Testing Predictive Relationships with Stationary Returns

Summary

This brief research note explains why asset prices are difficult to analyze directly: a broad equity index can drift over time, making price levels from distant periods poorly comparable. It distinguishes a predictive question from a contemporaneous association, using the example of whether a high VIX reading is followed by weaker S&P 500 returns on the next day.

The suggested approach is to work with returns rather than raw prices to improve stationarity, state the timing and hypothesis precisely, and visualize the data. The researcher should then challenge the preferred explanation and look for alternative reasons that could account for an apparent relationship. The document gives a research discipline, not an empirical answer: it presents no sample definition, statistical test, effect estimate, or evidence that VIX predicts subsequent index returns. Any observed association would need further testing and validation before it could support a trading decision.

Key ideas

• Price levels may drift, so comparisons across distant dates can be misleading.
• A predictive claim should specify whether one variable precedes another and by how long.
• Returns are generally more suitable than raw prices when seeking a more stationary series.
• Visualize the data and actively seek explanations that could undermine the initial hypothesis.
• The note frames a VIX and equity return question but provides no empirical result.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.