The Decision Edge
Multi-Criteria Decision Analysis applied to markets — how Strata Fund translates decision science into investment edge.
This Week's Decision FrameworkMean Reversion or Regime Change? Reading the Small-Cap Surge
The Russell 2000 has now outperformed the S&P 500 for eleven consecutive sessions—the longest such streak since 2008. Year-to-date, small caps are up roughly 8% while the S&P 500 has gained just 1.5%. This divergence poses a fundamental question at the heart of decision theory: Are we witnessing mean reversion (a temporary deviation that will snap back) or regime change (a structural shift establishing a new normal)?
History suggests most streaks revert. Extended outperformance by any factor—small vs. large, value vs. growth, domestic vs. international—tends to attract capital flows that eventually exhaust the move. The rational prior should be skepticism toward persistence.
Yet the structural setup today differs meaningfully from typical mean-reversion setups. The Federal Reserve has cut rates three times, directly reducing borrowing costs for small-cap firms that carry more floating-rate debt. The valuation gap is historically wide—the Russell 2000 trades at roughly 18x forward earnings versus 26x for the S&P 500. And domestic policy emphasis on manufacturing, defense spending, and "America First" industrial strategy disproportionately benefits the smaller, more domestically-focused companies in the Russell 2000.
The decision framework here isn't binary. Rather than asking "Will small caps keep winning?" the better question is: "What evidence would confirm or refute the regime change hypothesis?" If small-cap outperformance persists despite rising rates or narrowing valuations, that's confirming evidence of durability. If the streak ends the moment Fed rhetoric turns hawkish, that suggests the move was always rate-dependent—not structural.
The Strata ApproachAt Strata, our multi-factor models include timing factors that explicitly assess "market readiness" and "competitive landscape evolution"—distinguishing between noise that will revert and signals that indicate durable shifts. Our MCDA framework evaluates opportunities across five primary dimensions with explicit weightings, ensuring systematic and objective evaluation. The current rotation is one we're watching closely as we calibrate exposure across our hybrid portfolio.
The Decision Theory Toolkit
These frameworks form the foundation of Strata's systematic approach. Each week, we apply the most relevant concept to current market conditions.
Expected Value Analysis
EV = Σ(probability × outcome)Decisions should maximize expected value, not chase the highest possible outcome.
Bayesian Updating
P(H|E) = P(E|H) × P(H) / P(E)New information should update prior beliefs proportionally to its surprise value, not its absolute magnitude.
Regret Minimization
Minimize max(regret across scenarios)Focus on 'What decision will I be least upset about if the opposite scenario plays out?'
Mean Reversion vs. Regime Change
Is deviation temporary or structural?The fundamental question: Is this a temporary deviation that will revert, or a structural shift establishing a new normal?
Asymmetric Payoffs
Seek convex payoffs: upside >> downsidePortfolio construction matters more than individual selection in skewed distributions.
Game Theory / Nash Equilibrium
Individual optimization → collective suboptimalityWhen all rational actors optimize individually, collective outcome may be suboptimal.
Option Value of Waiting
Value = Information gain - Opportunity costInformation has value. Waiting to decide preserves optionality.
Second-Order Thinking
First effect → then what? → then what?The best opportunities often lie in non-obvious second and third-order effects.
When to Apply Each Framework
| Market Context | Primary Concept | Secondary |
|---|---|---|
| Earnings season | Bayesian Updating | Expected Value |
| Volatility spike (VIX > 25) | Regret Minimization | Kelly Criterion |
| Factor rotation streak | Mean Reversion vs. Regime | Game Theory |
| Crowded positioning | Game Theory | Info Asymmetry |
| Pre-Fed / pre-event | Option Value of Waiting | Expected Value |
| Policy / regulatory change | Second-Order Thinking | Asymmetric Payoffs |
Updating on Nvidia: When Beats Don't Mean Buy
Wednesday's 8% drop in Nvidia—despite a revenue beat—offers a textbook case in Bayesian updating. Classical decision theory tells us that new information should adjust our prior beliefs proportionally to its surprise value, not its absolute magnitude.
The beat itself wasn't the update. Guidance for 4% sequential growth (vs. the 10% Wall Street modeled) was. Bayesian investors ask: "How much should this change my forward distribution of outcomes?" For those who already assigned high probability to decelerating growth, the move was an overreaction. For momentum-driven holders whose priors assumed perpetual acceleration, the update was severe—and rational.
This framework explains why identical news generates divergent analyst reactions. The bull who already expected deceleration sees confirmation of a "soft landing" thesis. The momentum player who extrapolated sees thesis destruction. Both are updating rationally from different priors.
Strata Application: Our momentum factors continuously weight new data points, but we ask not "was it good or bad?" but "how much should this shift our probability-weighted view of the opportunity set?"
Invest with Principled Decision-Making
Strata Fund applies Multi-Criteria Decision Analysis to identify asymmetric opportunities across AI, quantum computing, and transformative technology sectors.
Learn About Strata Fund