Signal Turnover, Breadth and Neutralization
Turnover measures how much a portfolio trades. Learn how to calculate it, how it links signal decay to costs, and techniques to cut turnover without losing alpha.
Every time a signal changes, a portfolio must trade to follow it, and every trade costs money. Turnover measures how much of a portfolio is bought and sold over a period. Fast changing signals produce high turnover and high costs; slow signals produce low turnover. The art of portfolio management is to capture as much of a signal's predictive power as possible while keeping turnover, and therefore costs, under control.
Measuring turnover#
A common definition for a portfolio:
turnover = (sum of absolute changes in weights) / 2, per rebalance
annual turnover = turnover per rebalance × number of rebalances per year
An annual turnover of 100% means, roughly, that the whole portfolio is replaced once a year. Some sources count both buys and sells (without dividing by 2), so check definitions.
Turnover by strategy type#
| Strategy | Typical annual turnover |
|---|---|
| Passive index fund | Very low, often under 10% |
| Value and quality factors | Moderate, around 50% to 150% |
| Monthly momentum | High, often 200% to 400% |
| Short term reversal | Very high, often over 1,000% |
| High frequency strategies | Extremely high |
Approximate ranges; actual figures depend on construction.
Turnover, decay and costs#
Techniques to reduce turnover#
| Technique | How it works |
|---|---|
| Lower rebalancing frequency | Trade less often |
| No trade bands (buffers) | Only trade when a weight moves beyond a threshold |
| Signal smoothing | Use moving averages of signals so they change gradually |
| Holding period rules | Keep positions for a minimum time |
| Partial rebalancing | Move part of the way toward target weights each time |
| Netting across signals | Combine signals before trading so offsetting trades cancel. See Combining Signals |
| Cost aware optimisation | Include trading costs directly in portfolio construction. See Portfolio Optimization |
Buffers in practice#
A ranking strategy might buy stocks entering the top 10% but sell only when they fall below the top 20%. This hysteresis keeps stocks hovering near the cutoff from being bought and sold repeatedly, cutting turnover sharply with little loss of signal. See Rebalancing.
Optimal trading speed#
Nicolae Gârleanu and Lasse Pedersen (2013) showed mathematically that the optimal approach is to trade partway toward the target portfolio, moving faster for signals that decay quickly and slower for persistent ones, and to aim at a blend of current and expected future targets. This formalises the intuition that patient trading saves costs.
Turnover and capacity#
Higher turnover means more trading per dollar managed, which reduces capacity. A strategy with 1,000% turnover will hit market impact limits at a much smaller size than one with 100% turnover. See Alpha Capacity and Crowding.
Taxes#
In taxable accounts, high turnover realises gains frequently and may convert long term gains into short term ones, which are often taxed more heavily. See Trading Taxes and Capital Gains.
Frequently asked questions#
What is portfolio turnover?#
A measure of how much of a portfolio is traded over a period, often expressed as a percentage of the portfolio's value per year.
Why does turnover matter?#
Because every trade incurs costs, and high turnover can erase a strategy's edge, reduce capacity and increase taxes.
How can I reduce turnover without losing returns?#
Use no trade buffers, smooth signals, choose a sensible rebalancing frequency, net trades across signals and include costs in portfolio construction.
Next, learn the systematic sources of return in Factor Investing Explained.
3 quick questions on this lesson. Get them all right to finish it.
Turn on JavaScript to take the quiz.
Mentioned in
- Backtesting MethodologyResearch and Backtesting
- Portfolio and Multi-Asset BacktestingResearch and Backtesting
- Signal DiscoveryResearch and Backtesting
- Value FactorResearch and Backtesting
- Momentum FactorResearch and Backtesting
- Short and Long-Term ReversalResearch and Backtesting