Friday, July 5, 2013

Fees and Investment Performance: Confusing Correlation and Causation

Fees and Investment Performance: Confusing Correlation and Causation

Correlation and causation are two related but different concepts.  Correlation measures the degree to which two variables move in concert with one another.  Causation implies that one or more variables are responsible for or explain another variable.  A recent study published by the Maryland Public Policy Institute and the Maryland Tax Education Foundation[1] seems to imply causation between high fees and poor investment performance in public pension plans over the last five years. In my view, this conclusion is incorrect. 

There is, indeed, an inverse correlation between management fees and investment performance.  However, high fees aren’t the cause of poor performance.  Rather, its asset allocation that explains the relative poor performance.  Pension plans with large and growing commitments to alternative investments have tended to lag behind conventional portfolios of stocks and bonds.  Of course, alternative managers charge much higher fees than their conventional counterparts, which explains the correlation between fees and poor performance.  Moreover, the higher fees imposed by alternatives make it all the more difficult for these asset classes to produce adequate rates of return.  Even without the higher fees, alternatives would not have beat stocks and bonds over the past five years.
Market Expansion (1999)

The study is on the right track, but it is flawed in a number of respects.  Thus, public pension officials were given an opportunity to discredit the study even though its general thrust is correct.  For example, the study grossly miscalculates the management fees paid by the North Carolina Retirement System.  The study asserts that NC paid 0.71% in fiscal 2012, when, as the State Treasurer points out, the fees were 0.42%.[2]  Although I can’t quite figure out how the authors derived the fees for North Carolina, I sympathize with the problem of coming up with the numbers.  Unlike many states, North Carolina doesn’t provide fee information in its annual report.  Rather, you have to dig through the Government Operations report made to the General Assembly.  Even if you adjust the study for North Carolina’s fees, the study’s conclusion stands; a correlation between higher fees and lower performance.

Despite this error, the authors ought to be pretty confident in their conclusion about the correlation between high fees and poor performance.  Their study doesn’t estimate performance fees or underlying management fees in fund of fund portfolios.  Any pension program with a significant commitment to real estate, private equity, or hedge funds incurs a huge slug of fees above the basic management fee.  In fairness to the authors, performance fees and fund of funds management fees aren’t disclosed, so it’s impossible to evaluate the true burden of fees.

The Maryland Retirement System attacks the authors’ suggestion that pension plans ought to use more index funds by pointing out that their active managers beat their benchmark by 1.2% in the year ended May 31, 2013.[3]  This is a laughable criticism.  There is absolutely no statistical significance to a 1.2% performance advantage over a 12-month period.  In fact, in the past one, three, and five years covered by Maryland’s last annual report published in December 2012, the pension plan’s domestic and international equities underperformed in every time period save one.  The Maryland system’s international equities lost 14.49% versus the benchmark loss of 14.57%.[4]  Active management achieved a whooping 0.08% advantage over the index in Maryland by losing just a little bit less.

As I’ve written in numerous posts, the foray by public pension plans into alternatives is a fool’s errand.  While the fees are excessive and drag down investment performance, high fees are only small contributors to the problem.  Alternative strategies tend to work if only modest amounts of capital are invested.  When America’s public pension plans join the world’s sovereign wealth funds in chasing private equity, real estate, and hedge funds, the results are entirely predictable: performance is going to languish.

Even if the Maryland study had been flawless, it would have been attacked.  Public pension officials have bet hundred of billion of dollars on alternatives and made alternative managers unconscionably rich.  Do you really think pension officials are going to welcome criticism?



[1] Wall Street Fees, Investment Returns, Maryland and 49 Other State Pension Funds, By Jeff Hooke and John J. Walters http://www.mdpolicy.org/docLib/20130701_MarylandPolicyReport201302.pdf
[3] Ibid

No comments:

Post a Comment