Here's a question that keeps quant traders up at night: you've built a momentum factor that backtests beautifully, but how long does each signal actually stay valid before it turns to noise? It sounds like a data science problem, but it's really a market microstructure problem wearing a lab coat. Get it wrong and you're either trading too slowly — letting alpha evaporate — or churning your portfolio into transaction-cost oblivion.

The direct answer is that ASX mid-cap momentum signals typically exhibit half-lives somewhere between three and fifteen trading days, depending on the formation window and the specific factor construction. That range is wide enough to matter enormously. A signal with a five-day half-life managed like a thirty-day signal is roughly equivalent to serving yesterday's sushi with full confidence it's still fresh.

CONCEPTSignal half-life tells you how quickly a momentum factor's predictive power decays — it should directly govern your rebalance frequency.
WARNINGIgnoring decay speed in mid-caps costs you twice: stale signals reduce returns while delayed rebalancing inflates turnover when you finally act.
KEY IDEAAutocorrelation decay functions reveal the statistical fingerprint of a signal's useful life — not through intuition, but through the data itself.

Autocorrelation decay functions measure how correlated a factor signal is with its own past values across increasing lag intervals. Plot the autocorrelation coefficient on the y-axis against lag (in days) on the x-axis and you get a decay curve. The point where that curve crosses approximately 0.5 of its initial value is your half-life estimate. For ASX mid-caps, liquidity constraints and the relatively thinner order books mean information tends to get priced more slowly than in large-caps — but it also gets crowded out faster once discovered.

Momentum Signal Autocorrelation Decay — ASX Mid-Cap1.00.50.0-0.2½~12d half-life~7d half-life0102027dLag (Trading Days)12m formation6m formation

The practical machinery here involves fitting an exponential decay model — think AR(1) processes or more sophisticated HAR frameworks — to your factor's autocorrelogram. If the fitted decay constant gives you a half-life shorter than your rebalance cycle, you're systematically trading on stale information. ASX mid-caps add a wrinkle: index inclusion events, earnings calendars, and ASX200 rebalances create regime shifts that momentarily distort the decay curve. Robust practitioners segment their autocorrelation analysis by market-cap quintile and exclude the forty-eight hours around known corporate events. For deeper grounding in the statistical mechanics, the autocorrelation concept explained by Investopedia and the Wikipedia overview of momentum investing both provide useful context, while Investopedia's half-life explainer connects the statistical concept neatly to factor decay intuition.

The practical takeaway is simple enough to write on a sticky note above your monitor: measure your signal's half-life before you set your rebalance frequency, not after.

This content is for educational purposes only and does not constitute financial product advice. Past performance is not indicative of future results. Profit Logic Ltd (ACN 688 669 936) accepts no responsibility for errors or omissions in this content or anywhere on this website. Always seek advice from a licensed financial adviser before making investment decisions.