(Bloomberg) -- Richard Ennis knows a thing or two about how US public pension systems work. For half a century, he’s managed money for some funds and advised untold others at EnnisKnupp, a consulting firm he co-founded. He’s also the former editor of the Financial Analysts Journal and recipient of lifetime achievement awards for his work.
Now semi-retired, Ennis doesn’t pull punches: To him, the benchmarks that many public funds use to grade their investment performance raise questions about their integrity. “Bogus benchmarking is the single biggest problem in the field of institutional investing,” he said.
In his most recent broadside, in the April issue of the Journal of Portfolio Management, Ennis wrote that the public officials who manage $4 trillion for 26 million working and retired teachers, cops, and other public employees routinely set their benchmarks too low and in many cases receive bonuses for their accomplishments.
To Ennis, 78, the problem isn’t just that the benchmarks are too easy to hit. It’s also that pension fund employees and consultants who put together investment portfolios are often the same ones devising such yardsticks to measure their performance.
Ennis calls this benchmark bias. It can increase the already huge funding shortfalls at many public pension funds and shrink retirees’ monthly payments. It can also result in lower paychecks for public employees and higher taxes for everyone else. During the past two decades, the ratio of state and local government payroll spending used to fund pensions has tripled, to 15%.
Pension systems have wide latitude in selecting benchmarks. Some compare their results to plain-vanilla, transparent indexes, like the S&P 500 or the MSCI All World Index. Often, however, they use a complex mix of custom-designed formulas whose innards are impenetrable. “It’s the same at virtually every big fund,” said Ennis. “The staff cooks it up, in periodic consultation with the outside consultant. It’s then put before the clueless board for ‘approval.’ ”
The Public Employees Retirement System of New Mexico’s mind-numbing description of the benchmarks it has used since 2007 runs almost 800 words. Yet Deputy Investment Chief Kristin Varela said they’re far less complex today than they were a few years ago.
Sometimes pension funds simply modify existing benchmarks when they turn out to be too hard to hit. “What do you call a benchmark a pension fund didn’t beat? Its old benchmark,” joked Edward Siedle, a former Securities and Exchange Commission attorney who has audited dozens of pension funds. “No pension is going to keep a benchmark that makes it look bad, so the funds routinely change them to cast themselves in a more favorable light.”
The situation’s gotten so out of hand that about 10% of large institutions with investments in private equity don’t benchmark those investments at all, according to CEM Benchmarking Inc., which tracks institutional performance. Instead they use actual results as the benchmark: No matter how they perform, fund managers hit their mark.
In his article, Ennis took a deep dive into the benchmarks that 24 state pension systems used to rate their performance in the decade through June 2020. He constructed alternative benchmarks, using widely available data and indexes, to match the level of risk each fund took. (Bloomberg LP, the parent of Bloomberg Businessweek, owns benchmarks, including some Ennis used in his study.)
Theoretically, the performance of the pensions and Ennis’s risk-adjusted benchmarks should’ve matched. In fact, even though 19 of the 24 pension systems beat their own metrics, only two outperformed Ennis’s. On average, returns fell behind by 1.41 percentage points annually, suggesting they left $60 billion a year on the table for 10 years.
Ennis found that the total returns and benchmarks used by the California Public Employees’ Retirement System, the nation’s largest state pension plan, moved in lockstep, differing by just 0.03 of a percentage point in a decade. This suggests to Ennis that the yardsticks are being massaged to shadow returns.
The pension system lagged Ennis’s estimate of a best-fit benchmark by 1.14 percentage points, which adds up to a $5 billion annual shortfall. CalPERS declined to comment on Ennis’s findings.
Since the 1990s, many large investors have stopped benchmarking private equity against publicly traded equity, sometimes for good reason: Until sold, real estate, nonpublic firms, and other assets can be hard to value on a regular basis. Instead, pension funds use everything from aspirational hurdles, such as an arbitrary 15%, to measures “that lack sufficient detail to be classified,” according to CEM.
The North Dakota Public Employees Retirement System is among those using actual private equity returns as a benchmark. Its board is aware of the practice and its staff isn’t rewarded based on results, according to Chief Investment Officer Scott Anderson. “Any methodology used in the private area is an approximation,” he said.
“If you can’t come up with a good way to determine if your performance is good or bad, why are you doing it?” asks Alexander Beath, a senior research analyst at CEM. “Comparing performance to an actual return is giving up.”
The $14 billion in public pension money that Matthew Clark manages for the South Dakota Investment Council earned 9.4% annually in the decade through June 2020, trouncing most of his peers. Yet the benchmark South Dakota uses doesn’t fully reflect the risks taken to generate such returns, according to Ennis. The fund should have earned more like 12.8% from its relatively risky portfolio, which Ennis figured is equivalent to holding 81% equities.
Clark, whose bonus is based largely on how his fund performs long-term, believes Ennis’s model overstates the riskiness of South Dakota’s portfolio, yet praises his efforts. “I think his model is more meaningful when applied to a group rather than one fund,” Clark said.
The State Teachers Retirement System of Ohio, not part of Ennis’s study, paid five investment officials more than $500,000 each in salary and bonuses in 2020, even as the system lagged slightly behind its total return benchmark. The plan’s bonuses are based largely on how it does over the long term, according to a spokesperson. Board member Rudy Fichtenbaum, who has a Ph.D. in economics, said he has studied the benchmarks and the incentive formula but still feels in the dark. “I’m not a kindergarten teacher who just rolled in off the street,” said Fichtenbaum, “but when I read the pay documents it’s unclear to me how this works.”
©2022 Bloomberg L.P.
Advertisement