By Travis Morgan, CIPM
A common approach for calculating carved-out performance is to simply take the asset class performance from the firm’s portfolio accounting software and show it as supplemental information. This represents a quick and easy solution, but should still be scrutinized by all parties involved. The main reason for this is a technical difference between an asset class return and a total portfolio return. With a total portfolio return, all external flows are assumed to be client directed and the manager has no discretion over the timing of these. With asset class performance, the system would typically treat all purchases and sales as an external flow. An important component of return calculations is the timing assumptions used. Portfolio accounting software typically allows you to decide if you want to assume flows occur at the beginning or end of day. The difference between the two can cause a large difference in performance depending on the situation.
Let’s take an example of a carve-out that holds 10 securities. At the start of the day the assets are worth 100,000 and let’s assume that the 10 securities are evenly weighted. At the end of the day, 2 of the securities were sold for a total of 24,000 and the remaining securities were valued at 83,000. If we first calculate the return using an end of day assumption we would come up with a return of 7% [(83,000-100,000+24,000)/(100,000)]. This is pretty intuitive and makes the most sense. The portfolio made 7,000 off of an investment of 100,000. If we then use a beginning of day assumption we would come up with a much larger return of 9.21% [(83,000-100,000+24,000)/(100,000-24,000)]. The difference is due to the assumption that the sell occurred at the beginning of the day, meaning, the assets made 7,000 off of a capital base of 76,000. The same argument can be made for purchases. Let’s use the same example as above but instead of selling 2 securities, they bought 2 securities at the end of the day for 24,000 and the total portfolio was worth 109,000 at the end of the day. An end of day assumption would result in a return of (15.00%) [(109,000-100,000-24,000)/(100,000)] while the beginning of day assumption would result in a return of (12.10%) [[(109,000-100,000-24,000)/(100,000+24,000)].
Looking at this you could argue that the manager who utilizes an end of day assumption would face the consequence on the downside and the manager would incur a larger loss. The problem with this logic is that buy/sell decisions are completely up to the manager. This means that if they were using a beginning of day assumption, they could simply only sell during days that the market was up or buy during days when the market was down and as a result experience a significant bump in performance. Using the same example above, what if the manager sold 5 securities for 56,000 and an ending market value of 51,000? You would assume that the return would be the exact same because the profit of the portfolio was still 7. If you were to use an end of day assumption you would be right. However, a beginning of day assumption comes up with a return of 15.91% [(51,000-100,000+56,000)/(100,000-56,000)].
As we can see, timing assumptions can cause quite a difference in performance and may warrant further scrutiny from all firms to ensure no one is taking advantage of it to distort performance.