Implementation Shortfall

Implementation shortfall is a measure of the total cost of carrying out an investment decision. It compares the value that a portfolio would have had if the intended trade had been executed in full and at once, at the price prevailing when the decision was made, with the value actually obtained after trading. The difference captures commissions and fees, the price movement caused by the trade itself, the cost of delay, and the cost of any part of the order that was never filled.

The measure was set out by André Perold in a 1988 article, which contrasted a "paper" portfolio with the real one. It is now a standard tool of transaction cost analysis, used by asset managers to judge brokers, algorithms and their own trading desks, and to support best execution reviews. The price at which it starts, known as the decision price or arrival price, gives its name to a class of trading algorithms that aim to minimise it.

For a buy order, the shortfall in money terms can be written as:

• $Q$ is the number of shares the manager decided to buy.
• $q_j$ and $p_j$ are the size and price of each execution.
• $p_d$ is the decision price, the market price when the decision was made.
• $p_n$ is the price at the end of the measurement period, used to value shares that were not bought.
• Fees are commissions, taxes and other explicit charges.

For a sale the signs of the price differences are reversed. The result is usually divided by the value of the paper trade, $Q \times p_d$, and quoted in basis points.

Suppose a manager decides to buy 10,000 shares when the price is 50.00. The desk buys 8,000 at an average of 50.30 and pays 400 in commission; the remaining 2,000 are not bought, and the price ends the period at 51.00. The execution cost is 8,000 times 0.30, or 2,400. The opportunity cost is 2,000 times 1.00, or 2,000. Adding fees of 400 gives a shortfall of 4,800, which is 0.96 per cent, or 96 basis points, of the 500,000 paper trade. The same figure is reached by comparing returns: the paper portfolio gained 10,000, while the real one gained 5,600 on the shares bought less 400 in fees.

Analysts break the total into parts in order to see where cost arises.

• Explicit costs. Commissions, exchange fees and taxes such as a financial transaction tax.
• Delay cost. The price movement between the manager's decision and the moment the order reaches the market, including time spent with the trading desk.
• Market impact and spread. The difference between execution prices and the arrival price, reflecting the bid-ask spread and the market impact of the order's own demand for liquidity.
• Opportunity cost. The gain forgone, or loss avoided, on shares left unexecuted.

Part of the measured cost is not caused by the trade at all but by general market movement during execution. This timing risk can be positive or negative and averages out over many trades, which is why shortfall is assessed across large samples.

A common alternative is to compare execution prices with the volume-weighted average price for the day. That benchmark is simple, but it can be influenced by the order itself, since a large order forms part of the average, and it ignores delay and unfilled shares. A trader can look good against it by trading passively in line with volume while the price drifts away from the level at which the manager wanted to deal. The time-weighted average price has similar weaknesses. Implementation shortfall is harder to game because its reference point is fixed before trading begins, and it ties cost to the investment decision. Its drawback is that it requires an accurate record of decision times and is noisy for single trades. The narrower term slippage usually refers only to the gap between the expected and the achieved execution price.

Minimising shortfall involves a trade-off. Trading quickly reduces exposure to price drift but increases market impact; trading slowly does the reverse. Models of optimal execution, of which the framework published by Almgren and Chriss around 2000 is the best known, choose a schedule that balances expected impact against the variance of cost, according to the trader's aversion to risk. Arrival-price algorithms used in algorithmic trading apply this logic, typically trading more heavily at the start of the order. Choices about order routing, the use of dark pools and block trades are judged by their effect on shortfall.

Regulations requiring firms to take sufficient steps to obtain the best result for clients, such as those under MiFID II, have made systematic cost measurement a normal part of asset management. Managers report shortfall by broker, venue, strategy and order size, and investment committees use the results to set commission budgets and to decide how urgently orders should be worked. The study of how trading mechanisms create these costs belongs to market microstructure.