In this article, there are three steps to the right execution strategy. The first step is to resist the temptation to go for the lowest commission. Commissions are a small part of trading costs, and any saving from lower commissions can easily be swamped by an increase in indirect trading costs: market impact and the opportunity cost of delayed execution or not getting filled.
Order difficulty
The choice of execution strategy depends on order difficulty. The second step, therefore, is to classify orders by difficulty and map them into strategies. Order difficulty has many dimensions. In this article, we focus on the three primary dimensions – order size, liquidity and trade urgency – and describe a framework, the Order-Difficulty Cube, for mapping orders into strategies.
The third step to the right execution strategy is to use post-trade execution quality analysis to evaluate the effectiveness of the various execution strategies and algorithms, learn from the analysis and improve the mapping.
Order size is one dimension of order difficulty that is influencing the choice of execution strategy: the strategy for an order 1 per cent of average daily volume (ADV) is different from the strategy for an order 50 per cent of ADV. Stock liquidity is another dimension: the strategy for orders in large-cap stocks is different than for orders in small-cap stocks. A third dimension is trade urgency: the execution strategy for a two-day horizon differs from the strategy for a two-hour horizon. In figure 2, we combine these three dimensions of order difficulty into the Order-Difficulty Cube. The easiest orders in the cube are the low-urgency, small orders in large-cap stocks. The most difficult orders are the high-urgency, large orders in small-cap stocks.
We next quantify how to position orders along the cube’s three dimensions and map them into strategies. We begin by ignoring intermediate orders (for example, mid-size in mid-cap stocks). These orders are difficult to finesse into the right strategy and the cost of not using exactly the right strategy is small. We divide each cube dimension into three and focus on the eight corner cubes.
Liquidity dimension
We use stock capitalisation to classify orders along the liquidity dimension. Here is one possible classification of orders into the eight corner cubes: (a) orders in stocks with market cap greater than $10bn, and (b) orders in stocks with market cap less than $1bn. One possible classification of orders along the size dimension is: (a) orders less than 0.25 per cent of ADV, and (b) orders more than 15 per cent of ADV.
The trade urgency dimension is difficult to quantify but extremely important. One factor influencing trade urgency is execution risk. Execution risk results from random price changes: the price is equally likely to move up or down, and the longer a trader takes to execute an order the more the execution risk. Over many executions these random price changes average to zero. Because of the correlation between time to execution and execution risk, however, traders that dislike risk have higher urgency to trade, especially in volatile stocks.
Trading alpha
Another factor influencing trade urgency is “trading alpha”. Trading alpha measures the likely price change over the trader’s execution horizon, aside from the liquidity impact of the execution itself. While a portfolio manager’s strategic investment horizon may be months or years, a trader’s tactical execution horizon ranges from a few hours to a few days. Trading alpha depends on the alpha of the underlying investment strategy, as well as on the trading of other market participants. Passive investment strategies, for example, with no long-term alpha may have positive trading alpha.
Buy-side traders can estimate the trading alpha they face on different investment strategies using data on past executions and input from their portfolio managers. Trading alpha typically ranges from 0 to 80 basis points (bp) over a five-day horizon. Traders facing low alpha have low urgency to trade and traders facing high alpha have high urgency to trade. One possible way, therefore, to classify orders along the urgency dimension is: (a) orders in low-volatility stocks and with trading alpha less than 10bp, and (b) orders in high-volatility stocks and with trading alpha more than 50bp (see figure 3).
We can now position orders in the cube. For example: in figure 2, the easiest orders are those less than 1 per cent of ADV, in stocks with market cap more than $10bn, in low-volatility stocks with trading alpha less than 10bp.
