An edge is not static. It decays. Market structure shifts, participant behavior evolves, and information dissemination accelerates. Understanding expectancy decay is crucial for long-term profitability. Traders must quantify this decay and adapt.
Quantifying Expectancy Decay
Expectancy (E) is the average profit or loss per trade. E = (Probability of Win * Average Win) - (Probability of Loss * Average Loss)
Expectancy decay manifests as a reduction in E over time. This reduction can stem from:
- Decreased Probability of Win (Pw): The percentage of winning trades falls.
- Decreased Average Win (Aw): The average profit per winning trade shrinks.
- Increased Probability of Loss (Pl): The percentage of losing trades rises.
- Increased Average Loss (Al): The average loss per losing trade expands.
Often, multiple factors contribute simultaneously.
Sources of Decay
Market Microstructure Changes
High-frequency trading (HFT) algorithms continuously optimize. A strategy exploiting latency or specific order book imbalances may see its edge diminish as HFT firms adapt. For example, a strategy identifying large block orders on a Level 2 feed might see its fill rate decline as HFTs front-run or obfuscate order flow.
Increased Competition
As more traders adopt a profitable strategy, the edge diffuses. Arbitrage opportunities close faster. Statistical anomalies normalize. If 100 traders exploit a specific pattern, the profit per trader is higher than if 10,000 traders exploit the same pattern. Bid-ask spreads tighten, and available liquidity at desired price points decreases.
Information Efficiency
Markets become more efficient over time. New information integrates into prices faster. A strategy based on delayed news feeds or slow-moving indicators loses efficacy as information processing speeds increase.
Regulatory Changes
New regulations can alter market dynamics. For example, changes in margin requirements or tick sizes can impact strategy profitability. A strategy reliant on specific options spread mechanics might be affected by new exchange rules.
Calculating Decay Rate
To quantify decay, compare expectancy over different time periods.
Let E1 be the expectancy for Period 1 (e.g., Q1). Let E2 be the expectancy for Period 2 (e.g., Q2).
Decay Rate (DR) = (E1 - E2) / E1
A positive DR indicates decay. A negative DR indicates an improving edge.
Example: Futures Scalping Strategy
Consider a futures scalping strategy on the ES contract. Period 1 (Q1, 2023): 1,000 trades executed.
- Probability of Win (Pw1): 60%
- Average Win (Aw1): 2.5 ES points ($125 per contract)
- Probability of Loss (Pl1): 40%
- Average Loss (Al1): 1.5 ES points ($75 per contract)
E1 = (0.60 * $125) - (0.40 * $75) E1 = $75 - $30 E1 = $45 per contract per trade
Period 2 (Q2, 2023): 1,000 trades executed.
- Probability of Win (Pw2): 55% (decreased)
- Average Win (Aw2): 2.2 ES points ($110 per contract) (decreased)
- Probability of Loss (Pl2): 45% (increased)
- Average Loss (Al2): 1.8 ES points ($90 per contract) (increased)
E2 = (0.55 * $110) - (0.45 * $90) E2 = $60.50 - $40.50 E2 = $20 per contract per trade
Decay Rate (DR) = ($45 - $20) / $45 DR = $25 / $45 DR = 0.5556 or 55.56%
The strategy's expectancy decayed by 55.56% in one quarter. This significant decay demands immediate attention.
Identifying Decay Triggers
Traders must monitor key performance indicators (KPIs) for early signs of decay.
- Win Rate (Pw): A sustained drop in Pw suggests the strategy's entry criteria are less effective.
- Average Win (Aw): A shrinking Aw indicates price targets are harder to reach or exit conditions are less optimal.
- Average Loss (Al): An expanding Al means stop-loss orders are hit more frequently or price moves against the position more aggressively.
- Profit Factor (PF): PF = (Gross Wins) / (Gross Losses). A declining PF indicates decay.
- Maximum Adverse Excursion (MAE): An increasing MAE suggests trades are moving further against the entry before reversing or hitting a stop.
- Maximum Favorable Excursion (MFE): A decreasing MFE indicates trades are not reaching their full potential.
Adapting to Decay
When decay is identified, adaptation is mandatory.
Strategy Refinement
- Entry Criteria: Adjust parameters. For a mean-reversion strategy, widen the deviation threshold. For a breakout strategy, increase volume confirmation requirements.
- Exit Criteria: Modify profit targets or stop-loss placements. If Aw is decreasing, reduce profit targets to improve Pw. If Al is increasing, tighten stop losses.
- Timeframes: Shift to different timeframes. A strategy effective on a 5-minute chart might perform better on a 15-minute chart if market noise increases.
- Instruments: Apply the strategy to different assets. If a stock strategy decays, test it on a related ETF or a different sector.
Position Sizing Adjustment
If expectancy decays, reduce position size to manage risk. Optimal F (Kelly Criterion) is sensitive to expectancy. Kelly % = W - [(1-W)/R] Where W = Win Rate, R = Average Win / Average Loss
If W decreases and R decreases, Kelly % will decrease, advocating for smaller position sizes.
Example: Options Spreads and Position Sizing
Consider an options credit spread strategy on SPY. Initial Strategy (Month 1):
- Win Rate (W1): 70%
- Average Win (Aw1): $0.40 per share ($40 per contract)
- Average Loss (Al1): $0.60 per share ($60 per contract)
- R1 = Aw1 / Al1 = $40 / $60 = 0.6667
E1 = (0.70 * $40) - (0.30 * $60) = $28 - $18 = $10 per contract
Kelly %1 = 0.70 - [(1-0.70)/0.6667] = 0.70 - (0.30/0.6667) = 0.70 - 0.45 = 0.25 or 25% of trading capital. If trading capital is $100,000, Kelly suggests risking $25,000. Each contract requires $100 margin, so 250 contracts.
Decayed Strategy (Month 2):
- Win Rate (W2): 60% (decreased)
- Average Win (Aw2): $0.35 per share ($35 per contract) (decreased)
- Average Loss (Al2): $0.65 per share ($65 per contract) (increased)
- R2 = Aw2 / Al2 = $35 / $65 = 0.5385
E2 = (0.60 * $35) - (0.40 * $65) = $21 - $26 = -$5 per contract
The strategy now has negative expectancy. Kelly %2 = 0.60 - [(1-0.60)/0.5385] = 0.60 - (0.40/0.5385) = 0.60 - 0.7428 = -0.1428 or -14.28%. A negative Kelly percentage indicates the strategy is not profitable and should not be traded. If forced to trade, the optimal position size is zero.
Proactive Monitoring
Continuous monitoring prevents catastrophic losses from decayed edges.
- Statistical Significance: Do not react to every minor fluctuation. Use statistical tests (e.g., t-tests) to determine if changes in Pw, Aw, or Al are statistically significant or merely random variance.
- Rolling Expectancy: Calculate expectancy over rolling periods (e.g., last 50 trades, last 100 trades) to identify trends.
- Market Regime Analysis: Categorize market conditions (e.g., trending, range-bound, high volatility, low volatility). An edge might decay in one regime but remain robust
