SACRAMENTO, Calif. - Both mammoth California state pension funds soon may be using customized performance benchmarks in reaction to what they see as deficiencies in the widely used Trust Universe Comparison Service.
The $107 billion California Public Employees' Retirement System already has decided to use a new customized universe, while officials at the $67.8 billion California State Teachers' Retirement System are testing the system's own investment performance benchmark.
CalPERS is making the switch after being told by its general consultant of problems in using TUCS and the National Association of State Investment Officers' universe as its key indicators of plan investment performance.
"These universes (TUCS and NASIO) have some significant deficiencies," wrote Roz Hewsenian, a consultant with Wilshire Associates, Santa Monica, Calif., in a letter to Sheryl Pressler, CalPERS chief investment officer. The letter was sent when CalPERS was first considering forming a customized performance universe late last August and recently was obtained by Pensions & Investments.
Wilshire, CalPERS' general pension consultant, produces TUCS in cooperation with TUCS master trusts/custodians.
After citing a number of problems with TUCS and NASIO-type universes, Ms. Hewsenian stated, "Because of these deficiencies, neither of these universe comparisons (TUCS and NASIO) are useful, and in fact can be dangerous to use because of their lack of explanatory information."
Most universes have problems?
In a later interview, Ms. Hewsenian said the criticisms contained in her letter to Ms. Pressler are true of universe comparisons generally, including those of Callan Associates Inc., San Francisco, NASIO and others.
George Wolfe, a Wilshire vice president in the division of Wilshire that produces TUCS, disputed the claim that TUCS has any consistency or calculation problems.
And in response to the charge that TUCS might be dangerous to use, Stephen Nesbitt, senior vice president and head of Wilshire's consulting division, said: "That is something you can say about any universe. That's not a problem necessarily of TUCS, but a problem of every universe."
Mike O'Leary, an executive vice president with Callan, said he believes large pension plans generally are more aware of deficiencies in generic performance comparison databases today than five years ago. He said some large public and corporate plans are adopting comparison databases that are customized for them, and Callan has provided some of them.
Mr. Wolfe, however, said TUCS has expanded its use of global data and construction of sub universes to meet client needs. He said he believes clients are more satisfied today than previously that their needs can be met using TUCS.
Among the issues raised in Ms. Hewsenian's letter:
Universes' investment return data are collected from a number of sources. Differences caused by data quality of investment return calculation methodology can obscure the universe comparison results.
These universe comparisons have not evolved to capture the way plan sponsors are managing their investment programs today. Consequently, the data cannot be parsed in a manner to provide meaningful comparisons below the total fund level.
These universes provide no explanatory information. There is no performance attribution available and no means to explain why a particular plan sponsor ranks in a particular quartile. In fact, these universe comparisons completely ignore the risk exposures of their constituents, a major explanatory factor of performance results.
The universes also have no capabilities to develop expected future returns of the constituents, nor do they address the difference in investment policy among constituents at either the total fund level or the individual asset class level.
'Quality' comparisons critical
The issue of quality performance comparisons for pension funds is a critical one.
Comparison universes can be an indication to an investment committee or board, or legislature, of how well a fund is performing. The better the performance, the less the employer needs to contribute.
A comparison also can give an indication of how well the internal staff and external managers are doing and whether changes in asset allocation, money managers or investment approach are needed.
Neither California pension fund has dropped TUCS as an investment performance service, and CalPERS continues also to use NASIO. But CalPERS intends to begin relying on its newly created benchmark as a key indicator of performance; CalSTRS may do the same with a benchmark it is testing.
Wilshire is putting together the new CalPERS performance benchmark, which is made up of 30 similar large public and corporate pension funds. CalSTRS' benchmark would be made up only of public funds.
TUCS is probably the most widely used performance comparison system in the world. Some 300 investment plans with more than 8,000 individual investment portfolios totaling some $800 billion contribute investment performance numbers to TUCS. A network of master trust banks uses TUCS data to provide plans with investment performance comparisons.
Mr. Wolfe said CalPERS and CalSTRS are so large that they might be unique, and a special approach might be needed in making performance comparisons. But Mr. Wolfe said the charge that balance and consistency problems exists in TUCS performance data is "groundless."
Because master trust banks and not pension consultants control TUCS, the consultants might be unaware that TUCS can be tailored to provide meaningful peer groups, said Mr. Wolfe. "I believe it would work" for CalPERS and CalSTRS, he said.
Other reasons for the change?
Some observers said they suspect the two systems are bypassing TUCS as a key measure because they haven't fared well against it.
They also said the political aspirations of some CalPERS and CalPERS board members - as well as political infighting - could be the reasons for the change.
"CalPERS is absolutely reluctant to be compared with anyone," said Jake Petrosino, a former CalPERS board member. "CalPERS thinks it's unique in the universe. They want to create a number of benchmarks and then compare themselves to the one they want, unless none of them makes them look good. Then they won't talk about it," he said.
But in a letter to CalPERS board members, Ms. Pressler, the fund's CIO, said both the TUCS and NASIO universes "have some limitations."
Ms. Pressler also referred to the letter by Ms. Hewsenian explaining limitations with TUCS and NASIO and why CalPERS needed to create its own database.
Like CalPERS, CalSTRS uses the TUCS universe of public funds of $1 billion in size or larger for its key performance comparison. However, CalSTRS officials said that one-half of the 23 public funds in that universe have legislative constraints, including restrictions on the types of investments they can make. This limitation and other problems with TUCS and other universes limit the value of direct comparison with CalSTRS, according to fund officials.
Also, the TUCS universe includes outliers - funds that are distant from the norm of large funds in the way they invest. Another problem cited by Ms. Hewsenian is a difference in return calculations that are used in performance measurement.
Peer group identified
For CalSTRS, Pension Consulting Alliance, Studio City, Calif., has identified a potential peer group. It includes: Florida State Board of Administration, Tallahassee; Los Angeles County Employees Retirement Association Pasadena; Minnesota State Board of Investment, St. Paul; Virginia Retirement System, Richmond; Washington State Investment Board, Olympia; and Wisconsin State Investment Board, Madison.
In an interview, Ms. Hewsenian said CalPERS is facing greater scrutiny than ever before. A lot of investment funds of a billion dollar in size, which are in the large public fund universe, don't have the exposure to international and private equity investments that CalPERS has, she said.
In the short, run, she said, some private equity and venture capital style investments have "zero investment return" because it takes time for the investments to start generating return.
"So CalPERS has exposure to (these types of investments) and their rate of return is probably not going to be realized for years. That is not only true of CalPERS but any fund that starts one of these alternative investment type programs," said Ms. Hewsenian.
CalPERS is being penalized for having to diversify into those kinds of assets when the publicly traded markets are doing so well, she said.
Also, a comparable universe of funds, she said, could provide information about investment approaches have or haven't worked in the past, what the fund should be doing differently, what the fund could expect in the future from a particular investment approach.
Public funds in Illinois and Florida already indicated their willingness to be part of the new CalPERS comparison universe.
Ken E. Codlin, chief investment officer for the $7.5 billion Illinois State Universities Retirement System, Champaign, Ill., said he wrote Ms. Pressler on the subject.
His fund doesn't use TUCS, but uses a comparison universe provided by consultant Ennis, Knupp & Associates, Chicago.
The problem of fund comparison, he said, is "kind of analogous to what is the right market index to compare yourself to. There are endless debates on that too, and again it kind of depends on what question you are asking and how finely you are trying to address the question."
Tom Herndon, executive director at the $60 billion Florida Board of Administration, also said he would be willing to participate in the CalPERS universe. "We are always on the lookout for a better mousetrap," said Mr. Herndon.
He said his fund uses an Ennis, Knupp universe for comparison.