AAPL/NVDA: 19 days short the spread AAPL-0.6737*NVDABATS:AAPL-0.6737*BATS:NVDARORO_Labs**The trade, explained.** We are short the spread on AAPL/NVDA. In plain terms: short Apple, long NVIDIA — $10,000 of AAPL (32 shares) against $4,253 of NVDA (22 shares) on the hedge ratio. This is not a bet on Apple falling or NVIDIA rising on their own. It's a bet on the *gap* between them: the spread stretched unusually far above its normal level, and the position profits if that gap closes back toward average. Because one leg is long and one is short, the trade is roughly market-neutral — the broad market can rise or fall and the position mostly cares about the relative move between the two names. It is 19 days in, marked at −$397.47 (−2.79%), with a +$243.62 day as the latest bar pulled back. **The frozen trade rules.** These were locked by our tracker — our research stack, PairSync Terminal — the moment the signal fired, and they don't change mid-trade: - **Entry:** 2.74σ — the stretch at which the pair fired. - **Target:** exit at 0.02σ — essentially all the way back to the average, about 2.7σ of travel from entry. - **Stop:** ±4.45σ — the line where the idea is declared wrong. - **Max hold:** a timeout rule closes the trade if it overstays; no open-ended waiting. **Where it stands.** The tracker reads 3.64σ — the spread has pushed wider since entry, against the position. That leaves roughly 0.8σ of room to the frozen stop, with about 3.62σ of travel still to target. Not a comfortable print, but a bounded one: the exit lines were drawn before the trade began, and the tracker still flags the position as on track, within its expected range. **Optimisation history (in-sample).** Before this signal fired, our optimiser tested these exact rules against this pair's own history. That's in-sample — the rules were fitted on this data — so treat it as indicative of how the pair trades under these rules, not as a promise: - **Average return:** +5.44% per trade - **Average days in position:** 12.3 - **Max drawdown:** 7.4% - **Win rate / trades:** 3 for 3 — a small sample, and we say so. Sharpe 1.80. The three trades that set these parameters: Nov 21 → Jan 2 (+2.66%, 27 days), Apr 27 → May 1 (+7.10%, 4 days), May 14 → May 22 (+6.58%, 6 days). One yardstick worth watching: 19 days held against a 12.3-day in-sample average — this one is running long, though the slowest optimisation trade took 27 days. **Spread diagnostics in plain English.** Our research stack's scorecard on this spread is genuinely mixed, and that's worth being honest about: - **Hurst 0.23** — the rubber-band score. Well below 0.5, which says the spread's moves tend to snap back rather than trend. Strong. - **ADF p 0.474** — the formal "does it snap back to its average?" test. This one does *not* pass at conventional levels. A red ADF next to a green Hurst is a real tension: statistical tests disagree sometimes, that's normal, and this trade leans on the Hurst-style behaviour plus the pair-specific optimisation. - **Half-life 29 days** — a typical stretch fades halfway back in about a month. This is a patient trade by construction. - **Pearson R 0.17** — day-to-day co-movement is loose; the relationship here is about the spread's behaviour, not tight daily lockstep. - **Current Z 1.90** — our stack's latest general read of the stretch. It differs from the 3.64σ tracker print because the tracker runs the pair-specific terms frozen at signal time; the diagnostics run a fresh general model. Two lenses, one spread. **The chart read.** The spread's price pane shows a sharp push wider through late June into July — against the short position — and then a red bar on the most recent session, the first partial give-back. Wider first, mean later: that is exactly the sequence this trade is built to sit through, within its stop. **The second opinion: POSI v3.** The "POSI v3 by PairSync" script on this chart is a deliberately lightweight, one-size-fits-all solver running on the fixed ratio, learning its own window (72 bars here). Its read: MEAN REVERTING: YES — the script thinks the gap snaps back. Hurst 0.42 — rubber-band behaviour, independently of our tracker. Reward Risk 2.6:1 with an 83% probability of hitting target before stop, net edge +2.4σ, reason: "Optimal reversion trade." Its z-line prints 0.94, not 3.64 — expected, not a contradiction: the pane is a generic on-chart solver with its own learned window, while the tracker runs the pair-specific terms frozen when the signal fired. The pane's own suggested terms — entry 1.7σ, stop 3.0σ, target 1.8σ of travel — are tighter than our frozen 2.74σ / ±4.45σ / ~2.7σ set. The persistent difference is scope, not quality: an on-chart Pine script can't do per-pair discovery and frozen-at-fire tracking; the frozen tracker, in turn, is deliberately locked and doesn't re-learn mid-trade. The verdict in plain words: the second opinion agrees — stretched, mean-reverting, and worth the wait. **The takeaway.** If you want to watch this yourself, chart AAPL-0.6737*NVDA. One caveat: hedge ratios drift. A ratio solved months ago goes stale — static spread symbols need periodic re-solving. **How this strategy works.** Our research stack scans thousands of pairs of economically related stocks — same sector, same customers, same economic forces. For each candidate it tests whether the spread genuinely mean-reverts, optimises per-pair trade rules (entry, target, stop, timeout) on the pair's own history, scores the result to rank the best opportunities, and tracks the strongest signals live — freezing the rules at the moment the signal fires and monitoring the position to its exit. The core idea: the relationship is statistically proven at the point in time of the signal, not assumed forever. Relationships drift, which is why everything is re-tested and re-solved run after run. Not financial advice. Do your own research.