Thursday, August 9, 2012

Another victory for money-weighting!

The United States Governmental Accounting Standards Board (GASB) recently introduced new provisions that call for the reporting of money-weighting rates of return, using the internal rate of return (IRR).

The new provision, Statement No. 67 of the Governmental Accounting Standards Board, titled Financial Reporting for Pension Plans, includes the following details.

On page 12, paragraph 30.b.(4) you'll find

And then on page 17:

The document also includes a glossary, with the following definition:

I had been speaking with someone from GASB a few months back, answering his questions as well as offering my views on money-weighting. Obviously, I'm thrilled that they opted to go with this measure, as opposed to time-weighting, which was their original plan.

If you're involved with or support a government pension fund, you'll want to ensure you adhere to these new rules. And, if you're not a government fund, it's still worthwhile to become familiar with them, because money-weighting is catching on!

Note: The Governmental Accounting Standards Board (GASB) is the independent organization that establishes and improves standards of accounting and financial reporting for U.S. state and local governments. Established in 1984 by agreement of the Financial Accounting Foundation (FAF) and 10 national associations of state and local government officials, the GASB is recognized by governments, the accounting industry, and the capital markets as the official source of generally accepted accounting principles (GAAP) for state and local governments.


  1. Stephen Campisi, CFAAugust 9, 2012 at 9:35 AM

    Once again we see the importance of understanding what your performance information is used for, what question you are answering. Money weighted returns are the only way to evaluate the returns of a client portfolio, especially when the cash flows are a critical aspect of the investment plan. In the case of pensions, the contributions of the employer, the level of the benefits and the withdrawals to pay for those benefits are all part of the plan. Oh yes, and there's also the performance of the investment portfolio. It's easy to see that a money weighted return is required to evaluate the portfolio's performance, since we need to reconcile the beginning portfolio, the plan cash flows and the ending value of the portfolio. This is definitional; there can be no controversy regarding this (unless you fail to understand what's going on or what your responsibility is in evaluating the performance.)

    Now if we want evaluate the results of a single fund or product manager in isolation, we should use a time weighted return. The right tool for the right job.

  2. Steve, well spoken, and obviously I agree. When I spoke with the fellow from GASB I made this point: what is it that you wish to assess. I will shortly unveil another body that is moving to money-weighting ... entirely! Again, "the right tool for the right job." Thanks!

  3. As much as I am a proponent of the money-weighted return (MWR) methodology, I have mixed feelings about GASB’s new accounting standard. I should be happy as I have testified to GASB on October 27th 2010 and was subsequently invited by them. So I may have had some impact in their decision to adopt MWRs.

    Unfortunately, I wasn’t aware at that time that my specific argument against using expected returns as discount rates rested on a modified MWR. I used the term “money-weighted returns”, not realizing that my argument required a slight modification to the final cash flow: It need to be adjusted for changes in funding status that are attributable to plan sponsor decisions such as a benefit increase or decisions to not fully fund new benefit accruals. While a standard MWR can be called “investor return” in the mutual funds world, it falls short of being a “plan sponsor return” in the DB world.

    Like David, I too start with the question “What is it that you wish to assess?” I believe that public plan sponsors should measure their “plan sponsor returns” and overall pension plan success. GASB wrote in their Exposure Draft

    “ Additionally, the money-weighted rate of return provides information that is comparable with the long-term expected rate of return of plan investments, which is used in calculating information presented in other disclosures.”

    Without the modification from above, we won’t have a return measure that is comparable to the long-term expected return. Girard Miller recently wrote (“COLA Freezes: Pension Reform’s Third Rail”, June 7, 2012)

    “In Baltimore, Md.; San Jose, Calif.; and a number of other municipalities, a similarly idiotic "COLA" surrogate or supplement was installed, giving retirees a pension increase or "13th paycheck" in years that markets outperformed the average. That policy guarantees that the pension fund will never achieve its expected return.”

    So Miller agrees that investment returns alone are not appropriate to compare to one’s expected return. Plan sponsor returns are the combination of Investment Returns (measured as a standard MWR) and the “Returns on Governance”. I call those plan sponsor returns the “Pension Plan Internal Rate of Return” (PenPIRR).

    I applaud GASB for taking a step in the right direction. MWRs account for path dependencies in the actual return sequence and for the actual timing of contribution and benefit payments. Yet I am disappointed and frightened that GASB stops there. Why? Because I also believe that GASB’s MWR presents an even bigger danger of misinterpretation than time-weighted returns. If plan stakeholders start believing that this MWR is their relevant plan sponsor return and that it is comparable to their expected returns, than they will continue to make wrong decisions on both the investment side and on the benefit side of the plan.

    I believe that an isolated look at the investment side is likely to mislead. A highly volatile strategy leads to more or less predictable actions on the benefit side of the plan – e.g., a benefit increase when the plan is temporarily overfunded. Since this is not yet cash flow relevant, this action is not captured in a MWR. A MWR cannot account for various other interactions between plan sponsors and their investment managers, e.g., the systematic underfunding of new benefit accruals. The latter typically results in underfunding which may lead to riskier investment strategies.

    I understand PenPIRR goes beyond the typical scope of investment performance measurement. It translates the effects of typically uncovered plan sponsor actions into return space. For calculating the adjustment amount for the final cash flow we need the liability structures for each covered period – something performance measurement professionals may not be familiar or comfortable with. But I believe that it is PenPIRR and not a standard money-weighted return that is “the right tool for the right job”.


Note: Only a member of this blog may post a comment.