The correlation regime between NVDA-TSLA entered a weakening trend on 02 Sep 12:00 ET at +0.26, and is still ongoing (15 candles · 60 trading hours). Over this period NVDA fell 0.96%, TSLA rose 3.07%. The two pairs decoupled.
The composite correlation has since downed from +0.26 (02 Sep) to -0.195 now (last data: 09 Sep 08:00 ET) the weaking the regime flagged, playing out.
READING — NVDA–TSLA at WEAK × Weakening (this pair's own last 6): for this pair it has bee a moderate decoupling (hedge / spread) — 6 past runs, median 8 candles · 32 trading hours (range 5–18), 1 lasted 2+ days. So the play has a couple of days of room before the turn (ANTITHESIS).
GAME PLAN: Correlation is in a Weakening regime — the bond between NVDA-TSLA is loosening. This signals a shift into a decoupling regime: hedge and pair-trade (spread) strategies come to the fore, while one-way beta bets grow riskier.
ANTITHESIS: Risk: decoupling may be temporary; if correlation snaps back from the bottom, pair-trades backfire. Correlation is currently 0.41 points below the period (1 Month) average — decoupling continues.
The divergence has been TSLA outperforming while NVDA stalled — TSLA's move favored calls.
Methodology: Four-hour candle closes are used while the markets for both instruments are open (08:00, 12:00, 16:00, 20:00 ET).
The correlation coefficient is a composite score derived not only from the Pearson method but also from Pearson, Spearman, and EWMA calculations.
https://i.redd.it/gdxf4ra56ioh1.png
Posted by arslanefe
1 Comment
The chart shows the decoupling pretty clearly, TSLA grinding up while NVDA just sits there. The spread play makes sense for now but 0.41 below the monthly average isnt exactly screaming edge, more like a nudge. Would be curious to see what happens if NVDA catches a bid while TSLA cools off, that snapback risk is the whole ballgame