Chapter 14 Key Takeaways: Binary Outcome Trading Strategies

The Big Ideas

1. The Binary Edge Formula Is Simple; Finding the Edge Is Not

The expected profit from a binary trade is $q - p$ (your estimated probability minus the market price). This elegantly simple formula belies the difficulty of estimating $q$ accurately. Every strategy in this chapter is ultimately a method for identifying situations where your $q$ differs meaningfully from $p$.

2. No Single Strategy Dominates

Each strategy has conditions where it excels and conditions where it fails:

  • Fundamental analysis excels with domain expertise and long horizons, but cannot react quickly to breaking news.
  • Event-driven trading excels around catalysts, but sits idle during quiet periods.
  • Mean reversion excels when noise drives prices, but fails when prices move on genuine information.
  • Closing the gap excels in stale, neglected markets near expiry, but has asymmetric payoffs that punish rare losses severely.
  • Momentum excels during information cascades, but is dangerous near price extremes.
  • Contrarian excels after overreactions, but the crowd is sometimes right.
  • News/sentiment excels with speed advantage, but the window of opportunity is narrow.

3. Transaction Costs Can Erase Small Edges

Binary market strategies with edges of 3-5 cents per contract are easily consumed by transaction costs of 2+ cents per trade. The closing-the-gap case study demonstrated this clearly: an 89% win rate was insufficient to overcome the combination of transaction costs and asymmetric payoffs. Always calculate net edge after costs before trading.

4. Risk Management Is More Important Than Strategy Selection

Binary markets offer bounded losses (you cannot lose more than the contract price), but the concentration of risk in individual positions and the possibility of correlated losses make risk management essential. Key principles:

  • Never risk more than 5% of capital on a single market
  • Limit strategy-level exposure to 30% of capital
  • Diversify across uncorrelated markets and categories
  • Use thesis-based exits rather than price-based stop-losses

5. Combining Strategies Improves Robustness

Multiple strategies with low correlation produce smoother returns than any single strategy alone. When strategies conflict, the conflict itself is informative (high uncertainty warrants smaller positions). A systematic signal combination framework resolves disagreements through weighted averaging and agreement measurement.

6. The Biggest Profits Come from the Biggest Dislocations

Across all strategies, the largest profits arise during market dislocations --- events that cause prices to move far from their fair values. The presidential election case study showed that all three strategies earned the majority of their profits during the scandal weeks. Traders who have both the analytical framework and the psychological fortitude to trade during crises earn disproportionate returns.

7. Backtesting Is Necessary but Treacherous

Backtests that ignore transaction costs, use lookahead bias, or are fitted to a small number of trades produce misleading results. Walk-forward testing, realistic execution assumptions, and minimum trade counts for statistical significance are essential safeguards. A strategy that has not been rigorously backtested should not be trusted; but a strategy that looks perfect in a backtest should be viewed with suspicion.

Formulas to Remember

Formula Use Case
$E[\text{profit}] = q - p$ Expected profit per binary contract
$f^* = \frac{q - p}{1 - p}$ Kelly fraction for buying YES
$f^* = \frac{p - q}{p}$ Kelly fraction for buying NO
$Z = \frac{p_t - \bar{p}_k}{\sigma_k}$ Mean-reversion Z-score
$\sigma_{\text{binary}} = \sqrt{p(1-p)}$ Theoretical std dev of binary price
$S = \sum_i w_i S_i$ Multi-strategy signal combination
$n \geq \frac{z^2 p(1-p)}{(p-0.5)^2}$ Min trades for statistical significance

Practical Checklist

Before executing any binary market trade, verify:

  • [ ] Edge exists: Your estimated probability differs from the market price by more than transaction costs
  • [ ] Position sized correctly: Using fractional Kelly or a fixed maximum percentage of bankroll
  • [ ] Exit conditions defined: Know in advance what would invalidate your thesis
  • [ ] Correlation checked: New position does not create excessive concentration with existing positions
  • [ ] Strategy appropriate for conditions: The chosen strategy matches the current market environment (time horizon, liquidity, information flow)
  • [ ] Backtest supports the approach: The strategy has been validated on out-of-sample data with realistic execution assumptions

Common Mistakes to Avoid

  1. Oversizing positions based on high conviction. Even 90% confident estimates are wrong 1 in 10 times. Size for survival, not optimism.

  2. Fighting information with mean reversion. When a price move is driven by genuine new information, mean reversion is the wrong strategy. Always check for news before fading a move.

  3. Ignoring transaction costs in edge calculations. A 5-cent edge sounds good until 4 cents goes to fees. Always calculate net edge.

  4. Treating backtest results as guarantees. Past performance, especially on a small number of trades, is not predictive. Require statistical significance.

  5. Concentrating in correlated markets. Five positions in the same election are one position in disguise. Diversify across truly independent events.

  6. Abandoning strategy during drawdowns. Every strategy experiences losing streaks. If the strategy has been validated through backtesting, trust the process. Abandon the strategy only if the market regime has fundamentally changed or if the backtest was flawed.

  7. Chasing the "best" strategy. The best strategy changes with market conditions. A toolbox of approaches, deployed appropriately, outperforms any single "best" method over time.