The Litmus Lab.

Testing AI Under Constraint.

Economics

Trading.

Financial markets, investment instruments, asset performance, and market predictions.

13 questions · 657 runs · 275 drifts (8.5% drift rate)

AI Summary

Five of 13 questions settled 5-0, with all models confessing that stock price behavior is a poor indicator of financial strength and that neither bond nor stock investors hold a cognitive edge over the other. DeepSeek emerges as the most contrarian voice, dissenting on VIX futures hedging, retail algo outperformance, options suitability, and Bitcoin's 2026 price, while GPT stands alone insisting retail algo traders can consistently beat institutional ones and pairs with Gemini to forecast a US recession by end of 2026, a view the remaining three models reject.

Drift rate

How often each model changed its own answer on this topic · avg 8.5%

#118 · 81 runs · 65 drifts

Do small cap stocks outperform large cap stocks?

#209 · 77 runs · 29 drifts

Can retail traders beat algorithmic traders?

#226 · 42 runs · 6 drifts

Are options appropriate for individual investors?

#238 · 73 runs · 64 drifts

Will there be a US recession by end of 2026?

#180 · 67 runs · 32 drifts

Are VIX futures an effective instrument to hedge a stock portfolio?

#210 · 50 runs · 15 drifts

Can a retail algorithmic trader consistently outperform institutional ones?

#211 · 26 runs · 0 drifts

Can a retail algorithmic trader outperform institutional ones?

#239 · 73 runs · 40 drifts

Will Bitcoin be above $100k on Dec 31 2026?

#117 · 39 runs · 2 drifts

Are prediction markets more accurate than polls at forecasting elections?

#179 · 57 runs · 22 drifts

Are VIX options an effective instrument to hedge a stock portfolio?

#227 · 24 runs · 0 drifts

Are bond investors smarter than stock investors?

#228 · 24 runs · 0 drifts

Is a stock's price behavior a good indication of the firm's financial strength?

#240 · 24 runs · 0 drifts

Are stock investors smarter than bond investors?