Moving-average cross strategy shows no edge on TLT
Daniel Okafor, Strategy analyst (AI-assisted) ยท
The moving-average cross strategy, particularly using a 200-day moving average, is often employed by traders to identify potential entry and exit points. However, recent tests have raised questions about its effectiveness, particularly on TLT, the iShares 20+ Year Treasury Bond ETF. The question at hand is whether this strategy can provide a sustainable edge in out-of-sample conditions.
The Verdict on Performance
Based on our own test data from November 2, 2026, the moving-average cross strategy applied to TLT produced an out-of-sample Sharpe ratio of 0.21. This indicates a positive return relative to the associated risk; however, it does not indicate a significant advantage when considering the adjustments for multiple comparisons. The p-value of 0.274, following the Benjamini-Hochberg correction for false discoveries, reveals that the strategy does not have a statistically significant edge over the buy-and-hold approach [1].
Summary of Findings
| Metrics | Moving-Average Cross (n over 200) | Buy and Hold |
|---|---|---|
| Out-of-sample Sharpe ratio | 0.21 | 0.02 |
| Yearly growth | 1.7% | -0.8% |
| Worst fall | -25.9% | -48.4% |
While the moving-average cross strategy does show an improvement in yearly growth (1.7% compared to -0.8%), it's essential to note that its worst fall of -25.9% is significantly better than the buy-and-hold strategy's decline of -48.4%. However, this performance does not imply consistent outperformance, especially since the strategy's advantage was not statistically significant [1].
Testing Methodology and Results
When evaluating the moving-average cross strategy, The strategy underwent testing over a span of 10.8 unseen years, specifically from January 4, 2016, to October 30, 2026. The initial allure of a Sharpe ratio of 0.36 during the in-sample performance faded quickly in the out-of-sample period, where it could only muster a Sharpe ratio of 0.21 [1].
The testing framework was robust, incorporating 90 different strategies across various settings, with the moving-average cross strategy being just one among them. However, none of the strategies, including this one, maintained a significant edge after applying the correction, so none of the 90 tests survived [1].
Position Changes and Costs
During these unseen years, the moving-average cross strategy executed 23 position changes, each charged 5 basis points [1]. That is 1.15% in total costs over the period (23 x 0.05%, computed here), small beside the 1.7% yearly growth but counted in the result. The pass/fail rule was written on 2026-11-04, after the 2026-11-02 run had been read, so this is a labelled rule applied to a finished run, not a pre-registration [1].
Questions readers ask
What does the Sharpe ratio indicate about the strategy's risk?
The Sharpe ratio measures the return of an investment relative to its risk. A higher Sharpe ratio typically indicates a more favorable risk-return profile. In this case, the Sharpe ratio of 0.21 suggests that while the moving-average cross strategy provided some return, its risk-adjusted performance is not compelling compared to a benchmark.
Why is the p-value important in trading strategy assessments?
The p-value helps determine the significance of a strategy's performance. A lower p-value suggests that the observed performance is unlikely to have occurred by chance. In this instance, a p-value of 0.274 indicates that the strategy's edge was not statistically significant, implying that the results could be the result of random fluctuations rather than a reliable trading signal.