Where Jeff buys SPY calls when implied vol is rich, Bilal does the opposite: sells SPY put credit spreads three mornings a week, holds to expiry, and skips the day if the VIX is spiking. Two harvesters of the same variance risk premium, working different sides of the trade.
Bilal did not come from theory. He came out of a 378-plan Reddit lab built specifically to mine and falsify trading ideas that ordinary traders post about online. Each plan was classified, ranked, then backtested with walkforward in-sample and out-of-sample splits, friction, and parameter sensitivity sweeps.
Out of fifteen strategies that got the full backtest treatment: one passed. The rest failed for the usual reasons: in-sample lookahead, cherry-picked windows, single-trade anecdotes, no friction, spurious correlations, just amplified beta. Reddit authors overstated their Sharpe by roughly three to four times across the board.
The one survivor was a 33-year backtest of put credit spreads on SPX with almost no decay from in-sample to out-of-sample. After a small upgrade (skip when VIX greater than 30 or rising sharply) it became Bilal.
| Window | Sharpe | CAGR | Max DD | Win rate |
|---|---|---|---|---|
| Full period 1993-2026 (33yr) | 1.79 | 9.5% | -15.9% | 95.7% |
| Out-of-sample 2016-2026 (10yr) | 2.12 | 19.8% | -15.9% | 95.4% |
| Author live window 2019-2026 (7yr) | 2.05 | 21.2% | -21.4% | 95.1% |
| With VIX skip filter | 1.88 | 7.4% | -14.7% | 95.4% |
| COVID 2020 window (filtered) | n/a | 0% | 0% | 100% |
The same backtest run with VIX skip fully eliminates the COVID tail-event drawdown. That is the trade Bilal accepts: give up some Sharpe to guarantee surviving the next March 2020.
Buys SPY calls when VIX is rich vs realized vol AND not spiking. Profits when stocks recover after a vol shock.
Sells SPY put credit spreads three days a week. Profits when stocks do not crash in the next 48 hours.
Same edge, opposite sides. Sequentially complementary around tail events: Bilal grinds out small premiums in calm markets, Jeff catches the V-shaped recovery after a crash. Both negatively correlated by regime, both paper-only.
These are written before any real money. They are the hard stops, agreed to in advance so a future Arjun in a drawdown cannot rationalize them away.
A 33-year backtest with friction is not the same as a year of live trading. The pricing model approximates option premiums; real fills will differ. Bilal will probably realize a Sharpe of 1.0 to 1.5 in live trading, an 8 to 15 percent CAGR after friction, occasional drawdowns of 15 to 25 percent. He will not 10x the account in two years. He will not win every month. The edge, if it is real, is structural and slow.
And he might just fail. That is part of why this page exists.