Alpha models live on paper; portfolios live in markets, where every weight change costs money. This note quantifies the demand for trading: how weights drift even with no decisions (passive drift), how turnover is defined and read as a holding period, and — the central derivation — how the turnover of an active strategy is governed by one statistic of the signal, its forecast autocorrelation . The note closes with the information-horizon machinery (lagged and horizon ICs) that explains why some signals churn and others don't, setting up the cost–alpha trade-off completed in transaction costs and implementation.
Learning objectives. By the end of this note you should be able to:
- derive the drifted-weight formula and explain the long/short asymmetry;
- define one-way turnover and convert it to holding periods;
- derive the turnover formula and discuss its four drivers;
- define lagged IC and horizon IC and state the aggregation rule;
- explain how turnover budgets tilt the optimal alpha model toward slower factors.
In this lesson
- 6.1 Passive portfolio drift
- 6.2 Turnover defined
- 6.3 Turnover due to forecast change
- 6.4 Information horizon and lagged forecasts
The full lesson, with worked examples and exercise solutions, is for members.
Read the full lessonNew to SelfMadeQuant.com? See membership options or browse the Portfolio Theory and Quantitative Investment syllabus.