When Predictive Trading Strategies Are Not Enough: A Practical Framework for Evaluating Economic Viability
Many traders, investors, and financial decision-makers are drawn to strategies that appear to predict short-term market movement. A trading rule, indicator, model, or signal may seem promising when it shows favorable historical performance, strong forecast accuracy, or attractive backtested returns. However, predictive ability alone does not mean that a strategy is economically viable.
A strategy can look profitable on paper and still fail when real-world trading conditions are considered. Transaction costs, bid-ask spreads, slippage, liquidity limits, execution delays, changing volatility regimes, turnover, taxes, and downside risk can materially reduce or eliminate the value that appeared in a backtest. For short-horizon strategies, this issue is especially important because expected gains per trade may be small and repeated trading can magnify even modest implementation costs.
My recent doctoral research examined this issue through a rapid evidence assessment of empirical studies involving short-horizon U.S. equity and broad equity-index strategies. The purpose was to better understand when predictive trading strategies retain favorable net, risk-adjusted performance after realistic validation, transaction costs, liquidity constraints, execution conditions, and volatility-regime effects are considered.
The central finding was clear: predictive signals can add value, but their economic usefulness is conditional.
A strategy should not be evaluated only by whether it predicts price movement or produces a positive gross return. Instead, financial decision-makers should ask whether the strategy remains credible, implementable, and economically meaningful after a more complete review.
Why Prediction Is Not the Same as Viability
A trading strategy may appear successful because it identifies a pattern in historical data. That pattern may be connected to investor behavior, market sentiment, temporary mispricing, momentum, reversal, volatility, order flow, or other market conditions. Behavioral finance helps explain why these short-term patterns may occur.
However, a pattern that exists in historical data does not automatically become a profitable trading opportunity. The strategy must still survive the conditions of actual use.
For example, a short-term signal may disappear after trading costs are deducted. A model may work in one volatility regime but fail in another. A strategy may require trading volume that exceeds realistic liquidity. A backtest may look strong because it was overfit to a specific sample. A strategy may generate attractive raw returns but expose the investor to excessive drawdowns or tail risk.
This is why economic viability requires more than prediction.
Five Conditions That Matter
The research identified five major conditions that should be evaluated before a short-horizon trading strategy receives meaningful capital.
First, the strategy must demonstrate predictive relevance. The signal should have a logical and empirical relationship to future returns. It should not be adopted simply because it performed well in one historical test.
Second, the evidence must be methodologically robust. A credible strategy should survive out-of-sample testing, sensitivity checks, alternative specifications, and reasonable safeguards against overfitting or data mining.
Third, the strategy must be implementable. Transaction costs, spreads, slippage, execution timing, liquidity, capacity, and turnover should be evaluated before concluding that the strategy can produce real-world results.
Fourth, the strategy must be suitable for the intended market regime and time horizon. A strategy that works during high volatility may not work during calmer markets. A strategy that works intraday may not work over several days, and a weekly strategy may not translate to a one-minute chart.
Fifth, performance should be assessed on a net, risk-adjusted basis. Raw return alone is not enough. Decision-makers should consider drawdown, volatility, Sharpe or Sortino ratios, factor-adjusted performance, benchmark-relative return, and downside risk.
A Seven-Stage Evaluation Framework
Based on the evidence, I developed a staged evaluation framework that can help managers, traders, and financial decision-makers separate predictive promise from deployable economic value.
The seven stages are:
Confirm predictive relevance.
Test methodological robustness.
Evaluate implementation feasibility.
Assess regime and horizon suitability.
Measure net, risk-adjusted viability.
Conduct controlled piloting before full deployment.
Continue monitoring, reviewing, scaling, restricting, or retiring the strategy as conditions change.
This framework treats strategy approval as a process, not a one-time decision. A strategy may be rejected, redesigned, restricted, piloted, scaled, or retired depending on how it performs at each stage.
The most important principle is that failure at one stage should change the management decision. If a strategy predicts returns but fails after transaction costs, it should not be deployed as originally designed. If a strategy performs well in one regime but not another, it may need restrictions. If a model performs well in a backtest but lacks out-of-sample validation, it may require additional testing before capital is placed at risk.
Practical Lessons for Business and Investment Decisions
The same logic applies beyond trading. Business owners, investors, and managers often face decisions based on promising numbers. A model, forecast, dashboard, or historical report may suggest that an opportunity is attractive. However, the decision should still be tested against real-world constraints.
For trading strategies, the key question is not simply, “Does this signal predict movement?”
The better question is:
“Can this strategy produce favorable net, risk-adjusted results under realistic conditions, in the market environment where it will actually be used?”
That question leads to better risk management, stronger decision documentation, and more disciplined capital allocation.
Final Thought
Predictive performance is valuable, but it is only the beginning of the evaluation process. A trading strategy becomes more defensible when it survives validation, implementation costs, liquidity constraints, market-regime changes, and risk-adjusted performance review.
For short-horizon strategies, the difference between gross return and economically attainable return can be significant. A structured, evidence-based framework can help decision-makers avoid overreliance on attractive backtests and focus instead on whether a strategy can realistically create value.
Author Note: This article summarizes doctoral research conducted by Monica Johnson-Hall as part of her Doctor of Business Administration program. The discussion is for educational and research purposes only and does not constitute investment, tax, legal, or financial advice.
