Signal was wrong — that's the dominant factor. The long thesis on the LUMI-AI news failed outright: AMD never rose meaningfully after entry (best unrealized just +0.10%) and dropped as much as -4.89%, so the predicted overnight drift did not occur. Timing compounded it modestly — the fill at 471.21 was already ~1.2% above the prior close (vs. +0.54% at decision time), meaning some news pop was paid up for before entry. The mechanical next-close exit at -2.41% actually landed well off the worst point (-4.89%), so the exit didn't leave upside on the table — there was essentially none to capture.

Signal: wrong. The short bet on continued downside failed — AAPL rose ~2.6% against the position, and the best unrealized gain while open was only +0.35%, so there was essentially no favorable move to capture. Timing: the stock was already down 1.16% from prior close at decision time, so the news-driven decline had largely occurred before entry — the trade was chasing a move that had happened. Exit: the -2.6% realized loss sits between the best (+0.35%) and worst (-3.48%) unrealized, so the mechanical next-close exit neither compounded nor salvaged much; the dominant factor is simply a wrong directional signal (shorting after the drop, into a rebound), not exit execution.

Signal: Partially right — the short saw up to +2.73% unrealized, so downside did materialize after entry; but the close-to-close outcome was against it. Timing: The stock was already down 1.17% from prior close at decision time, so some of the news-driven move was consumed before entry — though the subsequent +2.73% drawdown shows meaningful downside remained. Exit: This is the dominant factor: the mechanical next-close exit gave back a +2.73% unrealized gain and locked in -0.92% (-$5.28), capturing roughly none of the available move. The data doesn't show intraday path timing, so whether the reversal was gradual or late-day is not settleable here.

Signal/timing: The predicted post-news drift never materialized — the trade only ever showed +0.18% unrealized, so the direction call was wrong for the holding window. Notably, entry at 220.49 was already ~1.35% above the prior close (217.55), meaning much of the news pop was priced in before the fill; the trade bought the announcement spike and rode the fade.

Exit: The exit (-1.49%) landed between the best (+0.18%) and worst (-2.31%) unrealized, so exit execution neither gave away meaningful gains (there were essentially none) nor caught the bottom — it's not the driver here.

Dominant factor: Timing — entering after the move had already happened left no drift to capture, and the position simply mean-reverted.

Signal: right. The short direction played out — GOOGL fell from entry 339.78 to 334.83 (-1.46%), consistent with the competitive-threat thesis, though whether the news itself drove the move can't be verified from this data.

Timing: partially late. At decision time the stock was only -0.57% from the prior close (346.59), but the fill at 339.78 implies roughly -2% had already occurred by entry — a meaningful chunk of downside was gone before the position opened.

Exit: good capture. The +9.90 P&L (+1.46%) took most of the +1.91% best unrealized, and the position survived a large -3.89% adverse excursion to end positive; the data doesn't show the sequence of those extremes, so whether the exit was skill or a lucky close-time can't be settled here. Dominant factor: correct signal with efficient exit, undermined somewhat by a late fill.

Signal: wrong. The predicted overnight drift never materialized — the stock managed only +0.21% at its best before sliding steadily to -3.2%, and the exit at 180.17 landed essentially at the worst unrealized level, so the fixed next-close exit captured nearly the full loss (though there was almost nothing positive available to capture). Timing: the entry data (-1.40% since prev close, flat last 30m) shows the earnings news was already public and the stock was fading, suggesting the beat was priced in, but the data here doesn't show how much of any post-earnings pop had already occurred and reversed. The dominant factor is a bad signal: betting on continued drift after a widely-reported earnings beat, when the move was over or reversing.

Signal: Wrong. The predicted overnight upside drift didn't materialize — the stock closed slightly below entry, and the excursion profile (+1.22% best vs -6.07% worst) shows the position spent most of its life deep underwater. Timing: The data doesn't settle this — the modest -0.59% pre-entry move doesn't tell us when the earnings hit the tape or whether the pop had already occurred earlier. Exit: The mechanical next-close exit actually performed well relative to what was available, landing at -0.19% versus a -6.07% worst case, though it also missed the brief +1.22% peak.

Dominant factor: signal failure — the bullish earnings thesis simply didn't produce the expected drift, and the trade was mostly saved by the exit landing near the top of a bad range.

Signal (dominant factor): wrong. The predicted overnight upside barely materialized — the trade never got better than +0.21% and sank to -2.24% while open, so the bullish earnings read simply didn't play out in price.

Timing: partially at fault — decision-time context showed the stock roughly flat (-0.08% vs prev close 368.79), but the fill at 370.62 was ~0.5% above prev close, meaning some of any favorable move was consumed between decision and fill.

Exit: actually fine — the -0.31% realized loss (-$2.30) is close to the +0.21% best available and far better than the -2.24% worst, so the mechanical next-close exit recovered most of the drawdown; the loss traces to the signal, not the exit.

Signal: Right, but weakly — the bullish call produced a small gain, and price at entry was below the prior close (-0.88%), so the anticipated move had not already run before entry; whether the news itself was truly fresh can't be verified from this data.

Exit: This is the dominant factor. The trade saw +0.77% unrealized (~$5.15 on 2 shares) but exited at +0.14% (+$0.92), capturing under a fifth of what was available — the position gave back most of its gain before the exit fired.

Bottom line: Direction and timing were acceptable; the exit left the bulk of the available move on the table.

Signal (dominant factor): The bearish call was simply wrong — AAPL rose ~2.7% against the short, and the trade never had more than +0.23% of favorable movement available.

Timing: Price context at entry (-0.05% since prior close, +0.10% last 30m) shows no prior downward move, so this wasn't a case of chasing a completed drop; the predicted decline just never occurred.

Exit: The exit captured most of what was salvageable — closing at roughly -2.7% versus a worst unrealized of -3.61%, and there was essentially nothing (+0.23% max) to have captured on the winning side. The loss is attributable to a bad directional signal, not exit execution.

Timing is the dominant factor. MSFT was already down ~1% from prior close (and the fill at 498.67 was ~1.7% below prev close of 507.29), so the bearish repricing the rationale predicted had largely occurred before entry; after entry the stock only went up, and the position was never profitable (best unrealized -0.06%). The exit was reasonable given what was available — the -0.42% realized loss avoided the -1.42% worst — there was simply nothing to capture. Whether the vulnerability news genuinely caused the pre-entry decline (i.e., whether the signal itself was right) can't be settled from this data; what's clear is that post-entry, the predicted direction did not play out.

Signal: Directionally right — price drifted lower after entry and the trade closed green — though the rationale itself is about SpaceX (a different Musk company), so whether the thesis actually drove TSLA's move isn't settled by this data. Timing: Mostly missed — TSLA was already down ~1.5% from prev close at decision time and the fill at 357.49 sits ~2.8% below prev close, so a large chunk of the bearish move preceded entry. Exit: Weak capture — +$2.70 (~0.38%) against a best unrealized of +1.26% means roughly two-thirds of the available move was left on the table; the dominant factor here is the exit, which surrendered most of what the (late) entry still had available.

Signal: Right direction — price rose steadily after entry (worst unrealized only -0.09%), though whether the blocked-acquisition news actually drove the move can't be determined from this data. Timing: Good — entry at 567.51 was below prev close after a -0.87% pullback, so the move had not already run; the trade captured essentially the entire favorable leg from entry. Exit: The +10.20 (+1.80%) realized is decent but left meaningful upside — best unrealized was +2.97%, so roughly 40% of the peak move was given back or missed.

Dominant factor: a correct, well-timed signal; the only shortfall was the exit capturing ~60% of the available move.

Signal: During the open window the short was barely ever in the money (best +0.19% vs worst -1.03%), so the bearish call didn't play out over the holding period — though the window was cut short by netting, so the longer-term thesis is unsettled here. Timing: The entry looks late — NVDA was already down 1.17% from prior close and rebounding (+0.89% in the last 30m), meaning the trade shorted into a bounce after the move. Exit: The exit at -0.64% (-$5.56) sat between the best (+0.19%) and worst (-1.03%) available, but it was mechanical — closed by being outranked by an opposing long decision, not by any read on the trade itself. Dominant factor: bad timing (shorting after the decline, into a recovery), compounded by a forced early exit that made the signal untestable.

Signal was wrong (dominant factor). The bullish repricing call never materialized during the holding window: the trade's best unrealized was -0.03%, meaning the position was essentially never in profit — MU only declined after entry. Timing isn't the issue in the "chased the move" sense (price was already down -0.74%/-1.53% at entry, so nothing bullish had happened yet); the predicted move simply never came. The exit at -1.70% (-16.20 USD) was between the best (-0.03%) and worst (-2.93%) unrealized, so it neither salvaged the breakeven opportunity nor rode to the bottom — but with no upside ever available, no exit rule could have made this a winner. Whether the bullish thesis plays out on a longer horizon isn't determinable from this data.

Signal was the failure. The bullish call never worked: the stock's best unrealized while open was still -0.17% (never above entry), so the predicted upside simply didn't materialize. Timing wasn't the issue — price was only down 0.43% from prior close at decision time with a flat last 30 minutes, so there's no sign the move had already run before entry. The exit was mediocre but secondary: realized P&L of -1.03% sits between the best (-0.17%) and worst (-1.38%) available, meaning most of the drawdown was absorbed, but with the trade never positive there was little to capture regardless.

Signal: Directionally right but barely — price drifted +0.30% from entry, a negligible move for a "record profitability" catalyst. Timing: The stock was flat-to-down at entry (-0.27% since prior close), consistent with the earnings news already being absorbed, though the data here doesn't show when the article broke or whether a larger move preceded it — that part is unsettled. Exit: Nothing to complain about — exit captured +0.30% against a best unrealized of +0.36% and worst of -0.43%, essentially the top of the range available.

Dominant factor: The signal was stale-or-weak — a headline this strong produced almost no tradable move during the holding window, so the +$2.48 is noise-level P&L on a well-executed exit of a nearly dead trade.

Signal: Directionally right but barely — the stock moved only +0.12% from entry, and the range while open (+0.32%/-0.33%) was symmetric noise, so the "superior chip benchmarks" thesis produced no discernible bullish move in this window.

Timing: Not a chasing problem — price was slightly down (-0.36% since prior close) at decision time, so the move hadn't already happened by entry.

Exit: Captured +0.12% of a +0.32% best case, which is acceptable given how little was available. The dominant factor is signal weakness: the catalyst simply didn't move the stock during the holding period, and this data can't settle whether the move came later or never — only that it wasn't there while the trade was open.