Quantifying Key Properties of Trend-Following Rules
摘要
This chapter argues that an effective trading rule must exhibit three fundamental properties: accuracy, responsiveness, and smoothness. To quantify these properties, we introduce three objective measures derived from the weighting function of return lags in the trading indicator. These measures rely on a key assumption: market returns randomly alternate between bull and bear states. However, unlike traditional approaches based on parametric regime-switching models, our framework does not require estimating transition probabilities or state-dependent parameters. Instead, it directly utilizes the return weights inherent in the trading rule. This approach allows for a model-independent evaluation, enabling a precise comparison of different trend-following rules. The proposed measures provide a transparent and systematic way to analyze the tradeoffs between signal precision, reaction speed to trend reversals, and stability of trading signals. This framework offers traders and researchers a robust method for evaluating the effectiveness of trend-following strategies beyond subjective visual comparisons.