The Pennsylvania Public School Employees' Retirement System, Harrisburg, has received a great deal of attention of late because of the recent discovery that a return they published was overstated. The result has been embarrassment to many parties, unexpected contribution increases and a federal investigation.
Last December, the $64.2 billion pension system reported a net annualized return of 6.38% for the nine years ended June 30, which was slightly above its actuarial-based objective of 6.36%. Though PennPSERS' return was reportedly audited by two independent parties, it was recently determined that the true return was 6.34%. While a 4-basis-point error might not usually seem material, a swing from outperformance to underperformance typically is, and it necessitated a change in contribution rates.
This revelation no doubt got the attention of other pension funds. No one wants to learn that a reported return, especially one that is a key determinant in defining contribution rates, was overstated. That said, there are lessons to be learned:
There are three major contributors to calculating returns: valuations, cash flows and what formulas are used.
Valuations can be tricky when they involve less-liquid assets or complex instruments. While certain assets, such as publicly traded stocks, can normally be priced daily, many others are valued monthly or, in some cases, quarterly. Funds need to ensure they have appropriate policies, procedures and controls in place to have confidence that the most accurate valuations are done.
The timing and size of cash flows can have an impact on the resulting return. In the case of pension funds, these flows, both in and out of the fund, are typically both regular and relatively stable. These flows, like the rest of the portfolio, will be impacted by what is going on in the market. Though it is common for funds to treat all flows as either start- or end-of-day events, we have found that the ideal default is for inflows to be treated as start-of-day events and outflows as end-of-day events. Over the past five or 10 years, we have seen more and more institutions and software vendors adopting this approach. The mistreatment of a cash flow can have an impact on the resulting return.
As for the calculation, we want to ensure it's both the right return method and that it is done in a proper fashion. There are broadly two approaches to calculating returns: time- and money-weighting. They serve different purposes, and it's important to understand their respective assumptions.
Time-weighting reduces or eliminates the effect of cash flows, which is important when evaluating the performance of managers that do not control the timing or size of the cash flows. The return tells us how the manager performed, taking the client's cash flows off the table.
Money-weighting takes flows into consideration, and is appropriate to evaluate the performance of managers that control the size and timing of cash flows (e.g., private equity managers). It answers the question, "How did the manager (that does control cash flows) perform?" Money-weighted returns are also appropriate to evaluate the fund itself, and answers the question, "How did the fund perform, taking cash flows into consideration?"
An argument against using money-weighted returns to evaluate the fund is mainly because the pension fund does not control the cash flows. While this is true to some extent, to exclude the impact of the flows, by using a time-weighted method, results in a return that does not speak to the question, "How did the fund perform?"
PennPSERS based its funding agreement with employees on time-weighted rates of return, which is arguably not the appropriate return because, as noted, it eliminates or reduces the impact of cash flows. The time-weighted return does not necessarily correlate with the overall gains or losses of the portfolio. For example, it is possible for a portfolio to lose money but have a positive time-weighted return. We sometimes see that occur when there is a sizable market downturn.
The preferred method to evaluate how a fund has done, relative to its actuarial return objective, is a money-weighted return: i.e., the IRR or internal rate of return. The IRR takes cash flows into consideration and reports how the fund has done. Because many pension funds have used time-weighted returns to evaluate performance for several decades, they have gotten used to that method, and sometimes seem a bit reluctant to change, despite the Government Accounting Standards Board's requirement for the annual reporting of an internal rate of return.
One question to ponder: Had PennPSERS complied with the all of the voluntary Global Investment Performance Standards, might the problem have been avoided?
Based on what we know, that is unclear. If it turns out that the valuations that were done were not up to what GIPS requires, then the answer would be "yes." However, there is no guarantee that the problem would have been avoided simply by complying. Two independent parties affirmed the reported return. Until we learn what was overlooked or done incorrectly, we cannot say.
That does not, in any way, suggest that pension funds shouldn't comply. The standards are recognized by many as best practice and dictate not only the formulas to use, but also how valuations should be performed. They require formal, written policies and procedures, which will result in the establishment of appropriate controls.
The standards require asset owners to report time-weighted returns, though they also recommend the publication of money-weighted returns. Time-weighted returns, as noted earlier, are helpful to determine "how the managers have performed," but not how the fund has done, especially relative to its return objectives. For that we need the internal rate of return.
The federal investigation, as well as the research that will be done by others, will likely provide answers to why the return was reported incorrectly. Hopefully, that information will be made public. But all pension funds can benefit from what has occurred, to ensure they have appropriate policies and procedures, as well as controls, in place; that they are valuing assets properly; and that they are calculating and reporting meaningful and correct performance results.
David Spaulding is founder and CEO of The Spaulding Group, a performance measurement consulting firm and GIPS standards specialist firm in Somerset, N.J. This content represents the views of the author. It was submitted and edited under Pensions & Investments guidelines but is not a product of P&I's editorial team.