Transaction costs in systematic investing: spread, slippage, and market impact
Gross returns and net returns diverge because of transaction costs. For a buy-and-hold investor who trades infrequently, this gap is negligible. For a systematic strategy that rebalances monthly, or a momentum strategy that turns over its entire portfolio several times a year, transaction costs can consume a substantial fraction of gross alpha. Understanding the components of transaction costs and how they interact with strategy design is a prerequisite for realistic performance expectations.
Components of transaction costs
Commission is the explicit fee charged by the broker or exchange per trade. In equity markets, commissions have fallen dramatically over the past decade and are near zero at most retail brokers. They remain meaningful in futures, options, and some fixed income markets. Commission is the most visible component of cost but often the smallest in modern equity trading.
Bid-ask spread is the difference between the price at which a market maker is willing to sell (the ask) and the price at which they will buy (the bid). A round-trip trade—buy then sell—costs at least one full spread. For liquid large-cap equities, the spread may be just one or two basis points. For small-cap equities, illiquid ETFs, or currency forwards, it can be tens of basis points. Spread is paid whether the trade succeeds or fails and is not visible on a trade confirmation.
Slippage arises from the difference between the price at which a trade is decided and the price at which it is executed. In fast markets or with delayed execution, the price can move against the investor between signal and fill. Slippage is especially relevant for strategies based on end-of-day signals executed at the next open.
Market impact occurs when a large order moves the market against itself. Buying a large position drives the price up; selling drives it down. Market impact is the cost of size and is non-linear: doubling the order size typically more than doubles the impact. It is the dominant cost for institutional strategies but negligible for most retail investors.
Estimating total transaction cost
A simple but useful approximation for total round-trip cost is:
Total cost ≈ Spread + Slippage + Commission
For a typical liquid equity ETF held by a retail systematic investor, round-trip cost might be 5–15 basis points per trade. A strategy turning over 100 per cent of the portfolio per year (annual turnover of 1.0) would incur approximately 5–15 basis points per year in transaction costs. A strategy turning over 400 per cent per year would incur 20–60 basis points—a drag that must be offset by at least that much gross alpha before the strategy is profitable net of costs.
Implications for strategy design
Transaction cost awareness changes strategy design at every level. Signal frequency matters: daily rebalancing incurs far more cost than monthly rebalancing, and the signal must be proportionally stronger to justify the higher frequency. Position sizing matters: concentrated portfolios trade fewer positions but each trade is larger; diversified portfolios trade many positions but each is smaller. The relationship between alpha decay (how quickly a signal loses its predictive power) and transaction cost determines the optimal rebalancing frequency—a signal that decays slowly favours less frequent, lower-cost rebalancing.
Limitations
Transaction cost estimates are themselves uncertain. Spreads widen in volatile markets. Slippage is hard to measure retrospectively because the counterfactual price is unobservable. Market impact models are based on historical data and may not extrapolate to extreme market conditions. Backtests that ignore or underestimate transaction costs systematically overstate net alpha—this is one of the most common sources of overfitting in strategy development.
Transaction costs in pfolio
pfolio applies a conservative transaction cost model to all backtested strategies, incorporating estimated spread and slippage for each asset class. The model parameters are documented in the platform's backtesting methodology disclosure. For live portfolios, actual transaction costs are visible in the trade log and in the performance attribution breakdown.
Related articles
Disclaimer
Get started now
