Home/Systematic Trading/Regime-Adaptive Sizing

Regime-Adaptive Sizing

regime-detectionLevel 3 — Advanced

What It Is

The practice of scaling position sizes and gross exposure differently across regime states — reducing exposure in ambiguous or high-risk regimes, scaling up in high-conviction regime states, and applying a constant-premium or risk-parity approach in intermediate states — rather than keeping constant sizing regardless of market environment.

Correct Execution

Practitioner defines explicit sizing rules for each regime state before implementation — not ad hoc during live trading. Core principle: use regime state to size risk, not to time entries and exits. In high-conviction trend regimes, full exposure. In ambiguous/transitioning states, reduced exposure. In high-vol/cascading regimes, minimum exposure or defensive positioning. The sizing is mechanical and rule-based; override by discretion is prohibited. For portfolio-level sizing, constant-vol targeting serves as a regime-adaptive baseline that automatically reduces size when vol rises.

Progression Levels

Diagnostic Tree

Coaching Cues

  • "Regime is for sizing, not timing. Know the difference." — Rodrigo Gordillo / Return Stacking framework
  • "In high vol with steep skew, put protection is already baked in by the crowd. That's when you buy calls." — Corey Hoffstein, Investment Magazine, 2021-07-10
  • "The constant-premium approach: it's not about the notional. It's about the spend." — Corey Hoffstein, 2021-07-10
  • "Half the portfolio should do well in increasingly bullish markets. Half should do well in increasingly bearish markets. That's it." — Corey Hoffstein, barbell framework

Common Errors

  1. Using regime for trade timing rather than risk sizing: Regime state should drive how much risk you take, not precisely when you enter or exit. Trying to time entries adds a second source of timing luck on top of regime detection lag.
  2. Ignoring the skew regime when sizing convexity: When skew is steep (puts very expensive relative to calls), that signals the market has already partially priced the downside. Constant-premium spend in that environment should weight calls over puts.
  3. Forgetting to size the defensive leg as aggressively as the aggressive leg: A barbell requires both ends to be real. A half-barbell (aggressive upside, token downside protection) is just a long portfolio with decoration.
  4. Not pre-defining regime-specific sizing rules: Making sizing decisions during regime transitions in real time introduces discretion at the worst possible moment. All sizing rules must be defined in advance.

Edges

Conventional Wisdom Is Wrong

Regime Should Drive Sizing, Not Entry/Exit Timing

Most regime filters are used as binary on/off switches — in regime, fully invested; out of regime, in cash. This creates a second-order timing problem on top of the original regime detection lag. The correct use of regime information is to scale position size (full, half, minimal) rather than to time entry/exit precisely. A position that is half-size during regime uncertainty captures half the opportunity while limiting drawdown from a wrong classification.

What most people do
Use regime filters to decide when to enter and exit positions entirely. Treat a "regime change" as a trigger for full portfolio repositioning.
What the best do
Define three exposure states (full, half, near-zero) tied to regime confidence, not binary signals. The half-exposure state is the perpetual holding state during ambiguous transitions. Only the high-conviction states (clear trend or clear risk-off cascade) trigger full or near-zero exposure.
Why it's an edge: Eliminates the catastrophic timing mistakes from full on/off regime switching while preserving the risk-reduction benefit of the classification.
How to exploit: Redesign any binary on/off regime filter as a 3-state system: (1) clear regime (>70% confidence): full exposure; (2) ambiguous/transitioning: 50% exposure; (3) high-cascade-risk composite: near-zero equity. Never jump directly from full to zero.
Rodrigo Gordillo & Corey Hoffstein, Return Stacking podcast, 2021-11-15
💎 Elite-Only Behavior

When Puts Are Expensive, Buy Calls Instead

When implied skew is steep and put options are expensive (relative to calls), the crowd has already partially priced the downside. The conventional response is to buy puts for protection — but this is the highest-cost protection at exactly the moment when protection is least value for money. The contrarian response: steep skew in a constant-premium spend framework automatically produces more calls than puts, creating an asymmetric long bias. When everyone is protected, the asymmetric risk is to the upside.

What most people do
Spend more on protection when they are most worried — buying puts when VIX is elevated and skew is steep. This is maximum cost for maximum protection at exactly the wrong moment.
What the best do
Fix the dollar amount spent on convexity (constant-premium spend), not the notional. When puts are expensive (high VIX/skew), the fixed spend buys less put notional and more call notional — automatically creating an upside asymmetry when the crowd is already defensively positioned.
Why it's an edge: Anti-cyclical positioning: most protection when puts are cheap (normal markets) and most upside exposure when puts are expensive (post-crisis, wall of worry). This is structurally superior to fixed-notional hedging which is pro-cyclical.
How to exploit: Replace any fixed-notional hedging program with a constant-premium spend. Set the monthly dollar budget for convexity. Allocate between puts and calls based purely on which side of the vol surface is cheaper. Track the notional size of each over time to observe the anti-cyclical behavior.
Corey Hoffstein, "Liquidity Cascades," Investment Magazine, 2021-07-10
🔑 Hidden Causal Lever

Constant-Vol Targeting Is A Lagging Regime Adapter That Needs A Leading Layer

Vol-targeting (scaling position size inversely to realized vol) is widely used as a regime-adaptive mechanism. But realized vol is a lagging indicator — the strategy reduces size after the first vol spike has already done damage. In cascade events, the first vol spike is the largest and most damaging before vol-targeting even responds. A positioning-based leading indicator must be layered in front of the vol-targeting mechanism to provide genuine pre-emptive adaptation.

What most people do
Implement vol-targeting and believe they have solved the regime-adaptive sizing problem. Experience that the first large spike still damages the portfolio before the size reduction kicks in.
What the best do
Use vol-targeting as the primary sizing mechanism, but add a positioning overlay that pre-emptively reduces size when systematic participant leverage is at extremes — before any vol spike occurs.
Why it's an edge: Correctly characterizes vol-targeting as a reactive (lagging) tool and adds the missing leading layer, providing genuine pre-emptive risk reduction in the scenario where it matters most.
How to exploit: For every vol-targeting strategy, measure the average time from cascade onset to when the vol-targeting mechanism has fully reduced size. If this is >5 days, the vol-targeting response is insufficient for sudden cascades. Add a composite positioning indicator that triggers a 30-50% size pre-reduction when extreme positioning is detected.
Corey Hoffstein, Investment Magazine, 2021-07-10

Sources

  • Corey Hoffstein, "Liquidity Cascades," Investment Magazine, 2021-07-10 — barbell architecture, constant-premium spend, skew-driven convexity allocation
  • Rodrigo Gordillo & Corey Hoffstein, Return Stacking podcast, 2021-11-15 — three-piston motor framework, return stacking with different exposure profiles
  • "Financial Advisors: Immunize Business Risk from Bear Markets & Inflation Regimes," YouTube, 2023-11-07 — equal-weight three-piston motor (equities/bonds/managed futures), 1/3 allocation framework for different client risk profiles