|
Calculation Error Chances are most plan sponsors are using time-weighted rates of return (TWRR) to calculate their performance. While TWRR is the preferred way to measure the performance of an investment manager, it falls short of providing information on how the portfolio did.1 Since plan sponsors typically control the cash flows, time-weighting is used to eliminate or reduce the impact of these flows, thus shielding the manager from client activities which could help or hurt their returns. However, at the same time, we lose the impact these returns actually may have on the value of the portfolio. Take the year 1987, which began as a continuation of the bull market. If our investor’s portfolio had $1 million invested at the start and they chose to add another $1 million in mid-September—just a few short weeks before the market adjustment of October—it’s highly likely they may have lost money for the year. However, it’s also possible that their manager may have reported a slight gain. This actually did happen a lot that year and many investors were quite confused as to how they could have a positive return but lose money. The use of time-weighting eliminated the impact of the return and the fact that the client lost money wasn’t consequential to the return measurement. Had we employed a money-weighted rate of return (MWRR) method, a negative return would have surely resulted, which would have made much more sense to our investor. The love affair with time-weighting began in the mid-1960s, when Peter Dietz wrote his study of the pension fund industry. He found that the internal rate of return (a money-weighted measure) was frequently being used to measure the performance of money managers. His argument against this was that the results were impacted by client cash flows. The Bank Administration Institute in the United States picked up on Peter’s work and published the first performance standards in 1968, strongly recommending the use of time-weighted measures. The Investment Council Association of America followed this up in 1971 with yet another standard, which also encouraged the use of TWRR. Unfortunately, along the way we seemed to have forgotten how money-weighting can actually be beneficial in many respects. We also mistakenly use TWRR for more than just calculating portfolio level returns: we also calculate the performance of securities, sectors, etc. by using TWRR, even though it’s typically the manager who makes these internal cash flow decisions. Because plan sponsors want to know how their account(s) are doing, as well as how the money manager is doing, we should see both TWRR and MWRR employed at the portfolio level. At the sub-portfolio level, where we measure the performance of securities, sectors, etc., we should see money-weighted returns, because the manager is controlling these flows. An argument can also be made that attribution should be done using MWRR. Where does the industry currently stand on this? For the most part, firms are using TWRR for just about everything. But, we are beginning to see some changes taking place as more and more firms are becoming aware of the benefits of MWRR. Portfolio managers, mutual funds, and software vendors are interjecting MWRR into their reporting. If a plan sponsor really wants to know how they’re doing, only money-weighted returns can provide the answer. Endnote |