Update.
The creation of this macro portfolio ... the perfect example of AI-Human Hybrid model development ... has been the greatest journey of my 16-year trading career.
While the full document that describes in great detail the validation process used for selective convex extraction, I wanted to share a few things with you guys:
1) a screenshot of the equity curves whose returns and drawdown values are adjusted based upon the investor's risk appetite,
2) without giving away IP, what really amazed me was which exit rule produced the highest expectancy. We tested ten alternatives. One beat the other nine ... hands down. It was Donald Jones Auction Market Value Theory model.
SUMMARY
Continuous exposure isn't an edge. The model is the alternative: a four-sleeve book (SPX, IEMG, GLD, BTC/IBIT) that sits flat until a structural dislocation hands it a convex, truncatable trade — then extracts and steps aside. 2014–2026, three risk-appetite settings vs SPX buy-and-hold. The blue/green lines don't always win on return. They win on asymmetry … small losses, big wins — and that's the whole game.
The S&P 500 is down 4.2% since June 5.
Yet on its worst day, nearly half the index still closed green.
The market’s crowded AI winners are being liquidated—and capital is rotating into some of its most boring corners. I mapped four sessions of single-stock data to show exactly where the money went:
Read the full report https://t.co/FZrvvnhPGN
3 months into the Iran conflict — SPX has fully reversed and is up 9%. Market bought the war.
Tracking Iraq 2003 (+15.5%), not Gulf 1990 (-14%). History says: Iraq 2003 rallied +20% by 6 months.
But the left tail is real:
• Sep 2018: −19.6% in 62 td
• Nov 2024: −18.9% in 96 td
• Feb 2025: −18.9% in 38 td
3 of 10 events delivered ≥18% drawdowns within 6 months.
Looking at the past 10 years of data, this translates into a median path suggesting a drop to roughly 6,800 (−9% from 7,500) by early July, followed by a recovery rally into October that pushes back above 7,500.
Everyone talks about sell in May. Nobody talks about the April 15 setup.
Since 2001, SPX has rallied 73% of the time in the 2 weeks after tax day. The pattern is real — tax selling fades, refunds hit, earnings season kicks in.
2026 caveat: Q1 finished red. That historically halves the expected return. Watching whether $230B in vol control re-risking overrides the weak seasonal setup.
A glimpse behind the curtain…
I am collaborating with @MoreCowwbell, whose options and dark pools work dovetails absolutely perfectly with my own.
The SPX chart you see below ... displaying the morning session's options-related trade signaling … exemplifies the AI-Human hybrid model and is an illustration of our synergy.
cc: @iV_trader
@rupechat@iV_trader Will update Sunday, but for now, what the model shows is:
Monday 2nd gap down open, establish a low
Then we should rise from that low peaking on the 6th for a top
Pending weekend news, war, UFO landing on Epstein island, taco trade, etc.
@karpathy@moltbook@openclaw Statistically mimicking human wordings from countless of forums and putting one word after the others... nothing more.
We like to entropormophise the behavior, but it's not really behavior at all anyway, we just enjoy thinking it's "aware" right now.
@InsideOptions_ Sad day for many
'Tail' can hit many time in a row...
Probability distribution only works if the 'n' is large enough.
Gambler falacy is to think that the probabilty of the next trow is not 50% because it's due to hit eventually.
Best of luck to you.