What’s Happening with the Pension Fund? -- Part 21

by Charles Schwartz, Professor Emeritus, University of California, Berkeley
schwartz@physics.berkeley.edu                          March 28, 2004


>> This series is available on the Internet at http://socrates.berkeley.edu/~schwrtz


New Documents Reveal Wilshire’s Con Game

A huge pile of documents were released following the lawsuit against the Regents over excessive secrecy in their investment activities. We can now see a clear pattern of duplicity by Wilshire Associates, with assists from some UC officials, in busting up the successful staff of the former Treasurer and moving billions into the hands of external investment managers.

Summary

     The main issue studied here is how UC’s investment consultant Wilshire Associates (with later assists from Treasurer David Russ) went about evaluating the past investment performance of the UC Treasurer’s office, mostly regarding the common stock portfolio of the retirement and endowment funds. The standard calculations for a professional evaluation answer the following questions:

•How do the past investment Returns, over various periods of years, compare to those of a designated Benchmark?
•How does the Risk (measured as the standard deviation of returns) compare to that of the Benchmark?
•How does the Risk-Adjusted Return (combining these two sets of data) compare to the Benchmark?
•How do these performance data rank compared to a relevant Peer Group?
We shall see that Wilshire misrepresented one of these criteria (Risk measurement), completely ignored another (Risk-Adjusted Returns), fabricated another (the appropriate Benchmark), and played switcheroo on the last (choice of a Peer Group).

     The story laid out in the now public set of documents starts when Wilshire was first hired by the Regents in 1999, goes through their policy recommendations to the Regents in early 2000 and finishes at the end of 2002 with the decision to move the management of all UC equity investments out of the Treasurer’s office.

First Reports from Wilshire

     A 10-page document from Wilshire Associates, “University of California; Total Fund Analysis; June 30, 1999”, appears to be their first technical report after being hired by the Regents to review UC’s investment operations. It starts with data on “Total Fund Universe Rankings.” The comparison “universe” they used is identified as “26 defined benefit [pension] plans with assets in excess of $10 billion, including 19 public funds and 7 corporate funds.” For the last 3-year period, UC ranked #2, with annualized Return 2.80% above the median; and for the last 5-year period (the longest covered by Wilshire), UC ranked #1, with annualized Return 4.61% above the median. WOW!

     The second page shows the results of Wilshire’s “Total Fund Risk Analysis.” Here we learn that, for the five years ending 6/30/99, UC’s total investment portfolio showed a level of Risk which was less (better) than the median (rank 38th percentile) at the same time that its annualized Return ranked at the very top of this universe. This, I think, says that UC’s outstanding performance should not be ascribed simply to luck.

     Subsequent pages of this report analyze separately the US Equity (stocks) and US Fixed Income (bonds) in the UC portfolio. The following are Wilshire’s numbers for the 5 year period. The UC Returns on Equity ranked at the 29th percentile of the peer group (almost in the top quartile), while the Risk for this portion was significantly lower than all of the peers (rank #1.) They also compare UC’s Equity performance to that of the Wilshire 5000 index: the Returns were almost identical while the UC Risk was much lower. For the UC Fixed Income investments, the Return was far ahead of all others in the peer group (rank #1), as well as far ahead of the Lehman Aggregate, which is shown for comparison; the Risk measurement here exceeded all the others’ (rank last.)

     This first report from Wilshire is a collection of numerical data, in tables and graphs, with explanatory footnotes; it is identified as the output of “Wilshire Compass”, the software database and analyzer developed by the company for their own and clients’ use. For a discussion of these results, along with Wilshire’s recommendations, we must turn to a separate document, “Review of the UC Treasurer’s Office Investment Program - Executive Summary, July 7, 1999.” Here we find quite a different story: “Total portfolio returns versus university peers is good, but misleading when used as a guide to evaluate the Treasurer’s investment program.” They describe both the stock and bond portfolios as having serious “risk factors”; and they repeatedly emphasize that the stock portfolio “has consistently under-performed the S&P 500 Index”, which they call its “benchmark.” We shall investigate these claims closely.

The Question of a Benchmark

     Lets dig into this question of what benchmark to use in studying UC’s past equity performance. The first Wilshire report did not mention the S&P 500 index at all; it only used the Wilshire 5000 Index for comparison (in addition to the peer group) and did not mention the word “benchmark.” I find a nice definition of this concept given by Wilshire’s own Stephen Nesbitt during the October 29, 2002 closed meeting of the Regents’ Committee on Investments (transcript released following out lawsuit: page 11.)

MR. NESBITT: I’m the honest broker. I work for you. I don’t work for the treasurer. And when you first hired us three-and-a-half, four years ago, when we did this analysis, we came to this conclusion [to outsource equity investments] back then.
For example, the reason you have an index fund representing 30% of your equity component, that was our recommendation because basically when we did the analysis back to 1983, and we saw the equity performance wasn’t good, our initial recommendation was for much more index. But we brokered a compromise and came up with 30%.
One problem three-and-a-half years ago in trying to critique the treasurer’s performance is that you - the Regents had no policies. So, in essence we could not hold the treasurer or staff truly accountable because you had no policies.
...
We all agree on some policies and then we can measure you against those policies. And then there can be no argument about what your objectives are.

     So, we understand that a benchmark is a specific standard of performance, pre-established as a matter of policy, against which one’s actual performance can be measured and one can be held accountable. Nesbitt says (in 2002) that prior to 2000, UC had no policy establishing any benchmarks for investment performance. Yet in July 1999, his Executive Summary asserts that the S&P 500 is the proper benchmark for UC’s equity portfolio, and he finds that the treasurer had underperformed this benchmark. But he spoke falsely both times.

     Among the pile of documents recently released is a comprehensive review of UC’s investment policies and practices, prepared by the Treasurer’s office and delivered to Wilshire early in 1999 at the Regents’ request. Here we find an extensive “Benchmark Study”, updated December 1998, which starts with a detailed chart showing two columns of specifications for each of the UC funds and the separate asset classes. The two column headings are: “Peer Group Benchmarks” and “Comparative Index”. For the Common Stock portions of UCRP (the pension fund) and GEP (the endowment fund) we find that the Peer Group Benchmark is, “SEI Equity Only Median”; and the Comparative Index is given as, “S&P 500, DJIA, Wilshire 5000.”

     A later description of “Frequently Used Benchmarks” tells us that “The SEI fund database covers 3,149 funds valued at $364 billion, of which 90% are pension funds and 10% are endowment funds.” And a table of historical returns for the total UCRP fund, annualized over ten years to 6/30/98, shows UC outperforming the SEI Balanced Fund Median by 2.8% and beating the SEI Large Plan (Over 1 Billion) Median by 1.1%. I did not find similar data for the Common Stock portfolio in those documents, but looking at the Treasurer’s annual reports we find that UCRP equities beat the SEI Equity Only Median over 10-year periods by 0.2% as of 6/30/98, by 0.6% as of 6/30/99 and 6/30/00.

     Further on in this document we find discussion of the role of the Comparative Indices.
The performance of each actively-managed asset class should be measured by an index which represents its passively-managed alternative. ... In other words, the Regents’ portfolios will not always “track” the chosen benchmark index as would a manager who is being paid to match a particular index’s sector weightings and portfolio characteristics. ... Typically, the Regents’ equity portfolios consist of large capitalization securities somewhat similar to the broad market as measured by the S&P 500 [index]. ... Over the long term, the common stock portfolios have slightly trailed the S&P 500.
     Thus, while the explicit “benchmark” established for evaluating performance of the UC Treasurer’s common stock portfolio was clearly the SEI peer group Median, a reader may conclude that the S&P 500 index also played a role for comparison purposes. It is entirely illegitimate, however, to assert - as Wilshire did without qualification or justification, in 1999 and thereafter - that the S&P 500 index is the benchmark by which to evaluate the past performance of the UC Treasurer’s domestic equity investments. One can redefine the benchmark for future reference, as the regents did in 2000 following Wilshire’s recommendation (adopting the Russell 3000 index for equities); but it is grossly dishonest to rewrite history as they have done. I raised this subject as a question in earlier reports and the new documents provide the answer: Wilshire fabricated this story of the S&P 500 as UC’s “historical benchmark” and, with help from Treasurer Russ, they got the regents to buy it.


Fudging the Evaluation of Portfolio Risk

     We noted that the June 30,1999 Wilshire Compass analysis of UC’s investments showed a quantitative measure of Risk in the stock portfolio which was outstanding: well below the market index and well below all other members of the peer group. However, in Wilshire’s July 7,1999 Executive Summary this data is not mentioned at all; and a completely different discussion of Risk is found.

[C]lose examination of the UC total portfolio reveals that it is unusual in composition and potentially very risky. The UC stock portfolio is concentrated in 65 large U.S. growth stocks with the largest 10 holdings representing almost 40% of the stock portfolio. This portfolio profile is very different from most large pension and endowment stock portfolios that are diversified into medium and smaller sized U.S. companies, and international stocks.

     Yes, UC’s investment style was unusual, and appeared, by simple counting of the portfolio composition, to lack diversification. But the purpose of diversification is to reduce risk and the calculation of risk in UC’s stock portfolio showed that UC carried a significantly lower level of risk than one might naively expect from such a “small” sized portfolio. In fact, the Wilshire 5000 Index, which is cited for comparison in their June 30 report, represents the most widely diversified portfolio of stocks; but UC’s stock portfolio showed a much lower measure of risk than this Index. Here, an honest professional should rely on the proper calculation, rather than on a superficial glance; or at least he ought to explain why these two approaches lead to such contradictory conclusions. But Wilshire simply pushed ahead with its own agenda and buried the conflicting data which it had itself produced in the first report.

     This same misrepresentation about the “risky” stock portfolio held by UC in the past appears again in the closed regents’ meeting in November 2002, and even in Treasurer Russ’ declarations during the lawsuit of 2003.

Looking at Risk and Return Together

     Even if we allow Wilshire to use the S&P 500 as a benchmark for measuring the past UC equity performance, there is a major roadblock to reaching the conclusion they do. This is the fact that while UC’s Returns over 10 years lag by about 1% per year - and this is the main argument used by Russ and Wilshire in the fall of 2002 - the UC stock portfolio shows a significantly lower Risk than the S&P 500. The correct way to balance these two factors (Was the lower return adequately justified by the lower risk?) is to calculate what are called “risk-adjusted returns”. This well established professional method is not mentioned anywhere in these Wilshire reports or in any of the newly released documents. In my earlier reports (see Parts 13-18 of this series) I pointed out this failure and did my own calculations, using official data from UC, with a resulting conclusion that was exactly the opposite of what Wilshire and Russ said.

     Was this a case of deliberate distortion/hiding of facts or might one imagine that Wilshire had some reason for not using such risk-adjusted calculations? A visit to Wilshire’s website and a display of their “Compass” program showed that they do indeed use risk-adjusted calculations - in the form of “Sharpe ratios” - and a look at the UC Equity Risk/Return graph in Wilshire’s June 30, 1999 report shows that UC’s Sharpe’s Ratio beats all the others (rank #1). One is thus led to believe that they did this calculation, saw the results, and decided to keep it quiet.

The 87th Percentile Ranking

     When the regents’ Committee on Investments met on October 29, 2002 to consider the proposal that all of UC’s equity investments should be given over to outside managers, one of the most damning “proofs” of poor performance by UC’s internal equity investment staff was a new comparison with peers, prepared by Wilshire and presented by Treasurer Russ. We read from pages 5-6 of the transcript of that closed meeting:

TREASURER RUSS: This chart, we’re on Page 4, shows the internal portfolio ... relative to the universe of large capitalization equity managers. ... That’s the other managers that manage in the U.S. equity arena. If you look at the far right, under 10-year, 10.34, 87 means 87th tile. Eighty-seven out of a hundred. And you can see for the 10 year, it’s at the bottom.

     The following table summarizes this data, shown on page 4 of item 603X, titled, “UCRP US Equity Quartiles.”

Table 1. Data from “Wilshire Large Core Universe” - Periods as of 06/30/2002
     
Annualized Total Return in %       (Percentile Ranking in parentheses)

1 Year
5 Years
10 Years
Top Quartile (25)
-15.71%
6.50%
12.89%
Median (50)
-17.94
4.45
11.94
Bottom Quartile (75)
-20.58
3.26
11.25
UCRP UC Treasurer
-22.66 (88)
2.29 (86)
10.34 (87)  <-----
S&P 500
-17.99 (54)
3.67 (67)
11.43 (65)




# of Managers in Ranking
179
165
  99
source: Item 603X, UC Regents Committee on Investments, meeting October 29, 2002

     The picture in Table 1 is a far cry from the first peer group comparison which Wilshire presented in its June 30, 1999 evaluation of UC. Without any explanation or justification they have replaced the former peer group (which was Wilshire’s own Large Pension Fund Universe) with a new peer group. They have pulled a switcheroo.

     Two things struck me as surprising in this new Wilshire comparison: the fact that the S&P 500 data are significantly below the median, when common lore says that most investment managers fail to keep up with this index; and the rather small size of the comparison sample (Morningstar, for example, gives comparison of mutual funds numbering well over a thousand for this asset class.) I first looked at some other UC published comparison data, which follows.

Table 2. Data from UC Treasurer’s Annual Report - Periods as of 6/30/2000
     Performance of Common Stocks      Annualized Total Return in %

1 Year
5 Years
10 Years
UCRP
11.4%
22.2%
17.4%
CRA Equity Only Median *
  8.6
22.2
16.8
*Capital Resource Advisors (CRA), formerly SEI, measures investment returns on approximately 5,500 portfolios. Those returns are gross returns and are before any investment management fees, which would be approximately 0.50% of average annual market value; UC’s total returns are net of (after) expenses of 0.04%.

     In Table 2 we see a much larger “universe” of peers; and UC’s performance stands well above the median. But this is data from a slightly different time period. So I went to the library and came out with the following data, from the best known industry source, which fits the exact same time frame and asset class as Wilshire’s data.

Table 3. Data from “Lipper Mutual Fund Quarterly”- Periods as of 6/30/2002
     Performance Benchmarks for Equity Funds      Annualized Total Return in %

1 Year
5 Years
10 Years
Large Cap Core Funds
-19.06%
2.06%
  9.57%
S&P 500
-17.99
3.66
11.43
source: BARRON’s (New York) July 8, 2002, page L-19

     Comparing the Wilshire data (Table 1) with the Lipper data (Table 3), we see that only the numbers for the S&P 500 agree. Wilshire’s Median returns are far larger than the Lipper averages, by 1.12%, 2.39% and 2.37% for the 1-, 5- and 10- years, respectively. The UCRP 10-year performance, which Wilshire ranked way down near the bottom (87th percentile), comes out well above the average in the Lipper data, by a margin of 0.77%

      There is a big discrepancy here. One expert I have consulted points out that the Lipper data includes management fees and expenses, which should be a lot lower for a large institutional investor like Wilshire’s clients. However, this correction would amount to only about 0.5% on the average, far short of explaining the 2.37% difference observed in the 10-year data for Wilshire vs. Lipper. As far as UC’s standing in the Lipper universe, this expense correction would cut UC’s 10-year performance to 0.31% above average, still very far from the “bottom” (87th percentile) ranking imputed by Wilshire.

     It is obvious, now, that Wilshire’s new comparison set is highly selected and not at all a fair basis for comparison of UCRP performance with peers. This looks to me like further evidence of a con job (fraud), set up by Wilshire and carried through by Treasurer Russ. What is further disturbing is the fact that nobody at that regents meeting commented on this serious bias, although plenty of investment experts were present, any one of whom should have noticed that something was amiss.

The $4.8 Billion Fracas

     The most outrageous example of cheap tricks in this now revealed saga, is provided by Regent John Moores, Chairman of the Board (and a well known investor from San Diego.) At the October 29 meeting, after Treasurer Russ had made his presentation showing that UC’s internal equity investments, over a decade or more, had trailed the benchmark by about 1% per year, we read, (starting on page 8 of the transcript)

REGENT MOORES: What do they look like cumulative...
TREASURER RUSS: We’re only showing the 10-year return here, but ...
REGENT MOORES: I meant the dollars. I just was curious about the dollars.
TREASURER RUSS: Well, we’d have to calculate that.
REGENT PARSKY: Why don’t you get it for John?
TREASURER RUSS: I’m not going to run a number off the top of my head on that.
REGENT MOORES: I’ll bet you can come close.
TREASURER RUSS: We can get a number, yes. ... [skip to page 13]
TREASURER RUSS: I have an answer to Regent Moores’ question. Dennis Sugino [from Wilshire] just handed me this. This just in. From the inception date of 12/31/83 to 9/02, the return on the equity portfolio ... is 11.01%. The policy index, 12.02. ... So 101 basis points.
     That was not the number Moores had asked for: he wanted to know how many dollars that shortfall amounted to. However, when it came to the November 13 meeting - when the Investment Committee recommendation was brought up for approval by the entire Board - Moores had his ducks all lined up. From the Minutes:
Chairman Moores praised the efforts of Regents Hopkinson and Parsky in revamping the University’s investment policies. He asked Treasurer Russ how many additional dollars would be in the University’s coffers if its funds had merely met the benchmark. Treasurer Russ reported that he and Auditor Reed had calculated a figure of $2,5 billion over the ten-year period. If they had been managed by outside professionals, another $2.3 billion might be added to that figure.

     These numbers were what newspaper reporters jumped on when these documents were released to the public, following the end of the lawsuit on January 12, 2004.  Any dope can say, “If I had invested in XYZ years ago, I would be very rich today.” Regent Moores shows himself as no better than a two-bit hustler; the suckers are the other members of the Board but they have no money at stake, we do.

Another Old Issue Clarified

     In my earliest paper in this series I criticized Wilshire’s March 16, 2000 Investment Strategy Study for its misleading presentation of data projecting future assets and liabilities for the pension plan, under alternative investment policies. My complaint was that a prediction of future quantities must be probabilistic and should be shown with the large uncertainties of such mathematical exercises, while Wilshire made it look like these were true numbers. What we learn now from the release of the minutes of those closed regents’ meetings is that this misleading data did indeed play a central role in convincing the Board to follow Wilshire’s recommendations.

Conclusion

     Wilshire’s dishonest campaign to get the money out of UC and into the hands of commercial investment managers, which I speculated on in my early reports, has now been confirmed by a wealth of new evidence. Earlier, it was difficult to provide a motive for this; but with recent revelations about the multitude of kickback schemes in the investment and investment consulting industry, that question seems answered.

     Finally, what blame falls to members of the UC Board of Regents? One or more were doubtless involved in the plot; it is likely that some others were biased toward the outcome (privatization); and most were merely dupes in the hands of slick pros. They all had opportunities to ask questions (they didn’t have to believe me in order to ask questions raised in my reports, which they all received). Yet they all failed, as far as the present records show, to exercise that minimum standard of responsibility.