WHY ORDER SIZE MATTERS
Integrating Portfolio Management and Trading
The traditional buy-side firm has separate groups responsible for portfolio management and trading. While in many cases this structure reflects a reasonable separation of labor, it is critical that portfolio management communicates enough information to allow the trader to employ the optimal trading strategy for each portfolio manager or fund strategy.
This is particularly relevant in the case of order size. From the PM’s perspective, it may be intuitive to avoid transmitting an entire position to the trader at once, in the hope that several smaller orders incur less market impact. However, in instances in which the PM is participating in a moving or trending market this may unnecessarily constrain the trader and result in higher implementation costs.
The most basic instruction a PM can give the trader is quantity of shares to trade. After the PM has determined a desired position size, it is to the PM’s benefit to communicate this full size to the trader rather than attempt to parcel out the order over multiple days.
Communicating Order Size and Participation Rates: An Example
Consider a PM with an average order life of 4 days and an average participation rate of 5% of volume. The PM has some skill as a market timer that warrants the trader being more aggressive when trading the PM’s orders.
In a scenario in which the PM communicates the full order quantity to the trader on the first day of trading, what had been a 4 day order at 5% participation would finish in 2 days at 10% participation.
By shortening the order duration, the trader has avoided the less favorable (i.e., higher for a buy) prices available on days 3 and 4 and made use of the information that the PM has some skill as a market timer. While the more aggressive execution strategy (10% participation) will incur higher market impact than the original execution strategy (5% participation), in a trending market the extra impact will be more than offset by the better cost basis obtained by completing the order in a shorter time frame. Below is an image illustrating this scenario.
Now consider the alternate scenario in which the PM instead chooses to reload orders over each of the 4 days. When the trader doubles participation this time, the order finishes earlier each day rather than in fewer days, and the trader is unable to take advantage of the PM’s skill as a market timer.
Because the prices available to the PM become worse with each passing day, the prices available on days 3 and 4 are worse than those available on days 1 and 2. Therefore, the trader who received the full order quantity on day 1 will have lower total trading costs, the manager will have a lower cost basis and the fund will have better performance than if the PM chooses to reload orders over each of 4 days.