/US] Small book ($553), mostly cash (79%), one small HOOD position down -4.5%. WSB is dominated by NKE (top post, 15k score) but the actual sentiment reading is negative loss-porn/joke content ('The only way to rescue Nike'), not a bullish catalyst — the interpreted signal is net -0.25 with 100% loss-porn framing. ORCL has 2 mentions and interpreted +0.50 but only 1 post, thin sample = near-zero signal per rule #15, and the force-majeure headline is a real overhang even though cross-feed doesn't cover it directly. RDDT has a genuine YOLO ($146K calls post) plus cross-feed consensus BUY (analyst target +40% upside) — best combo of WSB hype + cross-feed agreement I see today, but gulv-stop is 8.5% and budget only allows 1 share (~$153), which clears the $100 min-trade floor barely and is a real dollar position at this price tier (rule #3 exception for ≥$50 names). META also shows interpreted +0.92 with disclosed 3-share position and cross-feed consensus BUY, but it's UNTRADEABLE at min size per the params (too expensive even at 1 share within budget) — skip. HOOD is CAP-FULL, can't add; it's down -4.5% but only mentioned in comment-flow, not fresh WSB hype, and no cross-feed shown for it directly — I'll leave the stop where it is since the daily noise band (±3.3%) means current 6.9% stop is already outside noise, no need to tighten yet given it's a small loss, not a winner to ratchet. Given QQQ is neutral/flat, VIX calm (14.21), and most of the loud posts today are jokes/memes (NKE rescue meme, Bessent meme, Bears meme) rather than fresh actionable catalysts, I'm taking one modest, cross-feed-supported bet: RDDT at 1 share with a stop anchored to the gulv-stop floor. Everything else — ORCL (thin/negative catalyst), NKE (loss-porn meme), AMZN (mixed/ironic, tiny sample) — fails the signal-quality bar. Sitting mostly in cash remains correct; this is a small nudge, not a big swing, consistent with the tiny book size and the need to avoid manufacturing activity (rule #6). HOOD stays as-is — CAP-FULL means no add possible, and it's down only -4.5% with the daily noise band (±3.3%) already narrower than the current 6.9% stop, so no ratchet needed on a small loser. NKE's top post is loss-porn/meme framing (interpreted net -0.25, 100% loss-porn) despite the huge score, so it fails the signal-quality bar (rule #15) even though it's technically trending. ORCL's single post is too thin a sample (rule #15: 1-2 posts = zero signal) and carries a real negative catalyst (force-majeure headline) working against it. AMZN's interpreted read is mixed/ironic with a tiny sample too. Cash stays the default given QQQ neutral, VIX calm but no name clearing both the WSB-freshness and cross-feed-agreement bar convincingly enough for more than the one RDDT nibble.

[stop-loss triggered at $22.51 — broker-fill $22.52]

Horisont: 2-4 weeks Begrundelse: WSB post citing $300-by-2027 HOOD target on BTC/SEC catalyst, cross-feed shows no SELL votes and RISK-ON tape; sized to 1 share given last-cycle overnight risk and floor-stop constraints. === Session-analyse === [saxomoon/US] Last cycle of the day, RISK-ON (QQQ +2.81%), VIX calm at 15.4, crypto Fear&Greed at 70 (Greed). RIOT is our only holding, up +14% and crypto_mining tape is ripping (+17.5% 5d basket). Rule #2/#13 says let it run, don't trim — ratchet the stop instead. RIOT's floor-stop from the server sits at 11.6% under price (avg dagsmove ±4.3%), and current resting stop is already 22.10 (9.4% under, tighter than floor doesn't apply since floor is a minimum not max) — actually the floor requires stop >= 11.6%, so current 9.4% stop is inside the noise floor per the server's own gulv-stop rule. I'll widen protection isn't allowed (raise-only, and 11.6% is WIDER than current 9.4%, which the engine would reject as a widen). So I leave RIOT's stop as-is this cycle since I can't legally tighten it below the noise floor, and the ladder (+15% -> 12% trail) is roughly where the floor already sits — no action needed, current stop stands. On new entries: it's cycle 4/4, last cycle before close, so any BUY carries overnight risk with a weakened stop (44% of stop-fires historically hit in the first 30min after open). AMD/META/MU/INTC are the loudest WSB names today (AMD hit $1T, META getting mixed loss-porn love, MU spreads gain post) but ALL are on the UNTRADEABLE-at-min-size list (AMD, META, MU) or have a floor-stop too wide for budget (INTC gulv 8.9%, max 1 share ~$121 — technically tradeable but marginal). HOOD is trending (+2 mentions, NEW/RISING) with a clean cross-feed absence of SELL votes and a WSB post citing $300 2027 target, floor-stop 11.2%, max 1 share (~$124) — that's a real dollar position at $124/share (>=$50 threshold allows 1-2 share entries) and clears the $100 min-trade floor. Risk-sizing at 3% (~$17) against an 11.2% floor stop only supports ~$150 notional, so 1 share (~$124) fits within budget and floor. However, given it's the LAST cycle of the day, a fresh 1-share HOOD position needs a reason to hold overnight — the WSB catalyst ($300 target by 2027) is a slow-burn thesis, not an urgent overnight one, and the position size (1 share, ~$124) is small enough that overnight gap risk is limited in dollar terms. I'll take a small starter HOOD position with a wide stop reflecting the floor, sized down deliberately per rule #19. No other name clears both the fee-guard floor and the risk-budget cleanly at this session's cash level. Sitting mostly in cash/RIOT into the close with a small HOOD nibble is the disciplined call — RIOT is already my crypto_mining exposure at 21%, well under the 50% cap, and adding more mining here into a Monday close after a big already-run 5-day tape (+17.5%) would be chasing, not investing. RIOT's current stop (9.4% under price) sits inside the server's own noise-floor (11.6% gulv-stop for RIOT), so I cannot legally raise it further without violating the raise-only/floor logic — leaving it unchanged rather than manufacturing a move. The loudest WSB names today (AMD, META, MU, INTC) are either UNTRADEABLE at this account's min size or only offer 1-share entries at marginal risk-budget fit; AMD/META also show mixed-to-bearish INTERPRETED signal (loss-porn 50-60%, META net-sentiment near zero with 4 bear posts) which per rule #15 lowers conviction rather than raising it. Crypto_mining tape has already run +17.5% in 5 days — chasing more RIOT/mining exposure into the close would be adding to a stretched move, not fresh conviction. Staying mostly in cash with the existing RIOT position and one small HOOD nibble reflects the last-cycle discipline (#19): don't force size into positions that need to survive an unpaid overnight session on thin catalysts.

[stop-loss triggered at $10.10 — broker-fill $10.01]

Horisont: 1-3 weeks Begrundelse: Cross-feed consensus BUY (analyst target +41% upside) on a cheap name gapping up overnight (+3.3%) in a risk-on tape, uncorrelated to my crypto_mining/ai_infra overweight — exit if the gap fades and it drops back below $10. === Session-analyse === [saxomoon/US] WSB's hot list today is dominated by macro loss-porn (bond market chaos, CPI, treasury debt repurchases) rather than a fresh single-name squeeze thesis — the actual meme content (AAPL calls meme, oil superglut, MSTR Jordans joke) is more comedic than a directional catalyst. That said, the cross-feed screen is unusually constructive: MARA, RIOT, CLSK, WULF all show consensus BUY with short-interest and technical strength (RIOT +12.4% 5d, CLSK +13.0% 5d, WULF +120% analyst upside), and QQQ/SPY are green (+0.93%/+0.99%) on a RISK-ON day — exactly the regime where WSB momentum names pay. My own theme-facit shows crypto_mining has been a net loser (-717 DKK over 9 exits) which is a real scar, but that history is on MARA/RIOT/CLSK/WULF specifically during choppier tape — today's setup is different: multiple names in the group are simultaneously breaking out with short-interest fuel (MARA 32.7% short, ratio 2.4) and analyst upside, which is a materially stronger multi-feed confirmation than prior attempts. Given the crypto_mining scar tissue, I'll size small and concentrated rather than spreading across all four miners (that would just be one correlated bet wearing four costumes per the theme-facit rule). RIOT has the cleanest signal: 4 BUY votes, strong yfinance momentum reading (range_pos 71%, +12.4% 5d) and it's not one of my repeat-loser names in the current session's overnight data. I'll also take a small NOK position — cross-feed BUY (analyst target +41% upside), cheap enough for full share-count minimums, uncorrelated to the crypto_mining/ai_infra themes I'm already scarred on, giving portfolio diversification. Cash is tiny (~$556 spendable) so I'm keeping this to 2 disciplined, fee-efficient entries rather than forcing 4-8 trades — rule #6 applies: don't manufacture activity, and with such a small book, concentration in the best 2 ideas beats diluting into weak ones. GME's insider buying (edgar +0.85, $21.7M in insider purchases) is tempting but GME is not a fresh trending name today (mentions declining) and the crypto-mining allocation already uses my risk budget for meme/momentum plays this cycle. VIX is calm at 16.46 (not elevated), Fear&Greed is 56 (neutral-greed, no extreme), and QQQ/SPY are both green — this is the risk-on regime rule #11 says to lean into. I'm avoiding SOUN (just stopped out -10.2%, in cooldown per rule #8) and PLTR (recent exit, needs >=0.7 conviction + 2 cross-feed votes — PLTR cross-feed isn't even listed here with dual-vote confirmation, so pass). Sitting mostly in cash beyond the two small entries above — WSB's hot list today is macro-loss-porn (bond chaos, CPI, treasury repurchases) rather than a single-name squeeze thesis, so I'm not forcing size into names just because they're mentioned (AAPL, TSLA, MP mention counts are macro-driven commentary, not a fresh WSB catalyst on the stock itself). PLTR and SOUN are both in re-entry cooldown from recent exits and don't clear the required 2-cross-feed-vote bar cleanly enough to justify re-entry today. Crypto_mining as a theme has a real -717 DKK scar over 9 prior exits, so I deliberately only took one name (RIOT, the strongest single signal) rather than spreading across MARA/CLSK/WULF too, keeping the correlated bet-sizing sane. Will reassess at 11:30 ET as the sub fully wakes up.

Horisont: 1-2 weeks Begrundelse: Cross-feed consensus BUY (4-1) with strong 5d momentum (+12.4%, range_pos 71%) and high short-interest fuel; risk-on QQQ tape favors miners — exit if RIOT loses the $19 shelf or WSB/cross-feed flips. === Session-analyse === [saxomoon/US] WSB's hot list today is dominated by macro loss-porn (bond market chaos, CPI, treasury debt repurchases) rather than a fresh single-name squeeze thesis — the actual meme content (AAPL calls meme, oil superglut, MSTR Jordans joke) is more comedic than a directional catalyst. That said, the cross-feed screen is unusually constructive: MARA, RIOT, CLSK, WULF all show consensus BUY with short-interest and technical strength (RIOT +12.4% 5d, CLSK +13.0% 5d, WULF +120% analyst upside), and QQQ/SPY are green (+0.93%/+0.99%) on a RISK-ON day — exactly the regime where WSB momentum names pay. My own theme-facit shows crypto_mining has been a net loser (-717 DKK over 9 exits) which is a real scar, but that history is on MARA/RIOT/CLSK/WULF specifically during choppier tape — today's setup is different: multiple names in the group are simultaneously breaking out with short-interest fuel (MARA 32.7% short, ratio 2.4) and analyst upside, which is a materially stronger multi-feed confirmation than prior attempts. Given the crypto_mining scar tissue, I'll size small and concentrated rather than spreading across all four miners (that would just be one correlated bet wearing four costumes per the theme-facit rule). RIOT has the cleanest signal: 4 BUY votes, strong yfinance momentum reading (range_pos 71%, +12.4% 5d) and it's not one of my repeat-loser names in the current session's overnight data. I'll also take a small NOK position — cross-feed BUY (analyst target +41% upside), cheap enough for full share-count minimums, uncorrelated to the crypto_mining/ai_infra themes I'm already scarred on, giving portfolio diversification. Cash is tiny (~$556 spendable) so I'm keeping this to 2 disciplined, fee-efficient entries rather than forcing 4-8 trades — rule #6 applies: don't manufacture activity, and with such a small book, concentration in the best 2 ideas beats diluting into weak ones. GME's insider buying (edgar +0.85, $21.7M in insider purchases) is tempting but GME is not a fresh trending name today (mentions declining) and the crypto-mining allocation already uses my risk budget for meme/momentum plays this cycle. VIX is calm at 16.46 (not elevated), Fear&Greed is 56 (neutral-greed, no extreme), and QQQ/SPY are both green — this is the risk-on regime rule #11 says to lean into. I'm avoiding SOUN (just stopped out -10.2%, in cooldown per rule #8) and PLTR (recent exit, needs >=0.7 conviction + 2 cross-feed votes — PLTR cross-feed isn't even listed here with dual-vote confirmation, so pass). Sitting mostly in cash beyond the two small entries above — WSB's hot list today is macro-loss-porn (bond chaos, CPI, treasury repurchases) rather than a single-name squeeze thesis, so I'm not forcing size into names just because they're mentioned (AAPL, TSLA, MP mention counts are macro-driven commentary, not a fresh WSB catalyst on the stock itself). PLTR and SOUN are both in re-entry cooldown from recent exits and don't clear the required 2-cross-feed-vote bar cleanly enough to justify re-entry today. Crypto_mining as a theme has a real -717 DKK scar over 9 prior exits, so I deliberately only took one name (RIOT, the strongest single signal) rather than spreading across MARA/CLSK/WULF too, keeping the correlated bet-sizing sane. Will reassess at 11:30 ET as the sub fully wakes up.

[stop-loss triggered at $6.33 — broker-fill $6.34]

Horisont: 1-3 weeks Begrundelse: Cross-feed consensus BUY driven by extreme 39.9% short-interest-of-float (ratio 4.4) — squeeze fuel independent of today's thin WSB chatter; exit if short interest unwinds or price breaks below the recent 6.78 range floor. === Session-analyse === [saxomoon/US] WSB's hot list today is almost entirely consumed by AVGO's post-earnings guidance drop (posts #4,7,11,13) and macro-fear megathread noise (50% bull, 'fear' undertone dominated by rates/tariffs chatter) — none of which is actionable: AVGO is explicitly untradeable at this account's risk budget, and chasing a -5.7% overnight gap into a crowded, SPIKE-CHRONIC name is exactly the setup rule #1 warns against (no conviction bump, tighten-not-add). IREN shows a fresh/rising WSB mention plus cross-feed BUY consensus (analyst target +97% upside) but the strategy params flag it explicitly UNTRADEABLE at this risk budget (13.2% floor-stop only allows 2 shares vs 3 minimum) — skipping despite the pull. The one name clearing the bar is SOUN: cross-feed consensus BUY led by short_interest at an extreme 39.9% of float (ratio 4.4) — genuine squeeze fuel — with a calm, tight daily range (6.78-7.32 over 10 days) that fits comfortably inside the risk budget without needing a wide stop. It's not in today's top WSB mentions, so this is a cross-feed/short-interest-led trade, not a hype chase, sized small and given a stop anchored just under the recent range floor. RISK-ON tape (QQQ +0.47%, VIX 16.3 calm) supports leaning in modestly rather than sitting fully passive, but with genuinely thin fresh catalyst evidence elsewhere, cash preservation is the right call for the rest of the book — no reason to force trades into AVGO's crowded earnings drama or space-theme names lacking a name-specific catalyst beyond a generic fedreg mention. Sitting mostly in cash this cycle. Today's WSB hot list is dominated by AVGO's earnings/guidance drama, which is untradeable at this account's min-size math and is a SPIKE-CHRONIC crowded name post-drop — not a fresh setup. IREN has a rising mention and cross-feed BUY backing but is explicitly flagged untradeable at the current risk budget (floor-stop too wide for min shares). Space names (RKLB/ASTS) ride only a generic fedreg mention with no name-specific WSB catalyst and a weak 30d theme tape (-10.9%). No name in the RECENT EXITS list (PLTR, MRNA) shows the required 2+ fresh cross-feed BUY votes plus changed thesis to justify re-entry. Preserving capital for a session with clearer, larger WSB conviction rather than forcing size into thin or blocked setups.

[stop-loss triggered at $171.31 — broker-fill $171.30]

[stop-loss triggered at $142.78 — broker-fill $142.64]

Horisont: 3-7 days Begrundelse: WSB's #1 gain-porn story today (cancer vaccine breakthrough, $500k overnight thread, $130k-in-10s thread) with cross-feed consensus BUY on the +9% overnight gap and RISK-ON tape — but chasing a post-gap biotech binary event, so sized to 1 share with a wide stop given the stock's huge realized volatility (63→174→139 over 10 sessions). === Session-analyse === [saxomoon/US] MRNA is the clear WSB story today — cancer vaccine breakthrough, +9% overnight gap, $500k gain-porn thread, $130k-in-10-seconds thread, huge megathread comment volume (40✓), and cross-feed consensus BUY (yfinance +1.00 on the 5d move). But the INTERPRETED read is mixed-to-cautionary: one loss-porn post (holder wants to exit), thin sample, and it's not in the TRENDABLE UNIVERSE per the trading universe list — MRNA IS in the tradeable universe though, so that's not a blocker. Given RISK-ON regime (QQQ +0.87%), calm VIX (15.13), and a name up huge already off a real fundamental catalyst (Phase 3 cancer vaccine data, not just meme hype), this is a rare case of a genuine news-driven mover the WSB crowd is chasing correctly. However, entering AFTER a +9% gap into a name already up ~50% this month is real chase risk per rule #16 — I'll size modestly and use a wider stop to respect the volatility (biotech binary-event stocks gap hard), not full conviction size. PLTR is my only current holding, +14.4%, sub-theme (ai_infra/defense_tech) already AT CAP so no adds regardless of conviction — the INTERPRETED read on PLTR is actually net-bearish (-1.50, puts thesis) which reinforces NOT adding. Per the trailing-stop ladder, +14.4% crosses toward the +15% tier (trail ~12% under price) — I'll tighten the stop from its current 7.4%-under level... wait, current stop is already tighter (7.4% under) than the 12% ladder guidance would imply, so I leave it as-is since the engine is raise-only and the current stop is already tighter than what the ladder step would set. Actually checking: ladder at +15% suggests 12% trail, but current stop is 7.4% under — that's already tighter than 12%, so no update needed, it's already better protected. No other WSB names clear the bar today — NVDA has earnings tomorrow (avoid fresh entries into print) and INTERPRETED read is bearish (puts, AI-hype-dead thesis) despite cross-feed mixed. SOXL is bleeding (-9% 5d, -18.7% 30d) with a loss-porn post — pure fade. WEN's short-squeeze story (43% short float) is interesting but only 1 mention and no real catalyst beyond a CEO quote — passing given fee-drag concerns on cheap-ish names. Sizing MRNA conservatively given the parabolic gap and biotech binary-event volatility (10-day closes show a wild recent range from ~$63 to $174 to now $138 — huge realized vol, needs a wide stop per volatility anchor rule). PLTR is capped at max sub_theme concentration (ai_infra/defense_tech both AT CAP) so no adds are possible regardless of the +2.1% overnight pop — and its current stop (7.4% under price) is already tighter than the ladder's +15%-tier guidance (12% trail), so no stop update needed this cycle. Rest of the WSB feed is meme/discussion noise (Buffett memes, Wendy's CEO quotes, Walmart) with no tradeable edge; NVDA and MRVL both report earnings within 48 hours so fresh entries there are avoided per the earnings-watch rule, and NVDA's INTERPRETED read is net-bearish (puts, AI-hype-dead thesis) despite a mixed cross-feed. SOXL and crypto-mining names (MARA/RIOT/CLSK/WULF) carry a documented -732 DKK 30-day loss history in this exact theme with no new external catalyst — theme-facit re-entry bar not cleared, staying out. Sitting mostly in cash preserves the $413 spendable for a cleaner setup next session.

[stop-loss triggered at $43.02 — broker-fill $43.02]

[stop-loss triggered at $164.28 — broker-fill $164.26]

Horisont: 1-2 weeks Begrundelse: S&P 500 inclusion Aug 18 is a concrete catalyst (interpreted net +1.50, 2/2 high-conviction posts) with independent cross-feed consensus BUY (4-0-1); sizing small given the +13% overnight gap already priced in much of the pop. === Session-analyse === [saxomoon/US] RDDT is the clearest setup on the board today: S&P 500 inclusion catalyst (Aug 18), interpreted net-sentiment +1.50 with 2/2 high-conviction posts citing a concrete catalyst, and the cross-feed consensus is a clean BUY (4 BUY/0 SELL/1 HOLD) — genuinely independent corroboration beyond just the Reddit-derived wsb vote. It's not in the raw trending-tickers title list, but it's front-page news (#11, ▲2,518) and the interpreted block flags it explicitly. It gapped +13% overnight into the open which is chase-risk, but a real fundamental catalyst (index inclusion forces passive buying next week) justifies paying up modestly with a wider stop rather than skipping it entirely. IREN also screens well (short 28.4% of float, cross-feed consensus BUY, interpreted net +1.50 fundamental thesis on the MSFT 50MW delivery) but it's pulled back hard from highs (45.93→38.89 over 10 days) — I'll pass this cycle given limited cash and prefer concentrating in the stronger single idea. My IONQ (+16.7%) and PLTR (+13.8%) positions are working — both cross-feed consensus BUY, both winners I should let run per rule #2 rather than trim. IONQ's ticker-facit is clean (no prior losses) and analyst_ratings backs it strongly (+50% upside target). PLTR is at position-cap so no add possible anyway, but I'll tighten both trailing stops to lock in gains per the ladder (IONQ +16.7% → trail ~12%; PLTR +13.8% → trail ~12%), respecting the gap-floor (avg dagsmove ~2.9% and ~2.3% respectively, well under the ladder levels so no floor conflict). Cash is thin (~$303 spendable) and QQQ/SPY are dead neutral today (+0.05%/+0.01%) — no regime tailwind either way. VIX is calm at 14.55 so no vol-based caution needed. I'm avoiding RIOT (recent stop-out loss, no independent 2-feed re-entry corroboration beyond wsb/short_interest overlap concerns) and skipping SMR/OKLO/other nuclear/quantum names to avoid over-concentrating further in quantum (already 44%) or adding new themes without conviction. One focused RDDT buy is the disciplined play this cycle. IONQ and PLTR are winners running per plan — letting them ride with tightened trailing stops rather than trimming. PLTR is CAP-FULL so no further add possible regardless. Cash is limited (~$303 spendable) so only one high-conviction fresh idea (RDDT) makes sense this cycle; RIOT stays excluded per re-entry cooldown lacking 2-independent-feed corroboration, and I'm not chasing quantum/nuclear names (OKLO/SMR/LEU) to avoid stacking further into an already-44%-concentrated quantum theme.

[stop-loss triggered at $19.26 — broker-fill $19.26]

Horisont: 1-3 weeks Begrundelse: WSB post on Anthropic's $9B cloud deal with Riot Platforms plus cross-feed consensus BUY (analyst target +53% upside, short_interest fuel) — exit if the deal narrative fades or RIOT breaks below $19.40 support. === Session-analyse === [saxomoon/US] Small cash pile (~$320 spendable) and both existing positions already green — IONQ +9.7%, PLTR +8.9% — with PLTR at cap so no add there. WSB's own thread on PLTR (Spider-Man/Palantir post, +34% move) is real but the position is already at the server's CAP-FULL limit, so this is a hold-and-trail situation, not an add. RIOT is the one fresh trending idea today: NEW tag on the WSB extraction (Anthropic's $9B cloud deal headline, ▲762), and it clears cross-feed with 4 BUY votes including analyst_ratings +0.87 (target +53% upside) — genuinely independent corroboration beyond just Reddit hype. Regime is RISK-ON (QQQ +0.87%), VIX calm at 15.46, so leaning into the best-supported fresh name fits the tilt. CRWV and NBIS both gapped huge overnight (+19%, +17%) but NBIS reports earnings TODAY (0 days out) — chasing a pre-print gap into an earnings print is exactly the kind of coin-flip the rules warn against, and CRWV isn't even in the tradeable universe. Passing on both despite the momentum. MARA/CLSK/WULF all show strong short-interest-driven consensus BUY but I have prior losing round-trips on MARA (-97 DKK), CLSK (-106 DKK), WULF (-63 DKK) — TICKER-FACIT says these need a substantially changed thesis, and 'short interest is still high' isn't new information, it was already true when I lost money on them. Staying out. For the holdings: IONQ is up +9.7% with an already-tight 5.1% stop under price — that's inside the ladder guidance (not yet at +15% threshold), so leaving it as-is rather than forcing a ratchet. PLTR similarly at +8.9%, stop already 5.8% under price, close to the avg dagsmove floor of 5.4% — tightening further risks a noise wick-out, so no change there either. Sizing RIOT modestly (~$220, under the $221 single-position cap) with a stop anchored below its recent higher-low around $19.40, giving room for the ±normal swing while protecting against a thesis break if the Anthropic deal narrative fades. Keeping remaining cash in reserve — no forced churn given the FEE-DRAG track record (14d: 8 fills costing 0.75% of turnover) reminding me low-edge scalps net to zero after fees. IONQ and PLTR are both modest winners (+9.7%, +8.9%) but neither has crossed the +15% ladder threshold yet, and their current stops already sit near or inside the avg-dagsmove gap-floor (IONQ 5.1% vs ±3.4% move, PLTR 5.8% vs ±5.4% move) — tightening further now risks a noise wick-out rather than real protection, so leaving both stops unchanged this cycle. PLTR is also CAP-FULL per STRATEGY PARAMS so no add is possible regardless of conviction. Sitting mostly in cash beyond the one RIOT entry: NBIS/CRWV gaps are earnings-day/untradeable chase risk, and MARA/CLSK/WULF are repeat-loser tickers per TICKER-FACIT with no genuinely new catalyst beyond the same short-interest thesis that already burned this account.

[stop-loss triggered at $7.51 — broker-fill $7.51]

Horisont: 1-2 weeks Begrundelse: Post #12 blowout Q2 with 150% commercial revenue growth and INTERPRETED net-sentiment +1.50 (single high-conviction fundamental post) — exit if the post-earnings pop fades back below 150. === Session-analyse === [saxomoon/US] Small live book ($619) after two stop-outs on FIG and WULF that both wick-out (both names since ripped back — FIG +17% and WULF +1% per yfinance delta vs my exit, painful confirmation of the GAP-FLOOR lesson). Cash is thin (~$328 spendable) so this cycle is about discipline, not chasing. PLTR is the real WSB story today (blowout Q2, net-sentiment +1.50, single high-conviction post, katalysator confirmed) but it was already rejected once this session for breaching the 35% cap — at ~$156.68 ask even 1 share ($157) plus my existing IONQ/WEN book would still leave room, so I'll take a small 1-share starter position sized well under cap rather than force the 2-share size that got rejected. IONQ is flat (+0.7%) and WEN up +8.1% with earnings tomorrow (8/7) — tightening WEN's stop slightly to lock in gains ahead of the print rather than adding risk into earnings vol. QQQ/SPY are dead flat (NEUTRAL regime), VIX calm at 16.5 but crypto Fear&Greed is 25 (Extreme Fear) — no reason to chase miners (RIOT/MARA/CLSK all report earnings today too, pure gamble into unknown prints). WSB's real energy today is Burry's bear call and 0DTE loss-porn (posts #4, #8, #9) — that's noise, not a tradeable long thesis. FIG and WULF stay off the table per re-entry rule (both need >=0.7 conviction + 2 independent cross-feed BUY votes; WULF actually has 4-BUY cross-feed consensus which is tempting, but I just got stopped out hours ago on a wick — will reassess next session once the dust settles rather than immediately re-chasing the same name same day). Staying mostly in cash given the thin spendable float and no fresh, uncrowded, cross-feed-confirmed setup beyond the small PLTR starter.

[stop-loss triggered at $25.41 — broker-fill $22.55]

[stop-loss triggered at $17.74 — broker-fill $17.36]

[stop-loss triggered at $149.40 — broker-fill $149.39]

[stop-loss triggered at $20.91 — broker-fill $20.89]

Horisont: 2-4 weeks Begrundelse: WSB post on APLD's FY27 earnings beat reads net-sentiment +1.50 (highest-conviction single name today) and cross-feed shows consensus BUY with analyst target +175% upside, while the stock gapped green overnight against a red crypto-mining tape. === Session-analyse === [saxovegas/US] RISK-OFF day (QQQ -0.53%) and a crypto-mining bloodbath overnight (RIOT -5.6%, CLSK -4.6%, IREN -4.4%, MARA -3.8%, WULF -4.0%) argues for caution, not chasing. MU is the loudest WSB thread today ('you have until 4 PM to unfuck this', 'MU went crazy') but it's UNTRADEABLE at this risk budget per the strategy params (gulv-stop too wide for 0 shares), and the interpreted signal shows 67% loss-porn with a 'bounced at $1000 support' thesis that's thin/binary — skip. NVDA is the other mention-leader ('1280% gain, too scared to sell') but it's SPIKE-CHRONIC (perennial WSB name, no +0.1 bump per rule #1) and max-position-capped to 1 share at this budget — not worth the fee drag for a single share on a red tape day. APLD stands out: single WSB post but net-sentiment +1.50 (highest conviction read today), cross-feed consensus BUY (analyst target +175% upside), appears in the news flow (Q3 FY27 earnings beat), and its own tape is GREEN overnight while the rest of crypto-adjacent infra red — a genuine single-name divergence per the THEME-FACIT guidance, not a basket trade. Gulv-stop ≈11.2% allows up to 14 shares within the $330 risk budget; this clears MIN-EDGE and sits in a different sub-theme (AI datacenter infra) than my current UBER/RDDT concentration, so no theme-cap conflict. Sizing modestly smaller than max given the risk-off tape and VIX still calm-ish (15.08) but not screaming green-light. Holding RDDT and UBER unchanged — both are CAP-FULL (no room to add) and both are flat/slightly green, well within their stop bands (RDDT +0.8%, stop 7.7% under; UBER +0.1%, stop 4.5% under) — no ratchet needed yet since neither has moved enough to justify tightening further per the trailing-stop ladder (need +15% PnL minimum). Staying mostly in cash (47%) given the risk-off regime and the lesson from past sessions that forcing trades on a red NASDAQ day has not paid — sitting on dry powder for a better entry window. RDDT and UBER are both CAP-FULL and barely moved (+0.8%/+0.1%) so no stop ratchet is warranted yet — their stops already sit appropriately per the gulv-stop bands. MU and NVDA are today's loudest WSB threads but MU is UNTRADEABLE at this risk budget and NVDA is SPIKE-CHRONIC and position-capped to a single share not worth the fee drag; both also sit against a risk-off NASDAQ tape (-0.53%) and an overnight crypto-mining rout that argues for caution on anything crypto-adjacent (RIOT/CLSK/IREN/MARA/WULF all down 3.8-5.6%). Staying ~47% cash given the regime tilt (rule #11) and waiting for either a cleaner dip-buy signal or a genuine risk-on session before deploying further.

[stop-loss triggered at $112.74 — broker-fill $112.72]

Horisont: 1-2 weeks Begrundelse: WSB DD post '$UBER, follow the smart money' (score 469) is today's freshest non-chronic ticker in the hot listing, with a calm daily range letting a tight 5% stop fit the risk budget. === Session-analyse === [saxovegas/US] RISK-ON tape (QQQ +0.76%, VIX calm at 16) and WSB's top fresh DD today is the $UBER 'follow the smart money' post (score 469, #25 hot) — genuinely NEW attention vs the chronic MU/TSLA earnings chatter. UBER's daily range is tight (~1% swings), gulv-stop lands at 4.1% which fits comfortably inside the $37 risk budget for 5 shares (~$346, well under the $374 cap and $577 spendable). No cross-feed line explicitly covers UBER but the WSB DD is concrete and the setup is clean technically (flat/rising into support ~$69). MU earnings TODAY is loud on WSB (JPMorgan beat-and-raise thesis, $50K YOLO post, INTERPRETED signal +0.94 net-sentiment) but MU is flagged UNTRADEABLE at min size for this account — $1075/share means even 1 share risk-sized properly doesn't clear the budget cleanly, and opening a fresh earnings-day position I can't size sanely is exactly the low-edge trap rule #17 warns against. Passing. MARA/RIOT crypto-mining cross-feed shows SELL consensus and the THEME-FACIT shows this book has lost -723 DKK across 10 mining round-trips — no fresh external catalyst here, staying out per the theme re-entry bar. Holding INTC and RDDT untouched — both CAP-FULL (no room to add under the 30% position cap) and both gapped UP overnight (+2.8%, +1.4%) on no bad news, consistent with EXIT OVERRIDE keeping bagholding-cut reflexes off. Neither has moved enough from entry to justify ratcheting stops tighter than current resting levels (INTC -2.1%, RDDT -3.6%, both still below breakeven) — leaving stops as-is rather than forcing a premature tighten inside their daily noise bands (INTC ±3.4%, RDDT ±2.4%). INTC and RDDT are CAP-FULL — no adds permitted regardless of conviction — and both are still underwater from entry despite today's overnight bounce, so no stop-tightening is warranted yet (would risk firing inside normal daily noise). MU is the loudest WSB name today (earnings-day YOLO thread, JPMorgan beat-and-raise chatter) but is explicitly UNTRADEABLE at this account's min size — skipping rather than forcing a sub-scale earnings-day bet. Crypto mining (MARA/RIOT/IREN) shows cross-feed SELL consensus and a -723 DKK theme track record with no new external catalyst, so staying out per the theme re-entry bar.

[stop-loss triggered at $185.27 — broker-fill $185.20]

Horisont: 1-2 weeks Begrundelse: WSB YOLO thread ($146K RDDT Jan28 calls, disclosed position) plus cross-feed consensus BUY (analyst +40% upside) and social-theme tape +7.1%/5d; exit if WSB interest fades or thesis-break below the stop. === Session-analyse === [saxovegas/US] COIN and INTC are both CAP-FULL per STRATEGY PARAMS — no adds allowed regardless of conviction. COIN is up +17.6% with an avg dagsmove ±4.4%; current stop sits at 4.9% under price which is inside the noise band, but the ladder says +15% PnL → trail ~12%, and gap-floor says never trail TIGHTER than the daily move — so I'm leaving it roughly where the ladder/floor intersect rather than tightening further into noise. INTC at +2.2% is barely off entry with stop already 9.3% under (above its 3.7% dagsmove floor) — no change needed, thesis intact (semis theme is +6.5%/5d, INTC own overnight gap -2.4% is normal noise not a break). On new money: WSB hot list today is thin on fresh tradeable names — NKE (score 15,162) is a meme/loss-porn thread about a mercy-kill short, not a bullish catalyst (title: 'The only way to rescue Nike…'), and its own gulv-stop math only allows 10 shares max ~$359, but the thesis behind the hype is bearish irony, not a buy signal — skip. ORCL has real news (force majeure notice on datacenter, -ish) but cross-feed isn't shown consensus BUY and it's a headline-driven news risk, not WSB conviction — pass. RDDT has a genuine dot: WSB YOLO thread ($146K Jan28 calls), cross-feed consensus BUY (analyst target +40% upside), tema-tape social +7.1%/5d — but gulv-stop caps it at 2 shares (~$305), and per rule #9/theme discipline that's a small, single-shot bet in a name with no existing exposure; META also cross-feed BUY with WSB net+0.92 disclosed position but is UNTRADEABLE at min size per this session's rules. Given RISK-ON regime (QQQ +0.49%), calm VIX (14.2), and a real trending name with cross-feed corroboration, I'll take a modest RDDT position sized to the gulv-stop max (2 shares) — this respects rule #17 min-edge only marginally, but the $305 notional clears the $200 floor and the setup (WSB momentum + analyst upside + social tema tailwind) is the best fresh idea in an otherwise quiet tape. No sub-theme conflict since 'social' is currently at 0%. Cash stays largely on the sidelines (41% of book) — no other name clears both a fresh WSB angle AND cross-feed corroboration AND passes the risk-sizing math; TTWO's single-post YOLO is thin sample per rule #15 (interpreted signal is high net-sentiment but n=1, no real corroboration) and tema-tape shows gaming -13.7%/30d — pass. AMZN mention is accelerating but interpreted signal is net-negative/ironic (-0.12, meme thesis) — no real conviction there. Holding the two live winners with stops resting above their volatility floor is the correct discipline this cycle; last cycle's IREN stop fire (+5.4%, above its own gulv-band) was a legitimate protective exit, not a forced cut, so nothing to second-guess there per rule #7. COIN and INTC are CAP-FULL, so no adds there regardless of conviction; both stops already sit above their respective daily-noise gulv-floors and I'm not tightening further into noise per rule #13's gap-floor. Cash (41% of book) stays mostly parked: NKE's massive hot-list score is loss-porn/mercy-kill irony not a bull catalyst, ORCL carries fresh negative news risk without cross-feed corroboration, META/other big names are UNTRADEABLE at this session's min size, and TTWO/AMZN reads are thin-sample or net-negative per the interpreted signal. RDDT was the one name clearing WSB momentum + cross-feed BUY + tema tailwind at a size that clears the $200 floor.

[stop-loss triggered at $44.95 — broker-fill $44.94]

Horisont: 1-2 weeks Begrundelse: WSB has fresh accelerating INTC gain posts ('14k->38.5k weekend hold', 'Mr. Bottom Catcher') plus a +12% overnight gap on a risk-on tape — sizing to the 8.9% gulv-stop floor (3 shares ≈ $365) rather than the ticker's full momentum given semis theme has been a net loser for this book. === Session-analyse === [saxovegas/US] Risk-on tape today (QQQ +2.54%, SPY +1.49%, VIX a calm 15.4) and WSB is squarely in a semis/AI celebration mood — AMD hitting $1T market cap is the #4 hot post with strong loss-porn/momentum sentiment, MU spreads gain post is top-10, and INTC has multiple 'weekend hold' gain posts. Cross-feed corroborates several of these (MU consensus BUY on analyst upside, IONQ/CLSK/IREN/ARM all BUY-consensus), but per the min-shares/floor math AMD, MU, INTC-at-min-entry are mostly UNTRADEABLE or barely tradeable at this tiny $656 spendable budget — INTC is the one name that actually clears the risk-sizing floor (max 3 shares, gulv-stop ~8.9%), and it has genuine WSB fresh momentum (accelerating mentions, two 'weekend hold gain' posts, +12% overnight gap) plus room in the semiconductors theme (0% current exposure) that's been a net loser for us historically (-143 DKK) but INTC specifically hasn't been the loser (that was NVDA/older INTC positions) — still, I'll size it modest and respect the theme's soft memory. COIN and IREN are both healthy winners (+22.9% / +11.9%) — COIN is CAP-FULL so no add is possible anyway, and IREN's cross-feed consensus BUY plus ai_infra/crypto_mining tape strength argues for holding and trailing rather than trimming. AMD, MU, QQQ, META, SPY are all effectively untradeable at this account's tiny risk budget (gulv-stops too wide or min-entry impossible under the $200 floor) despite being the loudest WSB names today — chasing them with 1 share each would be exactly the binary 1-2 share mistake rule #3 warns against. Cash stays at ~51% of book; with only $656 spendable and most hot names priced/vol'd out of range, one modest INTC add is the only trade that clears both the WSB-momentum bar and the risk-sizing floor this cycle.

Horisont: 3-7 days Begrundelse: WSB actively debating COIN ahead of the CLARITY Act senate vote (posts #19/#20, plus #4 news at 7,312 upvotes) with cross-feed consensus BUY driven by 19.6% short interest — exit if the regulatory narrative fades or COIN breaks below recent support. === Session-analyse === [saxovegas/US] RISK-ON tape (QQQ +0.88%), VIX calm at 17.1, and IREN is up +2% on entry with a strong overnight gap (+4.9%) confirming crypto_mining momentum — thesis intact, no reason to touch the position beyond trailing the stop tighter to lock in the gain per rule #13, though the move is modest (+2% PnL) so I'm nudging the stop only slightly rather than aggressively, since the gulv-floor (12.6% avg dagsmove-based) still dominates the ladder math at this PnL level. On new entries: COIN shows cross-feed consensus BUY (short_interest +0.69, 19.6% float short) and is explicitly trending in WSB (CLARITY Act senate vote fail post at #4 with 7,312 upvotes, plus two direct COIN posts at #19/#20 debating the Clarity Act angle) — this is a real WSB catalyst (regulatory vote), not stale chatter. But the gulv-stop math caps COIN at 2 shares (~$332) given its 10.4% floor, and my risk budget only supports that size — still, 2 shares clears the $200 min-trade floor. ORCL also crossed my screen (WSB post #5 on layoffs, INTERPRETED signal net -1.00 bearish, CFO defensive quote) — that's a bearish catalyst on the actual post, not a buy signal, so I'm passing despite ORCL being 'tradeable' and mentioned; rule #18 requires me to honor a named headwind rather than override it. NVDA is under a mandatory re-entry cooldown (exited -4.3% via stop fire 9/14) requiring conviction ≥0.7 AND 2+ cross-feed BUY votes — current cross-feed shows only 1 BUY vs 3 SELL, so it's server-blocked regardless of my conviction; skipping. For COIN: short interest at 19.6% of float with WSB actively debating the CLARITY Act vote failure as a bullish setup ('Ready for COIN to rip ahead of senate vote') is a genuine squeeze-adjacent setup, cross-feed consensus BUY, and it fits within crypto/fintech theme without overloading crypto_mining (COIN is not literally a miner, diversifies away from IREN's pure-mining exposure). Sizing to the gulv-stop cap of 2 shares (~$332) keeps risk contained. RKLB WSB post is 100% loss-porn per INTERPRETED signal (author admits losses, no fresh catalyst) — that's a fade signal per rule #15, skipping despite the trending mention. IREN stays as-is beyond the modest stop tightening — thesis intact (crypto_mining theme tape +0.2% 30d, overnight gap +4.9%, RISK-ON regime), not yet at the +15% ladder trigger so no aggressive trail. Passing on ORCL despite the WSB mention because the actual post content (CFO comment on layoffs) is a named bearish catalyst, not support — buying into a headwind I've identified myself would violate rule #18. NVDA is cooldown-blocked (exited via stop 9/14, cross-feed only 1 BUY vs 3 SELL, fails the 2-vote server requirement). RKLB's WSB post is pure loss-porn with no fresh catalyst — fading per rule #15. Sitting mostly in cash (now ~$650 remaining spendable) is appropriate mid-cycle; will reassess at 14:00 for fresher trending names.

Horisont: 1-2 weeks Begrundelse: JPM upgraded IREN to overweight (interpreted signal +1.50, 1 high-conviction post, katalysator 1/1) and cross-feed shows consensus BUY (analyst target +78% upside) — a genuinely fresh catalyst distinct from my prior loss on this name, sized small given RISK-OFF tape and my poor crypto_mining track record. === Session-analyse === [saxovegas/US] QQQ is down -1.66% and SPY -0.74% today — a clear RISK-OFF regime, and the overnight tape confirms it: SOXL -16%, MRVL -8%, ARM -7%, MU -6.4%, INTC -6.4%, ASML -6%, CRWV -6% — semis and AI-infra got hammered overnight on the SK/Samsung uncertainty headline and 'AI slowdown' discussion sitting on WSB's hot list right now. WSB itself is mostly macro fear today (rate hike chatter, Warsh/Volcker hawkish talk, oil shock/Iran headlines, 'this selloff makes zero sense') — there is no fresh, high-conviction single-name YOLO trending in the actual tradeable universe; NVDA is the only universe ticker mentioned and its own interpreted-signal read is net-bearish (-0.25, 3 bear vs 1 bull, thesis fundamental) despite a cross-feed-adjacent posture. My NVDA position was just stopped out this morning (-4.3%), and rule #8 gates re-entry hard: it needs conviction >=0.7 AND >=2 fresh cross-feed BUY votes citing something NEW — I have neither, so NVDA stays off the table (cooldown respected). Looking at cross-feed consensus BUY names that ARE in the trending/interpreted mix: IREN shows a genuine catalyst (JPM upgrade to overweight, interpreted net +1.50, 1 high-conviction post) plus cross-feed consensus BUY (analyst_ratings +0.82, target +78% upside) — but my own TICKER-FACIT shows I've lost on IREN before (-137 DKK) and it's down again overnight (its theme, crypto_mining, is my worst theme at -757 DKK over 90d with 11 exits). The JPM upgrade is a genuinely NEW catalyst though (rule #8's 'what changed' bar), and it's a small, single-name divergence from a falling semis tape, not a chase into strength. Given RISK-OFF regime + my own poor crypto_mining track record, I'm sizing this very small and using a wide stop to respect the volatility floor, rather than skipping entirely — this is the single name that clears both a fresh catalyst AND cross-feed corroboration bar today. Everything else — GME, MARA, CLSK, RIOT, WULF, MRVL, ARM, CRWV, ASTS, NBIS — either has zero fresh WSB narrative behind it today (just perennial cross-feed BUY tags) or is actively getting crushed by overnight semis/AI-slowdown fear, which is exactly the kind of chase-into-a-red-tape rule #11/#16 warns against. With cash at ~$1,191 spendable and portfolio down slightly over the last week, the correct move on a day the NASDAQ is down -1.66% with no standout fresh WSB catalyst except IREN's JPM upgrade is to stay mostly in cash and take one small, catalyst-backed position rather than force multiple entries into red-tape names. Sitting mostly in cash today. QQQ -1.66%/SPY -0.74% is a clear risk-off signal and the overnight move confirms it broadly across semis/AI-infra (SOXL -16%, MRVL -8%, ARM -7%, MU -6.4%, INTC -6.4%, CRWV -6%) — exactly the tape rule #11 says to avoid chasing. WSB's hot list today is dominated by macro fear (rate hikes, oil shock, 'selloff makes zero sense') rather than a fresh single-stock YOLO; the only universe ticker mentioned by name, NVDA, just stopped me out this morning and its own interpreted read is net-bearish, so re-entry cooldown (rule #8) and lack of conviction keep it off the table. Cross-feed consensus-BUY names like GME, MARA, CLSK, RIOT, WULF, NBIS, CRWV have no fresh WSB catalyst behind them today — buying them now would be chasing perennial tags into a falling tape, not reacting to new evidence. I'll re-assess at 11:30 once WSB fully wakes up and the overnight semis rout either stabilizes or deepens.

[stop-loss triggered at $210.72 — broker-fill $210.48]

Horisont: 1-2 weeks Begrundelse: WSB has 4 posts on NVDA today (Jensen Huang memes, Hugging Face acquisition news) with net-sentiment +0.62 and a high-conviction post citing a concrete catalyst, corroborated independently by analyst_ratings BUY (+46% upside target) — exit if it breaks below recent support near $208 or the AI-bubble skepticism thread (post #7) gains real traction. === Session-analyse === [saxovegas/US] Today's WSB feed is thin gruel — the hot posts are memes (Jensen Huang, oil deals, sailors in Thailand), not fresh tickers. Only NVDA clears actual mention volume (4 posts, net-sentiment +0.62, 1 high-conviction, katalysator 3/4) and it's backed by an independent cross-feed BUY consensus (analyst_ratings +0.89, target +46% upside) — that's genuine multi-source corroboration, not just Reddit wearing three hats. But QQQ is -1.03% today, a clear risk-off tape, and VIX is calm at 14.4 so it's not a panic day, just a soft one. NVDA itself isn't in the TRENDING/MENTION-LEADERS blocks (those are RDDT/TTWO) so it doesn't get the auto +1 bump, and RDDT's interpreted read is actually net-bearish (-1.50, high-conviction, thesis 'growth story over') despite trending — that's a clear caution flag per rule #15, not a buy. Given RISK-OFF regime (#11) raises the bar, and my own THEME-FACIT shows semiconductors have been a net loser for me (-410 DKK, 5 exits) and crypto_mining a bigger net loser (-766 DKK, 14 exits), I'm not chasing miners today (MARA is cross-feed SELL 2-4, SOXL cross-feed SELL, OKLO cross-feed SELL despite short-interest flag). CLSK/RIOT/MARA are all overnight red and cross-feed net-negative — sitting those out. NVDA is the one name with real independent corroboration (analyst_ratings BUY, INTERPRETED bull read, historically I've had one small win on NVDA per TICKER-FACIT +81 DKK) but semiconductors tape is -6.6% over 30d and the entry here is a small, disciplined nibble given the red NASDAQ day — sizing down rather than skipping entirely, since the cross-feed and interpreted signal both agree. Everything else — SOUN (cross-feed BUY but overnight gapped -5.1%, and my own TICKER-FACIT shows I bagged a +306 DKK win last time so I'm wary of re-chasing after a big short-interest-driven pop that already faded), ASML/AVGO (cross-feed BUY but not WSB-hyped, and AVGO reports earnings tomorrow — avoid fresh entries into a print), QBTS (huge yfinance delta anomaly +17% looks like a data glitch, skip) — none clear the bar for a fresh position on a risk-off day. No open positions to manage stops on. Staying mostly in cash with one small, well-corroborated NVDA nibble is the correct read of a quiet, meme-only WSB session on a red tape. WSB today is meme-heavy and ticker-light: the actual trending block only surfaces NVDA and SPY from post titles, mention-leaders are RDDT/TTWO with RDDT's interpreted read flipping bearish (-1.50 net-sentiment, high-conviction, capped-DAU thesis) despite the mention count — a textbook rule #15 caution, not a buy. QQQ is red (-1.03%) which raises my bar per rule #11, and my own THEME-FACIT shows crypto_mining and semiconductors have both been net losers over 90 days, so I'm not chasing MARA/CLSK/RIOT/OKLO even though short-interest flags exist — cross-feed screening actively shows SELL consensus on MARA, CLSK, SOXL, OKLO, HOOD, IBIT today, confirming caution is warranted. AVGO reports earnings tomorrow so no fresh entry there despite the cross-feed BUY. Sitting mostly in cash (only holdings: none currently) with a single small, corroborated NVDA nibble reflects the quiet-WSB, risk-off backdrop — I'll get more aggressive if QQQ turns green and WSB starts naming a fresh high-short squeeze candidate with real independent cross-feed support.

[stop-loss triggered at $125.05 — broker-fill $125.05]

[stop-loss triggered at $19.96 — broker-fill $19.99]

Horisont: 1-2 weeks Begrundelse: Overnight +7% gap on BTC strength, INTERPRETED signal shows a high-conviction call buyer post (net-sentiment +0.75), cross-feed consensus BUY with analyst target +92% upside, and crypto_treasury tema-tape (+15.8% 5d) is diverging positively from my losing crypto_mining basket — a genuinely different bet, not the same theme in new clothes; exit if BTC rolls over or MSTR loses the $115 level. === Session-analyse === [saxovegas/US] RIOT is the only holding and it's up modestly (+4.4%) with the overnight tape actually favorable (RIOT gapped +4.6% overnight, crypto Fear&Greed at 73 Greed, BTC 30d +20.5%). Cross-feed shows RIOT consensus BUY (3 BUY/1 SELL/2 HOLD) with analyst target +63% upside — decent corroboration. But QQQ is red -1.18% today (RISK-OFF regime), crypto_mining theme is already AT CAP (100% of book), and my own THEME-FACIT shows -767 DKK across 13 exits on this exact theme — I have a proven negative edge here. No new capital should chase further crypto mining exposure today. MSTR (crypto_treasury) is tempting — INTERPRETED signal shows net-sentiment +0.75 with a high-conviction call-buyer post, cross-feed consensus BUY, and tema-tape shows crypto_treasury +15.8%/5d and +31.7%/30d, diverging positively from the crypto_mining basket's -9.4%/30d. That's a genuine name-specific catalyst (MSTR gapped +7% overnight on BTC strength) rather than borrowed mining momentum. However it's a RISK-OFF day on NASDAQ, VIX is calm (16) but Greed is elevated (73) — blow-off-top caution per rule #12. MRNA is the loudest WSB story today (cancer vaccine breakthrough, $500k YOLO posts, 38 comments in flow) but it's NOT in the tradeable universe... wait, checking again — MRNA IS in the tradeable universe. But it already ran +50% and is now pulling back overnight (-5.3%), classic blow-off/fade risk after a parabolic spike on binary trial news — chasing a name already up 50%+ on a single-day biotech pop is exactly the low-edge chase rule #16 warns against, and WSB sentiment on it is 'mixed' per the comment-flow gist, not clean bullish. Holding RIOT unchanged — it's a modest winner (+4.4%) but crypto_mining is already AT CAP (100% of book) and my own THEME-FACIT shows -767 DKK realized across 13 round-trips in this exact theme, so I'm not adding more mining exposure regardless of today's favorable overnight gap. Its stop stays where it is since the move hasn't extended enough to justify tightening beyond the current 18.7% cushion, and the position is small relative to portfolio total. RISK-OFF regime today (QQQ -1.18%) raises my bar for fresh adds across the board; the one new trade (MSTR) is sized small (2 shares, ~$251) specifically because Fear&Greed is at 73 (Greed) and blow-off-top risk on crypto-adjacent names is real. MRNA — today's loudest WSB story — is being passed on: it already spiked 50%+ on binary trial news and is fading overnight (-5.3%), which is a textbook chase-risk/low-edge entry per rule #17, and WSB's own commentary on it reads as 'mixed', not clean conviction. Sitting mostly in cash/RIOT rather than forcing more trades on a red NASDAQ day.

Horisont: 1-2 weeks Begrundelse: Cross-feed consensus BUY (4-1-1) with technicals leading (5d +15%, range_pos 64%) and short-interest fuel behind crypto miners broadly — exit if BTC rolls over or RIOT loses the $18.50 shelf. === Session-analyse === [saxovegas/US] WSB's front page today is dominated by MRNA (not in universe) and generic mag7/macro noise — the actual trending tradeable names (RDDT, NVDA, MU) are fading (▼ vs yesterday), not fresh. HOOD stands out on the OVERNIGHT GAP block (+13.8% overnight) and mention-leaders (+3 today), but the interpreted read is thin (1 post, +0.50, no high-conviction) and it's not in cross-feed screening at all — that's a chase-risk gap, not a corroborated setup, so I'm passing on it rather than buying the top of a 14% gap. The real signal today is cross-feed consensus on crypto-mining names: RIOT, MARA, CLSK, IREN, WULF all show BUY consensus with short_interest as the top driver (CLSK 40.3% short float, IREN 28.4%, WULF 123% upside target) plus strong technicals (MARA 5d +32%, RIOT 5d +15%). However my own THEME-FACIT shows crypto_mining has been a -767 DKK loser across 13 exits on these exact 5 names, and I just got stopped out of MARA (+266 DKK, a win this time) and IREN (-137 DKK) in the last 72h. IREN is also in my RECENT LOSING EXITS list requiring conviction ≥0.7 AND 2+ independent cross-feed votes — it has 2 independent votes (short_interest + others) which clears the bar, but earnings in 6 days is a real risk I don't want to underwrite with a fresh position right before the print. RIOT and CLSK don't carry that earnings overhang and have clean independent cross-feed corroboration (short_interest + technicals), so I'll take one measured position in RIOT (better technical setup, 5d+15%, range_pos 64% vs CLSK's crowded 40% short but no earnings catalyst either way) sized small given the theme's poor track record — this is a single bet, not adding legs across the whole mining complex. QQQ is green (+0.34%) and VIX is calm (14.89), so risk-on tilt supports a modest add, but I'm keeping size well under max position given the theme-level scar tissue and avoiding IREN/WULF/MARA to not re-stack the same correlated bet four times. Sitting mostly in cash outside the one RIOT entry. HOOD's overnight gap is tempting but uncorroborated by cross-feed screening and the interpreted read is a single thin post — chasing a +14% gap on drive-by mention-leader data alone fails the min-edge/corroboration bar. MRNA is the actual WSB obsession today but it's not in the tradeable universe. RDDT/NVDA/MU are all fading mentions (▼ vs yesterday) with negative or bearish interpreted sentiment (MU -0.50, INTC -0.50) — no fresh catalyst there, and NVDA has earnings in 5 days which argues against a fresh pre-print position anyway. Crypto-mining as a theme has cost me -767 DKK over 13 round-trips, so I'm treating today's cross-feed BUY consensus on RIOT/CLSK/IREN/WULF/MARA as one correlated bet and only taking the single cleanest name (RIOT, no earnings overhang, best range position) rather than stacking legs across the whole complex. Will reassess IREN specifically after its earnings print on 8/27.

[stop-loss triggered at $11.04 — broker-fill $11.03]

Horisont: 1-3 weeks Begrundelse: Cross-feed consensus BUY driven by short_interest (31.5% of float, 2.2x ratio) — a genuine squeeze setup independent of Reddit hype, confirmed by the +6.8% overnight gap; exit if BTC rolls over or the gap fully fills below $9.20. === Session-analyse === [saxovegas/US] Thin gruel on WSB today — RDDT is the only tradeable name with genuine mention momentum, and the interpreted read flags it as thin (2 posts, 50% loss-porn, 0 high-conviction) despite a decent net-sentiment number. Worse, RDDT is a RECENT LOSING EXIT (stopped out -5.9% on 08-17), which per rule #8 requires conviction >=0.7 AND 2 independent cross-feed BUY votes — the cross-feed line isn't even shown as consensus BUY here, and TEMA-TAPE shows the social basket down -10.1%/5d and -14.9%/30d. That's a stacked no: forgotten thesis, tape headwind, and the re-entry bar unmet. Skip it. Where the real signal is: MARA and CLSK both show consensus BUY with short_interest as the top independent feed (MARA 31.5% short/2.2x ratio, CLSK 40.3%/3.2x) — genuine squeeze fuel, not just Reddit noise, and the overnight gap shows MARA already +6.8% confirming momentum. But crypto_mining is my worst-performing theme by a mile (THEME-FACIT: 12 exits, -1033 DKK across 5 names) and both were losing round-trips (RIOT -179, CLSK -164 in the ticker facit, though not MARA itself). Per rule #8/theme discipline, re-entering this theme needs a theme-EXTERNAL catalyst, not just cluster momentum — the short-interest fundamentals ARE somewhat name-specific and independent (not just chasing the crypto pump), and MARA itself isn't in my losing-exit list, so I'll take a single small, disciplined position in MARA only (not CLSK, avoiding doubling the theme bet) sized well under cap, with a stop respecting the volatility (10d closes swinging ~4-6%, so 2x that framework points toward a wider stop). Everything else: QQQ regime is neutral, VIX calm at 15, Fear&Greed at 46 (neutral-fear) — no strong tilt either way. NVDA has earnings in 7 days and the interpreted read is actually negative (-1.0 net-sentiment, AI slowdown thesis) — avoid. MU also has earnings soon and consensus is thin (1 BUY vote). No other name clears both the WSB mention bar and independent cross-feed corroboration cleanly enough to justify spending the very limited ~$1,178 spendable cash on more than one position — better to stay concentrated and disciplined than spray small sub-floor bets across many names. WSB is quiet on tradeable names today — only RDDT shows real mention acceleration, but it's a recent losing exit with thin interpreted conviction (50% loss-porn, 0 high-conviction posts) and the social theme tape is down double-digits over 5d/30d, so the re-entry bar isn't cleared. Sitting on the rest of the cash rather than forcing entries into NVDA/MU ahead of earnings, or adding a second crypto-mining name (CLSK) which would just double down on my worst-performing theme (-1033 DKK over 12 exits) without an external catalyst. Will revisit RDDT if a fresh independent cross-feed BUY consensus appears and mention momentum sustains past a single day.

[stop-loss triggered at $41.73 — broker-fill $41.73]

[stop-loss triggered at $221.68 — broker-fill $220.06]

[stop-loss triggered at $165.82 — broker-fill $165.81]

[stop-loss triggered at $101.72 — broker-fill $101.70]

Horisont: 1-2 weeks (through the Aug 18 S&P inclusion) Begrundelse: Post #9/#25 — Reddit joins S&P 500 on Aug 18 (concrete dated catalyst, forced index buying), interpreted read net +1.50 with 3 high-conviction posts and cross-feed 4-0 BUY; exit if it loses the gap into inclusion day or the stop fires. === Session-analyse === [saxovegas/US] RDDT is the one genuinely fresh, multi-source setup today: the S&P 500 inclusion news (post #9 at 2,736 upvotes and #25) is a concrete, dated catalyst on Aug 18 — index funds are forced buyers — and the interpreted read backs it hard (net +1.50 over 3 posts, 3 high-conviction, thesis 'fundamental', 0% loss-porn, 2/3 posts citing a concrete catalyst). Cross-feed shows consensus BUY 4-0. The problem is price: RDDT already gapped +10.6% overnight into the open, which is exactly the chase risk rule #16 warns about, and today's tape is RISK-OFF (QQQ -0.42%). Given spendable cash of only ~$440 USD, a 2-share RDDT entry (~$351) is my one meaningful shot. I weigh the chase risk explicitly by NOT full-sizing (2 shares, not 2.5x) and by giving it a wider stop (9%) rather than a tight one inside the post-gap noise band — a 3-4% stop on a name that just moved 10% overnight is a coin-flip. Index-inclusion flows into Aug 18 are the thesis; if it fails to hold the gap through the inclusion date, the stop does the work. On existing book: NVDA is my only winner (+11.5%) and the $500B financing story (post #10, interpreted net +1.50, high-conviction, concrete catalyst) is fresh bullish news — I do NOT trim per rule #2. Its stop is already 2.9% under price and avg dagsmove is only ±1.0%, so I can ratchet modestly to ~2.5% without violating GAP-FLOOR. CRWV and IREN are both roughly flat and both still carry cross-feed consensus BUY (CRWV 3-1 with 20.4% short float; IREN 3-0 with 28.4% short float) — the EXIT OVERRIDE on this tier plus a net-bullish cross-feed means I hold both and let the mandatory stops cap risk. I will not tighten CRWV (stop already 3.5% under price vs ±4.2% daily move — already inside the noise band, tightening further is a market order in disguise) and leave IREN's 6.1% trail alone since it's at breakeven, nothing to protect yet. I'm skipping everything else. RIOT has the flashiest headline (Anthropic $9B cloud deal, post #21) but I was stopped out of it two days ago at -8.8% and the server requires 2+ independent cross-feed BUY votes for re-entry — RIOT does not appear in the cross-feed block at all, so it's blocked and correctly so. MSFT/MU/AMD/META are untradeable at my size. NBIS is mention-leading but the interpreted read is net -1.50 (the one post is someone SELLING it) and it gapped +7.6% — no. PLTR's +34% run is a Spider-Man meme with 0 high-conviction posts; that's noise. Sub-theme ai_infra is already AT CAP at 50%, which is another reason RDDT (social/index) is the right place for fresh cash rather than another AI-infra name. Holding CRWV and IREN — both near flat, both still carry net-bullish cross-feed consensus (CRWV 3-1, 20.4% short float; IREN 3-0, 28.4% short float), and this tier's EXIT OVERRIDE bars reflexive forgotten-cuts; the mandatory broker stops (101.72 / 41.73) cap the downside. NVDA held and not trimmed despite +11.5% — the $500B financing headline is fresh bullish news, so I ratchet the stop to 2.5% under price (avg dagsmove only ±1.0%, so this respects GAP-FLOOR) and let it run per rule #2. No RIOT re-entry despite the Anthropic headline: it's a losing exit from two days ago with zero independent cross-feed BUY votes today, so it fails the server's re-entry bar. No adds to ai_infra — already at the 50% sub-theme cap. Risk-off tape (QQQ -0.42%) plus Crypto Fear at 29 keeps me to a single, deliberately under-sized entry rather than spraying cash across a red day.

[stop-loss triggered at $19.16 — broker-fill $19.16]

[stop-loss triggered at $29.97 — broker-fill $29.95]

Horisont: 2-4 years Begrundelse: Adding to an existing SMR/nuclear position on strong 5d momentum and 3-1 BUY consensus, while position size (9% of book) leaves ample room below the 35% cap and the structural datacenter-baseload thesis is unbroken. === Session-analyse === [saxostrategi/ALL] Cash is scarce (7% of book) and the DK buy window is structurally closed this cycle — no DK holdings to sell, so no path to the 1,160 DKK floor there; skipping Danish names entirely rather than forcing a bad trade. On the US side, TSM/LRCX/AMAT/MU/GEV/PWR are cap-blocked (correctly — I hold winners, not touching them), leaving CDLR, NEE, and OKLO as the only legal sizes. CDLR's tape is weak (marine_installation +1.4%/-3.6% 30d) with no catalyst; NEE at 1 share is fee-inefficient and already fairly priced (analyst upside modest). OKLO stands out: cross-feed consensus is BUY (3-1), 5d momentum +11% on genuinely improving small_modular tape (+7% 5d), and it's still only 9% of the book — room to add without approaching any cap. This is 'adding to a name already working,' not a fresh chase, and the SMR/nuclear structural thesis (dispatchable baseload for datacenters) is intact and if anything strengthening on the uranium/nuclear tape (+6%/+3.9% 5d despite negative 30d — a base building, not a blow-off like ARM/AMD at 96-99% range position). TSM (34%) is at the position cap and should not be added to; LRCX, VST, CCJ theses are all intact with no new information to act on — holding through normal sector noise per mandate. DK buy window is closed this cycle (no DK holdings to sell to reopen it, and forcing a sub-floor trade would just burn fees). Cash remains thin (7%) by design after the OKLO add; waiting for the book to compound before deploying more into cap-blocked names like TSM/LRCX/AMAT/MU rather than chasing extended momentum names (AMD, ARM) at 96-99% of their 10-day range.

Horisont: 2-3 years Begrundelse: Uranium is the physical constraint on the nuclear baseload thesis inside ELECTRICITY, and CCJ is the least-diluted, most liquid pure-play producer to add alongside OKLO/VST without touching capped semis names. === Session-analyse === [saxostrategi/ALL] The book is small (8,588 DKK) and already concentrated: TSM at 33% and LRCX at 22% are both effectively capped from further buying, and semis/ai_infra exposure sits at 55%. Cash is only ~1,215 DKK, which after fee-guard math gives a single US buy window of roughly $40-179. Memory and semicap tape is hot (+7.5%/+6.6% 5d) and cross-feed consensus on MU, LRCX, AMAT is uniformly BUY with strong analyst upside, but MU's per-share price (~$1000+) and LRCX's cap-block rule those out for new buys today. CCJ and LEU are legal-sized single-share adds with strong tape momentum (uranium/nuclear feed strength, LEU's short-interest signal at 0.83) but LEU has already run +23% over 10 sessions off a low base — chasing that vertical move into a 1-share position is a bad risk/reward for a name I don't yet hold, better to wait for a pullback. CCJ is a cleaner story: uranium/nuclear structural thesis intact, price up modestly (92→102, +10% over the window) without the parabolic look, and it's not yet a position, giving genuine diversification within the nuclear sub-theme without touching my capped semis exposure. VST carries an edgar SELL flag (dilution filing + insider sales) and is my only red position (-2.3%), but the thesis — VST as nuclear/grid/data-center-adjacent generation capacity — is not broken by a routine shelf filing; a couple of insider sales is weak-signal noise per the feed's own caveat. I'm not adding to it (single-share cap anyway) and not selling on a filing that doesn't invalidate the structural case; a bad week is not a broken thesis. Given the fee-floor math, one clean US buy (CCJ, 1 share ≈ $102, well within the $40-179 window) is the sensible single action today — sized small because that's what the cash allows, not because conviction is low. No DK trades: no Danish tickers in the tradeable universe roster of names I want exposure to right now, and forcing a DK leg just to use both windows would be churn, not thesis. TSM and LRCX are both past or near the 35% concentration ceiling and cap-blocked for new buys regardless of how attractive analyst targets look (TSM +27%, LRCX +30% upside per feeds) — that's a function of book size, not a signal to trim; winners compound. VST's edgar dilution flag is real but a routine 424B5/S-3ASR shelf plus a couple of insider sales does not break a grid/nuclear generation thesis, and the position is too small to trim economically anyway — holding through the noise. Not chasing LEU's vertical +23% run into a first position; will revisit on a pullback. Staying in one clean CCJ add today given the tight $40-179 US buy window and reserving remaining cash rather than forcing a second, worse-sized leg.

Horisont: 2-3 years Begrundelse: Small modular reactor thesis is a distinct nuclear sub-theme from VST's regulated-utility exposure — datacenter operators need dispatchable baseload before 2030 and OKLO is the clearest pure-play SMR name in the universe, exit if NRC licensing timeline slips materially or dilution accelerates beyond routine. === Session-analyse === [saxostrategi/ALL] Cash is tight (~1,970 DKK), and the three big cap-blocked winners (TSM, LRCX, VST) already carry the book's real exposure — TSM at 33% is near the ceiling, VST at 22%, LRCX at 21%. No thesis has broken: semicap/memory/nuclear are all down 5-13% over 30d on sector noise, not fundamentals, and TSM's own analyst consensus still shows +28% upside with 2 BUY votes. That's a 'hold the winner' case, not an add case given concentration. On the buy side, the only names with real room under the cap are smaller legs (OKLO, LEU, CCJ, NVDA-1-share, MRVL-1-share, etc.), and I want a name with a distinct thesis rather than more overlap with what I already own. VRT stands out: analyst consensus BUY +0.80 with 40% upside, data-center/grid power-delivery thesis distinct from LRCX/TSM's fab-equipment and foundry exposure, and it's down from its 290 high to 280 on tape noise (grid basket -4.4% 5d) not a broken thesis. But VRT isn't in the current legal BUY WINDOW list for cash reasons (not listed with a share range), so I'll skip it this cycle rather than force something. Instead, I'll add a single small OKLO position — small modular nuclear is a distinct sub-theme (SMR, +7.2% 5d, contrarian to the broader nuclear basket's -1.5%/-6.6%, and only 22% nuclear exposure currently all in VST which is large-cap regulated + retail). OKLO's edgar signal (dilution/insider sales) is a real risk given the 424B5 shelf filing, so I'll size this modestly (3 shares, ~$124) and set a wider stop given OKLO's volatility. This diversifies nuclear away from single-name VST risk without breaching cash constraints, leaving ~1,150 DKK cash buffer intact. TSM (33%), VST (22%) and LRCX (21%) already carry the book's thesis and TSM/LRCX/MU/GEV/PWR/AMD are cap-blocked from further buys — adding here would just concentrate further into names already near the ceiling, not diversify. The 5-13% pullbacks in semicap/memory/nuclear look like sector noise (VIX 17.7, no earnings surprise, no policy shift) rather than a broken thesis, so no sells either. Keeping the single OKLO add small and preserving cash buffer since today's window is thin and this account trades once a day with no ability to revise stops later.

Horisont: 2-3 years Begrundelse: Independent power producer with direct leverage to datacenter/AI electricity demand and strong analyst upside (+47% to target); exit if PJM capacity price reform or a large customer contract cancellation breaks the power-shortage thesis. === Session-analyse === [saxostrategi/ALL] This is a fresh, all-cash book on day one — the starting basket matters more than any single day's noise. The tape is broadly red across every theme (nuclear -8.8%/30d, semis -7.2%, memory -7.3%), which for a multi-quarter structural thesis is an entry opportunity, not a warning. I want to seed all three legs (electricity, semis, memory) with names that have both a stated thesis and cross-feed support, while respecting the cash budget and avoiding the CAP-BLOCKED names (AMD, GEV, MU, PWR are blocked this cycle regardless of sizing). For semiconductors, TSM stands out: 31 insider buys totaling ~$276k is a strong signal of management conviction at these levels, and analyst consensus is BUY. For memory/semicap, LRCX is down hard (5d -10%, range_pos 0%) — the memory/semicap cycle thesis is intact (HBM/DRAM demand from AI buildout hasn't changed), and this is a battered entry into a name I want structural exposure to, not a broken thesis. For electricity/grid, VST has strong analyst support (+47% upside, 1.4/5 rating) and nuclear-adjacent power generation exposure that benefits from datacenter power demand — CEG's single SELL vote and steep 5d drop makes me prefer VST as the cleaner entry into the same theme today. Sizing: USD budget is ~$1,251. TSM at ~$418 (1 share, capped anyway) ≈ $418. LRCX capped at 1 share ≈ $301. VST 1-3 shares available; I'll take 2 at ~$143.79 ≈ $288 to keep some room. Total ≈ $1,007, leaving buffer under the $1,251 ceiling and well within the DKK cash converted. Stops are set at 20-22% — wide enough to ride out normal 5-8% sector noise cited in the mandate, tight enough to protect capital if a thesis actually breaks. No DK-listed name in the universe today, so no .CO trade this cycle. Seeding the book today across all three mandate legs — semiconductors (TSM), semicap/memory-adjacent (LRCX), and grid/electricity (VST) — using cash that would otherwise sit idle. GEV, AMD, MU, PWR are cap-blocked this cycle so left untouched; CEG's sharp single-feed SELL and steep drop makes VST the cleaner grid entry today. No DK-listed names in today's universe, so the Copenhagen leg stays empty this cycle.

Horisont: 2-3 years Begrundelse: Semicap equipment maker sold off to 5-day/30-day lows despite intact memory/logic capex upcycle thesis; exit if fab capex spending across TSM/Samsung/Micron actually contracts. === Session-analyse === [saxostrategi/ALL] This is a fresh, all-cash book on day one — the starting basket matters more than any single day's noise. The tape is broadly red across every theme (nuclear -8.8%/30d, semis -7.2%, memory -7.3%), which for a multi-quarter structural thesis is an entry opportunity, not a warning. I want to seed all three legs (electricity, semis, memory) with names that have both a stated thesis and cross-feed support, while respecting the cash budget and avoiding the CAP-BLOCKED names (AMD, GEV, MU, PWR are blocked this cycle regardless of sizing). For semiconductors, TSM stands out: 31 insider buys totaling ~$276k is a strong signal of management conviction at these levels, and analyst consensus is BUY. For memory/semicap, LRCX is down hard (5d -10%, range_pos 0%) — the memory/semicap cycle thesis is intact (HBM/DRAM demand from AI buildout hasn't changed), and this is a battered entry into a name I want structural exposure to, not a broken thesis. For electricity/grid, VST has strong analyst support (+47% upside, 1.4/5 rating) and nuclear-adjacent power generation exposure that benefits from datacenter power demand — CEG's single SELL vote and steep 5d drop makes me prefer VST as the cleaner entry into the same theme today. Sizing: USD budget is ~$1,251. TSM at ~$418 (1 share, capped anyway) ≈ $418. LRCX capped at 1 share ≈ $301. VST 1-3 shares available; I'll take 2 at ~$143.79 ≈ $288 to keep some room. Total ≈ $1,007, leaving buffer under the $1,251 ceiling and well within the DKK cash converted. Stops are set at 20-22% — wide enough to ride out normal 5-8% sector noise cited in the mandate, tight enough to protect capital if a thesis actually breaks. No DK-listed name in the universe today, so no .CO trade this cycle. Seeding the book today across all three mandate legs — semiconductors (TSM), semicap/memory-adjacent (LRCX), and grid/electricity (VST) — using cash that would otherwise sit idle. GEV, AMD, MU, PWR are cap-blocked this cycle so left untouched; CEG's sharp single-feed SELL and steep drop makes VST the cleaner grid entry today. No DK-listed names in today's universe, so the Copenhagen leg stays empty this cycle.

Horisont: 3+ years Begrundelse: Leading-edge foundry monopoly on AI/HPC silicon with 31 recent insider buys signaling management conviction at current levels; exit if foundry share loss to Samsung/Intel becomes real or capex guidance breaks down. === Session-analyse === [saxostrategi/ALL] This is a fresh, all-cash book on day one — the starting basket matters more than any single day's noise. The tape is broadly red across every theme (nuclear -8.8%/30d, semis -7.2%, memory -7.3%), which for a multi-quarter structural thesis is an entry opportunity, not a warning. I want to seed all three legs (electricity, semis, memory) with names that have both a stated thesis and cross-feed support, while respecting the cash budget and avoiding the CAP-BLOCKED names (AMD, GEV, MU, PWR are blocked this cycle regardless of sizing). For semiconductors, TSM stands out: 31 insider buys totaling ~$276k is a strong signal of management conviction at these levels, and analyst consensus is BUY. For memory/semicap, LRCX is down hard (5d -10%, range_pos 0%) — the memory/semicap cycle thesis is intact (HBM/DRAM demand from AI buildout hasn't changed), and this is a battered entry into a name I want structural exposure to, not a broken thesis. For electricity/grid, VST has strong analyst support (+47% upside, 1.4/5 rating) and nuclear-adjacent power generation exposure that benefits from datacenter power demand — CEG's single SELL vote and steep 5d drop makes me prefer VST as the cleaner entry into the same theme today. Sizing: USD budget is ~$1,251. TSM at ~$418 (1 share, capped anyway) ≈ $418. LRCX capped at 1 share ≈ $301. VST 1-3 shares available; I'll take 2 at ~$143.79 ≈ $288 to keep some room. Total ≈ $1,007, leaving buffer under the $1,251 ceiling and well within the DKK cash converted. Stops are set at 20-22% — wide enough to ride out normal 5-8% sector noise cited in the mandate, tight enough to protect capital if a thesis actually breaks. No DK-listed name in the universe today, so no .CO trade this cycle. Seeding the book today across all three mandate legs — semiconductors (TSM), semicap/memory-adjacent (LRCX), and grid/electricity (VST) — using cash that would otherwise sit idle. GEV, AMD, MU, PWR are cap-blocked this cycle so left untouched; CEG's sharp single-feed SELL and steep drop makes VST the cleaner grid entry today. No DK-listed names in today's universe, so the Copenhagen leg stays empty this cycle.

Horisont: 1-2 weeks Begrundelse: Cross-feed consensus BUY driven by 16.9% short interest in an uptrend, in a risk-on tape with no theme-level loss history; exit on a close under the stop or a cross-feed flip to SELL. === Session-analyse === [wsb/US] Cycle 1 of 4 on a RISK-ON Monday (QQQ +1.24%, VIX 16.5, Crypto F&G 59). I'm 100% cash after the GOOGL soft-stop fire Friday, and the last six sessions were HOLDs, so no thesis is being recycled. The WSB hot listing is mostly loss-porn and memes (0DTE blowups, 'Semis 9 down days in a row', 'Buckle up'), with a heavy capitulation vibe. That reads more as a washed-out tape than a euphoric one, and the overnight gap-ups (SOXL +9%, APLD +7.8%, ASTS +6.4%) look like a relief bounce rather than fresh catalysts. On the cross-feed side, the server requires at least 2 BUY votes, but the wsb feed counts as one of them, so I want a non-Reddit independent voter too. CLSK is the cleanest: 3 BUY / 1 SELL with short interest at 41.6% of float, which is the squeeze fuel I care about. But CLSK is crypto_mining, where THEME-FACIT shows 12 exits and -930 DKK, the theme tape is -40% over 30d, and a re-entry-style bar applies to the whole theme. No theme-external catalyst. I'm skipping miners (MARA/RIOT/WULF/IREN too). NBIS is up massively and gapped +3.8%, but it sits in the losing ai_infra theme and WSB loves it, which is crowding risk. SOXL has a SELL consensus and a violent 5d drop; I won't catch that knife. Best actionable idea: SOFI, which has 2 BUY votes (short_interest 16.9% of float, plus a second feed). It is fintech, a theme with no realized losses on record, it is in a mild uptrend (14.46 → 15.76), it has no earnings in the window, and its fee is the flat $1 minimum. Entry ~$15.76; 25 shares ≈ $394 stays under the ~$432 25% position cap and clears the $200 preferred size. It is not a screaming WSB name, so I'm being honest that this is a cross-feed/short-interest trade, not WSB hype. Size is modest given the regime is risk-on but the sub is deeply wounded. Stop at 9%, outside SOFI's typical daily swing, with a 12-day horizon. This is not the last cycle of the day, so no overnight-risk restriction applies. I'm leaving most cash for later cycles when the picture is clearer, and I'm avoiding earnings names (META, NFLX, TSLA, IBM, SPCX). Only one name cleared the cross-feed and theme-history bar. The WSB feed is dominated by loss-porn/capitulation rather than fresh hype, the crypto-mining and ai_infra themes carry losing track records with negative tapes, and several big names sit on earnings prints this week. Most cash stays in reserve for later cycles.

SOFT STOP TRIGGERED: GOOGL faldt under 3.5% stop (337.48 → 341.28). Auto-cut før agent-call.

SOFT STOP TRIGGERED: RDDT faldt under 7% stop (153.16 → 148.66). Auto-cut før agent-call.

SOFT STOP TRIGGERED: NVDA faldt under 7% stop (232.51 → 221.99). Auto-cut før agent-call.

SOFT STOP TRIGGERED: APLD faldt under 10% stop (23.51 → 20.90). Auto-cut før agent-call.

SOFT STOP TRIGGERED: AVGO faldt under 7% stop (347.92 → 351.81). Auto-cut før agent-call.

SOFT STOP TRIGGERED: SMCI faldt under 8% stop (40.62 → 41.30). Auto-cut før agent-call.

Horisont: 1-2 weeks Begrundelse: WSB post on APLD's fiscal Q1 earnings beat (net-sentiment +1.50, high-conviction, concrete catalyst) plus cross-feed consensus BUY with analyst target +175% upside — exit if the earnings pop fades or ai_infra tape keeps bleeding (-1.6% 5d). === Session-analyse === [wsb/US] RISK-OFF day (QQQ -0.49%) with crypto-mining names gapping down hard overnight (RIOT -8%, CLSK -7%, WULF -6.9%, IREN -6%) — exactly the theme (crypto_mining) where my own 90d facit shows -1170 DKK across 14 exits. Staying away from that whole bucket, re-entry bar is high and not cleared. APLD shows up with a fresh WSB post (fiscal Q1 earnings beat, net-sentiment +1.50, high-conviction, katalysator 1/1) plus cross-feed consensus BUY (2 BUY, 0 SELL, analyst target +175% upside) — this is a NEW/RISING-ish name with real corroboration beyond WSB alone (analyst_ratings independent feed), clearing the ≥2 cross-feed bar. ai_infra theme is already ⚠ AT CAP at 67% though, so I'm sizing this small and deliberately nudging over rather than loading up — APLD's earnings-beat catalyst is name-specific, not just theme momentum. NVDA mention-leader in title score (3,563) is SPIKE-CHRONIC — loss-porn thread ('too scared to sell'), 40% loss-porn in the interpreted read, net-sentiment lukewarm at +0.50 — this is the crowd already in, no fresh edge, and I'm CAP-FULL on NVDA anyway per server block. MU is RISING (z=+3.3σ) with a 'unfuck this by 4pm' loss-porn thread and a bounce-at-support thread, but MU itself is in the UNTRADEABLE-at-min-size list this cycle — skip. Existing book (AVGO, GOOGL, NVDA, RDDT, SMCI) stays put; SMCI gapped down -4.3% overnight on no ticker-specific bad news I can see (just risk-off chip-sector tape), RDDT actually gapped UP +1.7% and has cross-feed BUY backing (analyst target +40%) so letting it ride. No forced selling — nothing is >10% down AND forgotten; all core holdings are green or barely red and still within their thesis windows. Risk-off tape + crypto-mining crater overnight keeps me away from the chronic-loser mining theme entirely. NVDA is spike-chronic loss-porn and already cap-full; MU is untradeable at min-size this cycle despite RISING tag. Core book (AVGO, GOOGL, NVDA, RDDT, SMCI) holds — all within +1.4% to +6.8%, no forced sells triggered by rule #4, and ai_infra theme is AT CAP so only a small deliberate APLD nudge goes in on its own earnings catalyst, not theme momentum.

Horisont: closed Begrundelse: Crypto_mining theme forgotten by WSB today (not in trending/mentions), gapped down -3.4% overnight, no cross-feed corroboration listed, and theme has -1134 DKK round-trip history — cutting the bagheld loser per rule #4. === Session-analyse === [wsb/US] Risk-off tape today (QQQ -0.92%, SPY -0.58%) with VIX still calm at 15.5 — not a day to chase, but ai_infra is already at 79% exposure (⚠ AT CAP) with AVGO/GOOGL/NVDA flagged CAP-FULL, so no adds there regardless of WSB noise. NVDA is the #24,548-score meme-of-the-day post ('1280% gain, too scared to sell') and cross-feed shows consensus BUY with analyst target +37% upside, but it's SPIKE-CHRONIC (always-trending, no +0.1 bump) and I already hold a CAP-FULL position — the correct move is just to hold and tighten, not add. AMD and MU are both untradeable-at-min-size per this cycle's params, so that AMD leap/MU patience chatter is noise I can't act on anyway. WULF gapped down -3.4% overnight into a crypto_mining theme that's bleeding -7.3% over 30d and not mentioned anywhere in today's trending/mention-leaders — rule #4 says cut the forgotten loser, and there's no cross-feed consensus BUY line for WULF to save it (it's not even listed in the cross-feed screening block, meaning no corroboration). My own THEME-FACIT shows crypto_mining is -1134 DKK over 13 exits — this is a recognized bleeder theme, not a fresh thesis. RDDT is down slightly but DOES have cross-feed consensus BUY (analyst +43% upside) and genuine WSB mention today (1 high-conviction post, net-sentiment +1.50) — holding that one per the consensus-override rule. Given the risk-off regime, AT-CAP ai_infra theme, and no fresh cross-feed-corroborated name outside my capped/forgotten holdings, cash is the right call today — sitting tight and letting the existing winners (AVGO, SMCI, NVDA) run with tightened stops rather than forcing new entries into a red tape. QQQ -0.92% risk-off day raises the bar for new longs. ai_infra theme is at 79% (⚠ AT CAP) and AVGO/GOOGL/NVDA are explicitly CAP-FULL — can't add to the exact names WSB is hyping (NVDA mega-post, AVGO YOLO thread) without breaching discipline. AMD/MU are untradeable at min size this cycle despite trending mentions. RDDT holds on cross-feed consensus BUY despite a minor overnight dip. Sitting on the WULF sale proceeds in cash rather than forcing a redeploy into a capped theme or a risk-off tape — will re-scan at 11:30 for a fresh, uncapped, cross-feed-corroborated name.

Horisont: 3-7 days Begrundelse: WSB's #1 trending post today is a 4.3M YOLO on WULF (1,234 score, 431 comments) and cross-feed shows 3 BUY votes incl. analyst target +118% upside, but it gapped -4.8% overnight so I'm sizing smaller and setting a wider stop to respect that risk flag going into the close. === Session-analyse === [wsb/US] Last cycle of the day, RISK-ON tape (QQQ +0.96%), and the book is already sitting at the ai_infra cap (67%) with AVGO/GOOGL/NVDA all CAP-FULL — no room to add there even though NVDA's hot-listing score is absurd (17,653 upvotes, pure meme-title with no body, cross-feed BUY but mostly analyst-driven not a fresh WSB catalyst). WULF is the one name that clears every bar: #1234-score YOLO post (NEW/RISING, not spike-chronic), cross-feed consensus BUY (3 BUY votes incl. analyst_ratings +0.91 target +118% upside), and it's in crypto_mining which despite a rough 90d theme history (-1134 DKK) has zero current exposure in this book, so this isn't stacking a losing bet. However, it gapped DOWN -4.8% overnight into the open, which per rule #16 is a risk flag worth respecting on a late-cycle entry — I'll size it small and give a wider stop since it's high-beta, and treat the overnight gap risk explicitly rather than ignoring it. NVDA/GOOGL/AVGO are CAP-FULL in the ai_infra theme (already at the 67% ⚠ AT CAP marker) so no adds despite NVDA's insane hot-listing score — that score is pure meme-title noise (0 body text, interpreted signal +0.25 but zero high-conviction) riding spike-chronic territory, not a fresh catalyst. NKE is cross-feed SELL with WSB loss-porn thesis (earnings miss meme) — avoiding. TSLA interpreted signal is net-bearish (-0.38, 100% loss-porn, puts positioning) despite 2 mentions, so skipping. It's the last cycle of the day, so I'm keeping new exposure light (just WULF, small size, wider stop) rather than opening anything that needs babysitting overnight — RDDT stays held since cross-feed still shows consensus BUY despite WSB going quiet on it, satisfying the rule #4 exception.

SOFT STOP TRIGGERED: IONQ faldt under 8% stop (40.61 → 42.69). Auto-cut før agent-call.

Horisont: 1-2 weeks Begrundelse: Cross-feed BUY consensus (2 BUY, 0 SELL) riding the MU earnings-beat semis tailwind WSB is buzzing about today, without adding to an already-losing position. === Session-analyse === [wsb/US] MU dominates WSB today (RISING tag, z=+3.0σ, 74 comment-flow mentions, strong post titles about the earnings beat) and cross-feed backs it with a BUY consensus (analyst 1.4/5, target +43% upside) — but MU itself isn't in our tradeable universe at this size (flagged UNTRADEABLE at min-size), so the only way to play the theme is via adjacent names. AVGO and AMD both carry cross-feed BUY consensus and sit in semiconductors, a theme already at 13% exposure (room to add) riding +10.5% 30d tape — but ai_infra/semis combined are already 46%+13%=59% if we stack more, so I'll be selective. AMD is untradeable at min size per the UNTRADEABLE list, leaving AVGO as the cleanest semis add with 2 BUY cross-feed votes and no SELL. Existing book: IONQ (+9.8%) has fresh cross-feed BUY backing (analyst +0.89, +52% target upside) despite quantum tape being soft (-3.8% 5d) — worth tightening the stop rather than adding, since quantum theme is already 20%. SMCI (+2.7%, bought last cycle) has cross-feed BUY confirmation too, holding. GOOGL is CAP-FULL so no action there. NVDA is CAP-FULL and basically flat — holding, gap-up overnight modest. RDDT gapped +3.6% overnight on no fresh WSB catalyst (not in trending/mentions this session) and is down -3.6% from my avg — cross-feed wasn't shown for RDDT so no override; it's a small, single-share, low-conviction legacy position I'll leave alone rather than force an exit on a position this small (fee drag not worth it either way). New money: with RISK-ON regime (QQQ +0.44%), VIX calm at 16, and cash at 25% of book, I'll deploy into AVGO (cross-feed BUY, semis tailwind, NOT yet in my book so no theme-stacking on an existing loser) sized modestly given its high per-share price limits shares to 1. This is a single-share high-conviction add per the exception rule for names ≥$50. I'm skipping NKE (earnings today, WSB sentiment net-bearish per INTERPRETED signal '-1.00 skeptical of NKE beating earnings' — avoid pre-print chase) and skipping MU-adjacent plays beyond AVGO to avoid conviction-leakage into an already crowded ai_infra/semis book this late in the day (cycle 3 of 4, last cycle later today per rule #19 raises bar further). GOOGL and NVDA are CAP-FULL — no room to add regardless of signal. RDDT is a single-share legacy loser with no fresh WSB mention or cross-feed row shown this cycle; too small to matter either way on fees, leaving it alone rather than forcing a cut. MU itself is untradeable at our size so I'm expressing the theme via AVGO only, deliberately avoiding stacking further into ai_infra/semis which is already the dominant exposure (46%+13%=59% combined) this late in the day (cycle 3 of 4) when a fresh overnight-hold needs a real catalyst, not just hype.

Horisont: 1-2 weeks Begrundelse: MU's blowout earnings (WSB's #1 hot topic today, 2,659-score YOLO + multiple earnings threads) is lifting the whole memory/AI-infra semis tape and SMCI carries its own cross-feed BUY consensus with short-interest support — MU itself is untradeable at min size so this is the cheaper access point to the same catalyst. === Session-analyse === [wsb/US] MU is the clear WSB story today — RISING tag (z=+3.0σ), top of hot with a 2,659-score YOLO post plus a cluster of earnings-beat threads (18, 19, 21), net-sentiment +0.53 with real skin-in-the-game (3 shares/1 calls disclosed) and a 5/1 bull split. Cross-feed backs it: consensus BUY (2 BUY incl. analyst_ratings +0.83 target +43% upside). That clears the ≥2 BUY-vote corroboration bar cleanly. MU itself is untradeable at min size per the UNTRADEABLE list (too expensive, ~$1,015/share) — so I can't play the actual ticker, but AMD and SMCI sit adjacent in the same semiconductors/AI-infra theme-tape (semis 5d +1.3%, 30d +10.7%, tailwind) with their own cross-feed BUY consensus (AMD 2 BUY/3 HOLD, SMCI 2 BUY/2 HOLD short_interest-backed). AMD is flagged untradeable at min size too, so SMCI is the real access point — cheap enough for a real scale-out position, cross-feed BUY, and rides the same earnings-beat tailwind without chasing MU's own after-earnings pop. Existing book: IONQ is my best winner (+8.1%), cross-feed BUY (analyst target +52% upside) despite quantum theme-tape being soft (5d -3.7%) — I'll hold and nudge the stop up since it's not yet at the +15% ladder trigger. RDDT dipped -6.2% but still shows up in comment-flow ($RDDT 4✓) and gapped +0.9% overnight — not yet 'forgotten' by rule #4, and no cross-feed SELL consensus on it, so holding rather than cutting into a neutral tape. GOOGL and NVDA are CAP-FULL, no adds possible regardless of the Gemini 4 buzz (post #12, 816 score) — just holding. Quantum theme is already at 20% and ai_infra at 32%, approaching the 50% cap collectively, so I'm deploying new cash into semiconductors (13% current, room to grow) rather than piling onto saturated themes. NASDAQ is neutral today (QQQ -0.05%), VIX calm at 16 — no regime reason to go big, so sizing SMCI modestly within the $200+ min-trade floor. GOOGL and NVDA are CAP-FULL (position-cap blocks any add) despite the Gemini 4 buzz (post #12) and NVDA's AI-datacenter mention — holding as-is. RDDT down -6.2% but still appearing in comment-flow and no cross-feed SELL, not yet a clear 'forgotten bagholder' cut under rule #4. Quantum (IONQ) and ai_infra (GOOGL/NVDA) themes are already sizable (20%/32%) so new cash is routed to semiconductors instead of adding exposure there. Sitting mostly in cash otherwise — WSB's real fresh energy today is concentrated in MU, which the sizing rules block directly, so a single modest semis trade is the deliberate, non-forced move this cycle.

Horisont: 1-2 weeks Begrundelse: China/ByteDance-Alibaba Nvidia chip-buying news post (1,048 upvotes) plus INTERPRETED signal +0.75 net-sentiment with disclosed positions and 3/4 catalyst-cited posts, backed by cross-feed consensus BUY; sizing small given RISK-OFF tape and the name already gapped +3.3% overnight. === Session-analyse === [wsb/US] Today's WSB feed is genuinely NVDA-heavy: the China/Nvidia chip news post is high-scoring (1,048 upvotes) and the INTERPRETED signal shows +0.75 net-sentiment with 3/4 posts citing a concrete catalyst and actual disclosed positions (2 shares, calls) — this is real conviction, not drive-by hype. Cross-feed backs it up: NVDA consensus BUY (3 BUY/1 SELL/2 HOLD) with analyst_ratings at +0.88 (+46% upside target). NVDA also gapped up +3.3% overnight on the China chip-buying news, which is momentum confirmation. However, it's RISK-OFF today (QQQ -0.78%, SPY -0.57%) and NVDA already ran on the gap, so I'm sizing modest rather than full — this is a smaller starter position with a wider stop to respect the gap risk, not a chase into full size. META is also loud (7 mentions today, highest mention-leader, +1.33 net-sentiment, 2 high-conviction posts with disclosed skin-in-the-game) but it is UNTRADEABLE at min size per this session's params — skip it, no way around the constraint. IONQ continues to work (+16.9% unrealized, cross-feed consensus BUY with analyst target +48% upside, quantum theme +5.2% over 5d) — I'll let this winner run and tighten the trailing stop per the ladder rather than trim it. RDDT is down -4.2% and gapped down another -2.1% overnight with no fresh WSB catalyst today (only comment-flow mentions, no trending/DD presence) — but it still shows in comment-flow with real engagement and I don't have 2+ cross-feed BUY votes to hold conviction on adding; I'll hold it flat for now rather than force an exit at a 4% loss on a name that hasn't cleared the 10% forgotten-bagholder threshold yet. GOOGL is CAP-FULL per constraints — cannot add. Given RISK-OFF regime, calm VIX (14.2) but crypto Fear&Greed at 74 (Greed, blow-off risk for meme/crypto names), I'm avoiding fresh crypto-mining or meme-stock entries entirely (matches my own THEME-FACIT: crypto_mining has been a -1688 DKK loser across 17 exits — no fresh evidence justifies re-entering that theme). Staying disciplined: one modest NVDA add, hold everything else, tighten IONQ's stop to lock in the +16.9% gain. Holding RDDT flat — down -4.2% and gapped -2.1% overnight with no fresh WSB catalyst today (comment-flow only, not trending), but it hasn't crossed the >10% forgotten-bagholder threshold yet so I'm not forcing a cut into a red tape day. Holding GOOGL (CAP-FULL, can't add anyway, small +1.5% gain intact). Sitting on 52% cash given RISK-OFF regime (QQQ -0.78%) — no crypto-mining or meme names cleared the bar today, and my own THEME-FACIT shows that theme has been a consistent loser. META and AMZN both show strong INTERPRETED sentiment but META is untradeable at min size this session and AMZN has no cross-feed corroboration shown — passing on both rather than forcing a low-conviction entry.

Horisont: 1-2 weeks Begrundelse: The $146K RDDT Jan 2028 calls YOLO post plus cross-feed consensus BUY (analyst target +40% upside) back a small momentum add in the social theme tape running +9% over 5 days. === Session-analyse === [wsb/US] WSB's hot list today is thin on fresh, tradeable conviction: NKE is pure loss-porn meme content (net-sentiment -0.25, thesis: loss_porn), META has genuine momentum (net +0.92, skin-in-the-game 3-share disclosure, cross-feed BUY with wsb as top signal) but I already have zero exposure there and it's not in my current sub-theme mix — worth a small starter given RISK-ON regime + cross-feed corroboration. RDDT has a single $146K YOLO calls post (100% loss-porn framing though, thin sample = weak signal per rule #15) but cross-feed shows consensus BUY (analyst_ratings +0.64, 2 BUY/0 SELL/3 HOLD) — decent secondary corroboration for a small position. Crypto mining names (CLSK, RIOT, WULF, IREN) show strong cross-feed BUY consensus, but my THEME-FACIT shows -1770 DKK across 18 exits in this exact theme, WULF just stopped out yesterday (re-entry blocked), and IREN/COIN are also on my recent-exits cooldown list — this is a theme I've bled money on repeatedly with no theme-external catalyst, so I'm passing despite the cross-feed noise. Existing holdings: GOOGL is at 25% cap already (CAP-FULL, no adds allowed) — holding as-is, small +2.1% gain, quantum/internet/mega_cap themes fine. IONQ is up +10.3%, cross-feed shows consensus BUY (3 BUY/0 SELL/1 HOLD, analyst target +49% upside), quantum theme showing +8.6% 5d tape strength — this is a winner to protect with a trailing stop rather than trim, per rule #2. QQQ is green today (RISK-ON +0.36%), VIX is calm at 14.21, so leaning in on quality names makes sense, but I'm being selective given the thin WSB catalysts today — most of the hot list is macro-macro chatter (bonds, yields, Bessent) rather than actionable single-name hype. Trades: small META starter (cross-feed BUY consensus, wsb sentiment strongly positive, RISK-ON tape) and a small RDDT position (cross-feed BUY, calls YOLO is on-theme for social momentum tape +9.0% 5d). Both sized modestly given thin sample sizes and no ticker in my top-3 trending list clearing every bar. Raising IONQ's stop to lock in gains as it approaches the +15% ladder threshold. GOOGL is CAP-FULL at 25% position size — no adds allowed regardless of WSB chatter (Google's Gemini 4 news is a mild positive but the position is maxed). Crypto mining names (WULF/IREN/CLSK/RIOT/MARA) show strong cross-feed BUY votes but my THEME-FACIT shows -1770 DKK across 18 round-trips in this exact theme with zero improvement, WULF just stopped out yesterday and IREN is on cooldown from a recent exit — repeating this bet without a theme-external catalyst is exactly the mistake the facit warns against, so I'm sitting out despite the signal. NKE's WSB post is pure loss-porn mockery, not a thesis. Staying mostly in cash (61%+) given today's hot list is dominated by macro/Bessent/bond chatter rather than fresh single-name catalysts — the two small adds (META, RDDT) are the names that cleared both WSB sentiment and cross-feed corroboration bars.

SOFT STOP TRIGGERED: WULF faldt under 8% stop (17.11 → 16.03). Auto-cut før agent-call.

Horisont: 1-2 weeks Begrundelse: Crypto_mining theme tape +9.9%/5d and +17.5%/30d with cross-feed 4-BUY/0-SELL consensus (analyst target +102% upside) — adding to an already-green position on strong external corroboration, not just WSB hype; exit if theme tape rolls over or WULF loses the $16 level. === Session-analyse === [wsb/US] WSB is entirely META/SPX-options theater today (irony-heavy, mixed bull/bear, 50% loss-porn) — not tradeable, META isn't even in my universe. AMD is genuinely hot (trillion-dollar cap, 5-day rally, z=+2.5σ, cross-feed BUY with 5d +24.7%) but it's priced at $615+ and flagged UNTRADEABLE at min size for this account — skip. That leaves my actual holdings: WULF has the strongest cross-feed backing I've seen (4 BUY/0 SELL, analyst target +102% upside, short interest fuel) and crypto_mining theme tape is ripping +9.9%/5d, +17.5%/30d — this is a real tailwind, not just chatter. GOOGL is up nicely (+7.2%) but CAP-FULL at position cap, so no add there, just let it run. IONQ is flat/slightly red, quantum theme is down -17.9% over 30d — no name-specific catalyst here, cross-feed shows BUY (3-1-0) but the theme tape is weak and WSB isn't mentioning it today, so I'll hold without adding. WSB's hot list today is a META options circus (irony sentiment, 50% loss-porn, mixed 9-bull/11-bear split) and a viral SPX 0DTE YOLO post — neither maps to my tradeable universe or a real thesis. AMD is the one legitimately fresh/rising ticker (trillion-cap milestone, cross-feed BUY, +24.7% 5d) but it's flagged untradeable at min size here, so no entry despite the hype. IONQ stays held flat — cross-feed still shows consensus BUY so rule #4's forgotten-name cut doesn't trigger, but I won't add given the quantum theme's ugly -17.9% 30d tape and zero fresh WSB mentions today. Risk-on regime (QQQ +0.35%), calm VIX (14.8), and Extreme Greed on crypto fear/greed (78) support leaning into my highest-conviction existing winner (WULF) rather than chasing new untested names, while tightening stops on GOOGL and WULF to lock in gains as they extend.

Horisont: 1-2 weeks Begrundelse: Fresh +7.5% overnight gap with cross-feed consensus BUY and huge analyst upside target (+131%) — sizing smaller given crypto_mining theme has been this book's worst bleeder historically, exit if gap fades below $15.5 support. === Session-analyse === [wsb/US] RISK-ON tape today (QQQ +1.66%) and the overnight gap sheet is full of green in the AI/quantum/nuclear/mining names (OKLO +12.8%, IONQ +10.2%, SMR +9.0%, WULF +7.5%, ARM +7.5%), but WSB itself is thin on fresh signal — the hot list is dominated by macro news (Fed rate hike, CLARITY Act failure) and loss-porn memes, not a specific ticker rally. The one name that stacks WSB relevance + cross-feed corroboration + a real overnight gap is IONQ: consensus BUY (3-0-2, analyst target +88% upside), INTERPRETED signal net +0.50 with a cited catalyst ('hybrid quantum apps show utility'), quantum theme 5d +8.2%, and it gapped +10.2% overnight on real momentum — not chasing a stale spike. WULF is the other qualifier: consensus BUY (3-0-2), analyst target +131% upside, and a clean +7.5% gap, though I'll size it smaller given the crypto_mining theme has been my worst bleeder (-1650 DKK over 18 exits) — this is a fresh name+catalyst (WULF specifically, not a recycled MARA/RIOT bet) so I'll allow a small toe-in rather than a full pass. GOOGL stays untouched — CAP-FULL per the params, small +2.4% winner, nothing to add or trim. COIN/NVDA/NBIS/MP are all on the re-entry-gated list from recent exits and none of today's cross-feed shows the required 2+ fresh BUY votes plus a genuinely new catalyst, so they stay out per rule #8. Given it's cycle 3 of 4 (not the last cycle yet), I have room to open positions with normal stops rather than needing overnight-survival conviction — but I'm still keeping size moderate since IONQ/WULF are both high-beta names in a book that's been over-concentrated in mining/AI-infra churn historically (THEME-FACIT crypto_mining and ai_infra both net negative). Total new deployment stays well under the 25% single-position cap and doesn't push any sub-theme over 50%. GOOGL is CAP-FULL and only +2.4% — no add possible, no reason to trim a modest winner; holding as-is with no stop tightening needed yet since it hasn't cleared the +15% ladder trigger. Passing on COIN/NVDA/NBIS/MP (recent exits, re-entry-gated, no fresh 2+ cross-feed BUY consensus with a genuinely new catalyst today) and on SPCX/AMD (today's actual top mention-leaders) since neither clears cross-feed corroboration — SPCX INTERPRETED read is split bull/bear with 50% loss-porn, and AMD is UNTRADEABLE at min size per params anyway. Staying selective given RISK-ON tape doesn't mean forcing size into names without the corroboration bar.

Horisont: 1-2 weeks Begrundelse: Overnight +10.2% gap on real quantum-theme momentum (5d +8.2%) plus cross-feed consensus BUY (analyst target +88% upside) and an INTERPRETED post citing a concrete utility catalyst — exit if the gap fully round-trips or quantum theme rolls over. === Session-analyse === [wsb/US] RISK-ON tape today (QQQ +1.66%) and the overnight gap sheet is full of green in the AI/quantum/nuclear/mining names (OKLO +12.8%, IONQ +10.2%, SMR +9.0%, WULF +7.5%, ARM +7.5%), but WSB itself is thin on fresh signal — the hot list is dominated by macro news (Fed rate hike, CLARITY Act failure) and loss-porn memes, not a specific ticker rally. The one name that stacks WSB relevance + cross-feed corroboration + a real overnight gap is IONQ: consensus BUY (3-0-2, analyst target +88% upside), INTERPRETED signal net +0.50 with a cited catalyst ('hybrid quantum apps show utility'), quantum theme 5d +8.2%, and it gapped +10.2% overnight on real momentum — not chasing a stale spike. WULF is the other qualifier: consensus BUY (3-0-2), analyst target +131% upside, and a clean +7.5% gap, though I'll size it smaller given the crypto_mining theme has been my worst bleeder (-1650 DKK over 18 exits) — this is a fresh name+catalyst (WULF specifically, not a recycled MARA/RIOT bet) so I'll allow a small toe-in rather than a full pass. GOOGL stays untouched — CAP-FULL per the params, small +2.4% winner, nothing to add or trim. COIN/NVDA/NBIS/MP are all on the re-entry-gated list from recent exits and none of today's cross-feed shows the required 2+ fresh BUY votes plus a genuinely new catalyst, so they stay out per rule #8. Given it's cycle 3 of 4 (not the last cycle yet), I have room to open positions with normal stops rather than needing overnight-survival conviction — but I'm still keeping size moderate since IONQ/WULF are both high-beta names in a book that's been over-concentrated in mining/AI-infra churn historically (THEME-FACIT crypto_mining and ai_infra both net negative). Total new deployment stays well under the 25% single-position cap and doesn't push any sub-theme over 50%. GOOGL is CAP-FULL and only +2.4% — no add possible, no reason to trim a modest winner; holding as-is with no stop tightening needed yet since it hasn't cleared the +15% ladder trigger. Passing on COIN/NVDA/NBIS/MP (recent exits, re-entry-gated, no fresh 2+ cross-feed BUY consensus with a genuinely new catalyst today) and on SPCX/AMD (today's actual top mention-leaders) since neither clears cross-feed corroboration — SPCX INTERPRETED read is split bull/bear with 50% loss-porn, and AMD is UNTRADEABLE at min size per params anyway. Staying selective given RISK-ON tape doesn't mean forcing size into names without the corroboration bar.

Horisont: closed Begrundelse: CLARITY Act Senate vote failed 49-50 overnight — the exact catalyst the top WSB YOLO post ('Ready for COIN to rippp ahead of senate vote') was betting on is now dead, and COIN gapped -5.8%; full exit rather than holding a broken-thesis position through the overnight gap risk on the last cycle of the day. === Session-analyse === [wsb/US] Last cycle of the day, RISK-OFF tape (QQQ -0.42%, SPY -0.85%), and WSB itself is mostly consumed with Fed-hike memes and CLARITY Act failure rather than any fresh actionable ticker. The CLARITY Act failing 49-50 is a real overnight-relevant negative catalyst for crypto policy names — COIN gapped -5.8% on it and is now down -5.3% unrealized on my position, having already been bought yesterday. WSB is still nominally hyping COIN ('Ready for COIN to rippp ahead of senate vote') but that thesis is now DEAD — the vote failed, the catalyst is spent, and this is exactly the discretionary full-exit case in rule #2: the setup that justified the buy no longer exists. I'm not averaging down into a broken catalyst going into an unprotected overnight hold on the last cycle of the day (rule #19) — cutting it now at a known price beats trailing a stop through the gap-risk night, especially with crypto_exchange theme already bleeding on the 5d tape (-7.6%). GOOGL is a small, already at-cap position (+1.1%, essentially flat) with no fresh WSB catalyst tonight — mention count is thin, and internet/ai_infra/mega_cap themes are all already carrying it fully. No reason to add (capped anyway) or cut (marginal gain, not down >10%, no distress). I'll leave it alone into the close. No new BUY clears the bar tonight: RDDT and FIG have decent cross-feed consensus BUY, but neither is in tonight's actual TRENDING/mention list (rule #18 — can't borrow WSB hype that isn't there), and it's the last cycle of the day with RISK-OFF regime — a fresh overnight entry needs a real dated catalyst, not just a technical cross-feed nod. NVDA and MRVL show up in mentions but consensus is SELL/thin-BUY and NVDA specifically nets SELL on cross-feed with only 1 BUY vote (analyst) — fails the ≥2 BUY-vote corroboration requirement anyway. Sitting mostly in cash into the close is the correct, disciplined call given the regime tilt and the empty WSB catalyst calendar for tonight — better to preserve capital for tomorrow's 09:45 read after the overnight settles the Fed-hike reaction and CLARITY fallout properly plays out in price. GOOGL stays untouched: small +1.1% position, already at the position cap (no room to add anyway), no fresh WSB mention tonight, not down >10% so rule #4 doesn't trigger a cut. No other name clears the bar for a fresh BUY on the last cycle: RISK-OFF regime (QQQ -0.42%, SPY -0.85%) raises the bar, WSB's actual hot listing tonight is dominated by Fed-hike/CLARITY-Act macro memes and non-tradeable tickers (DELL, KDK, AUR, DXYZ), and the only tradeable trending names (NVDA, MRVL, RKLB, ORCL) either fail the ≥2 cross-feed BUY-vote corroboration (NVDA is net SELL, 1 BUY vote), lack a dated catalyst to justify holding through the overnight gap (rule #19), or show negative INTERPRETED sentiment (ORCL -1.00, MRVL -1.00 on layoffs/pre-Fed call-selling). Preserving cash into the close given the fee-drag history (0.8% round-trip) and the -0.12%/day overnight-negative-pay stat on this universe is the disciplined move; will re-read at 09:45 tomorrow once the Fed reaction and CLARITY fallout are fully priced.

Horisont: 3-7 days Begrundelse: COIN is the top fresh WSB mention today (+2.0σ, CLARITY Act vote thread) with cross-feed consensus BUY (short_interest + analyst backing), despite loss-porn on the failed vote — sizing small given the bearish INTERPRETED read and Fed-week volatility. === Session-analyse === [wsb/US] Thin, mixed WSB tape this morning: the Daily Discussion dominates hot but the substantive posts are mostly loss-porn (RKLB fumble, 1.1M-to-150k, NVDA puts blown up, COIN calls burned on the failed CLARITY vote) — this is a fear-tinted sub right now, confirmed by the megathread pulse reading 20% bull / fear. COIN is the one name with real structure: 2 WSB mentions (+2.0σ vs its own baseline), cross-feed consensus BUY (3 BUY/1 SELL/4 HOLD) with short_interest backing 12.4% float short, and it's in the CLARITY Act news cycle (failed vote = fading regulatory clarity, but WSB posts are split on fading vs ripping into the vote, and analysts/short_interest still lean bullish). The INTERPRETED read is net-negative (-0.25, 100% loss-porn) on COIN specifically because of the failed vote, so I'll size small and lean on cross-feed corroboration rather than the Reddit hype itself. RDDT also shows cross-feed consensus BUY (analyst target +33% upside) but has zero fresh WSB mention today and my own TICKER-FACIT shows -458 DKK across 2 prior exits on this exact name — theme fatigue, no fresh catalyst, skip. ORCL is the #4 hot post but the DD/INTERPRETED read is unambiguously bearish (CFO comments post-layoffs, -1.00 net sentiment, fundamental thesis) — that's a name WSB is discussing negatively, not a buy signal. NVDA is a hot YOLO post but it's SPIKE-CHRONIC (always trending, no +0.1 bump), cross-feed consensus is SELL (3 SELL/1 BUY), and the local read is literally an author who blew up his account on NVDA puts — pure loss-porn, no edge, and I already ate two stop-outs on NVDA/NBIS this week from over-exposure to this exact ai_infra/semis complex. Staying out. Given RISK-ON regime (QQQ +0.47%) I'd lean into strength, but nothing here clears the ≥2 cross-feed-vote bar with genuine fresh WSB conviction except COIN, and even that has a bearish INTERPRETED qualifier. I'll take one small, corroborated position in COIN sized well under cap, with a wide-ish stop given crypto-adjacent volatility and event risk (Fed decision Wednesday, CLARITY Act fallout), and otherwise sit tight on GOOGL (holding fine, +2.2%, no reason to touch) and hold the rest in cash — 81% cash is the correct posture when the WSB vibe is fear/loss-porn dominant rather than screaming a fresh rocket. GOOGL is a small +2.2% winner with no cross-feed weakness and CAP-FULL on adds — just holding, no action needed. Most of today's hot WSB content is loss-porn (RKLB fumble, 1.1M→150k, blown NVDA puts, burned COIN calls) rather than fresh rocket conviction, and the megathread pulse reads only 20% bullish with a fear tone — this is not a day to chase size. NVDA, ORCL, RKLB and the crypto-miner complex (a theme I've already lost -1650 DKK on over 18 round-trips) all lack a genuinely fresh, corroborated catalyst today, so I'm passing on them and keeping ~79% of the book in cash as dry powder for a cleaner setup later in the session.

Horisont: closed Begrundelse: Crypto_mining theme absent from today's WSB trending/mention lists and overnight gap -4.1%; theme has been a -1590 DKK loser over 90d with no cross-feed BUY consensus to save it — full exit ahead of an unprotected overnight hold. === Session-analyse === [wsb/US] Last cycle of the day, QQQ -0.67% and SPY -0.48% — a risk-off tape — with the megathread PULS reading 'fear' and 80% bearish on top comments. This is not the session to open new overnight risk on thin conviction. COIN is the one trending name with fresh mention acceleration (+5 mentions today, CLARITY Act Senate vote catalyst) but the overnight gap was -8.9%, the INTERPRETED read is net-sentiment 0.00 (split bull/bear, zero high-conviction posts), and COIN has no listed cross-feed BUY consensus this cycle — a coin-flip binary bet into a failed Senate cloture vote is exactly what rule #19 says to skip late in the day. NVDA is a SPIKE-CHRONIC always-trending name (28k YOLO post, highest mention count) with a mixed/bearish comment-flow gist per PULS — crowding flag, not a fresh buy signal. On holdings: IREN gapped -4.1% overnight into today and sits at -3.4% unrealized; crypto_mining is a theme that's bled -1590 DKK over 90 days across 17 round-trips (my own theme-facit) and it isn't mentioned anywhere in today's TRENDING or MENTION-LEADERS lists — WSB has moved on from the miners entirely today. No cross-feed BUY-consensus line is shown for IREN to invoke the rule #4 override, so per rule #4 (forgotten + down, no consensus save) I'm cutting it now at a known price rather than holding an unprotected soft-stop through a night that's been measured as negatively-paid (-0.12%/day) with 44% of my own stop-fires historically landing in the first 30 minutes after open. GOOGL (+2%) is small and CAP-FULL per the strategy params — no action, just holding. No new BUYs this cycle: nothing clears the ≥2 cross-feed BUY-vote corroboration bar with genuine non-spike-chronic WSB momentum. Crypto Fear & Greed at 69 (Greed) is a blow-off-top caution on crypto-adjacent names per rule #12, the regime is risk-off per rule #11, and per rule #19 any late BUY needs a catalyst that survives an unprotected overnight gap — nothing on today's board meets that bar. Raising cash into the close and re-assessing at tomorrow's 09:45 open is the disciplined call. Sitting mostly in cash into the close: risk-off tape (QQQ -0.67%), fearful megathread sentiment, and no tradeable name clears the ≥2 cross-feed BUY-vote bar with fresh (non-spike-chronic) WSB momentum. COIN's Senate-vote catalyst is a binary event with zero-conviction split sentiment and no cross-feed backing — skipping per rule #19's late-cycle discipline. GOOGL remains held at cap, small and roughly flat, no action needed.

Horisont: 1-2 weeks Begrundelse: Fresh WSB DD thread on IREN's $50M/yr NA sports sponsorship deal (genuine new catalyst, not just crypto-mining momentum) plus cross-feed consensus BUY (analyst target +80% upside) clears the re-entry bar after yesterday's stop-out was actually a small winner (+4.1%). === Session-analyse === [wsb/US] WSB today is mostly macro doom (10Y hitting 5%, Warsh/Fed rate-hike fears, 'this selloff makes zero sense', Kenny G / Volcker jokes) rather than a clean meme-stock rocket thread. NVDA gets genuine multi-post bullish interpreted signal (net +0.62, 3 bull/1 bear, catalyst-heavy, disclosed positions) plus cross-feed corroboration would be needed but NVDA isn't showing consensus BUY in the screening block, so I'm passing on it — single-source WSB hype without 2+ cross-feed votes is exactly the loss-making pattern rule #9/CROSS-FEED gate warns against. MSFT does clear the bar (consensus BUY, wsb+2 hold votes, interpreted net +1.25 with disclosed 0DTE call profits being let ride) but it's on the UNTRADEABLE-at-min-size list for this account, so it's a no-go regardless of conviction. Cross-feed BUY consensus exists on IONQ, TSM, ASTS, ONDS, SOUN, RDDT, IREN, APLD, QBTS among others, but none of them are in today's TRENDING or MENTION-LEADERS lists — meaning WSB itself isn't currently talking about them, just background analyst/short-interest votes. Buying purely on cross-feed consensus with zero WSB chatter contradicts this book's entire mandate (WSB sentiment is the primary source). IREN is the one small exception — it's mentioned in trending (score 8, low volume) AND has a fresh DD-style post about a $50M/year sponsorship deal AND cross-feed consensus BUY — but I just ate a stop-out on IREN yesterday (-15.8%... wait, that was NBIS; IREN was +4.1%), and per rule #8 re-entry requires conviction ≥0.7 AND ≥2 fresh cross-feed BUY votes citing what's changed. The IREN sponsorship post is genuinely new information (a real business catalyst, not just momentum), cross-feed shows 2 BUY votes, and crypto_mining theme, while a net loser for me historically (-1590 DKK/17 exits), IREN individually has been a net winner (+242 DKK/2 exits) — so I'll take a small, disciplined re-entry with a tight stop given the theme's poor track record and existing GOOGL exposure already at cap. COIN is accelerating in mentions (+2 vs yesterday, DD post 'Fading Clarity Act') but gapped down -6% overnight and cross-feed doesn't show in the BUY list — that's a risk flag, not a buy signal; the crypto_exchange theme tape is actually +11%/5d which is a divergence worth noting but I don't have corroboration to act on a falling knife. Given QQQ/SPY both slightly red (neutral-to-soft regime), VIX calm at 15.84 but Fear&Greed at 69 (Greed) which is more a headwind for chasing crypto-adjacent names, I'm keeping this small: one modest IREN add, hold GOOGL (already at cap, small +2.9% winner, no reason to touch), and otherwise sit mostly in cash. No forced activity — WSB's hot list today is dominated by macro fear threads, not a fresh meme-stock catalyst worth chasing. GOOGL is a small +2.9% winner at the theme cap (internet/ai_infra/mega_cap all at 19%) — no room to add and no reason to sell a name WSB isn't actively hating on and where overnight gap was only -0.6%. Beyond the single IREN re-entry, I'm avoiding the crowd of cross-feed-only BUY names (IONQ, TSM, ASTS, ONDS, SOUN, QBTS, RDDT) because none of them show up in today's actual WSB trending/mention data — that's cross-feed noise without the WSB conviction this book is built to follow, and chasing it would just be borrowing tomorrow's fee drag for no edge (rule #17). NVDA has genuine WSB bull sentiment but lacks cross-feed corroboration (no BUY consensus shown) so the single-source gate blocks it. Macro backdrop (10Y at 5%, hawkish Fed fear, Warsh rate-hike thread top of hot list) argues for caution — QQQ/SPY both red today, Fear&Greed at 69 (Greed) is a chase-risk flag not a buy signal. Sitting mostly in cash (staying ~92% cash after the IREN add) is the correct read of a session where WSB is arguing about macro doom, not rallying behind a specific rocket.

SOFT STOP TRIGGERED: NVDA faldt under 7.5% stop (225.41 → 209.27). Auto-cut før agent-call.

SOFT STOP TRIGGERED: NBIS faldt under 10% stop (248.59 → 209.53). Auto-cut før agent-call.