At the last STRS Board meeting, a Board Member asked, “What would have happened to STRS’s return if it had used index investing?” In that meeting, CEM reported that STRS earns a return of 7% compared to its policy return (a.k.a. benchmark return) of 6.3%. Thus, they claimed that STRS outperformed its benchmark by 0.7%, i.e., by 70 basis points (bps). Let’s assume that is true. Does that mean that if STRS used index investing it would have only earned 6.3% as claimed by Callan (STRS’s investment consultant) and CEM? The answer is no. How is that possible? The rest of this essay provides the answer.
STRS, like most pensions, uses “custom benchmarks” and peer comparisons to evaluate its performance. But neither “custom benchmarks” nor peer comparisons are true benchmarks. A real benchmark measures what is known as opportunity cost, i.e., how much you would earn using passive investments rather than active investments.
Passive investments involve buying all publicly traded assets that satisfy certain fixed, publicly known criteria. For example, for the S & P 500, the criterion is to be among the 500 largest publicly traded corporations in the U.S. stock market.
Active investing means choosing some segment of a market, or within a segment picking certain players, who the investment manager thinks will be most likely to outperform the market segment within which the picked players lie.
One thing important to remember is every time someone buys a stock, someone else must be selling a stock. More specifically, for every winner there must be a loser, and the amount won by one party matches the amount lost by the other. That is, trading is a zero-sum game. But trading is not free! So, even those traders fortunate enough to win must still pay to trade, and those expenses make it exceedingly difficult to outperform the market on a consistent basis.
So, what most pension managers have figured out is that if they can create a custom benchmark for their style of investing, they have a better chance of beating that benchmark. Then the game that gets played is each pension compares its performance against its custom benchmark.
Think about it this way. Each pension creates a measuring stick that has an arbitrary length. Then it divides that arbitrary length into 12 equal parts. Pension A reports a score of 6 out of 12, and pension B reports a score of 4 out of 12.
Callan, CEM, and the STRS investment staff would declare pension A did better than pension B, i.e., pension A beat its peer pension B. But what if the stick that B is using is twice the length of A? If you use one common system of measurement (a.k.a. an index) and that index is feet and we find that pension A’s stick is 1 ft long, then pension A gained 6”. But since pension B’s stick is twice as long as A’s stick, it is 2 feet (divided into 12 equal parts). The two pensions are not using one common system of measurement! So, B’s score of 4 parts is really 8”, and it has outperformed pension A. The only way for us to truly know who is performing better is if we all use the same measuring stick -- one common system of measurement.
So, when Callan, CEM, and the investment staff reports that STRS gained 7% with a benchmark of 6.3% -- i.e., beat its benchmark by 70 bps -- and someone asks, what would STRS have earned if it had invested in index funds, the answer is not necessarily 6.3% because STRS’s benchmark is not a true index. Suppose the S & P 500 index went up 7.1% and to buy that index fund STRS would have paid .02% (2 BPS). It would have earned 7.08% after expenses, i.e., slightly less than the index. But it still would have outperformed STRS which only earned 7%.
Richard Ennis, who helped create the field of institutional investment consulting at A.G. Becker & Co. and founded EnnisKnupp (the firm Ohio hired to help clean up the Ohio Bureau of Workers Compensation after Coingate), analyzed STRS’s performance using a weighted average of three indices. They were the Russell 3000, the MSCI ACWI ex US Hedged (Morgan Stanley Capital International All Country World Index excluding the U.S.), and the Bloomberg U.S. Ag (Total U.S. Bond Market Index). Ennis’ weighted average of the three indices outperformed STRS by 87 bps between 2009 and 2021. During that period, STRS had an annualized net return of 8.18%, whereas Ennis’ weighted average of the three indices specified above -- a diversified passive portfolio of stocks and bonds -- earned 9.06%.
So, once again it would appear that when asked a simple question, we did not hear the simple truth. That makes it hard to build trust, which everyone seems to agree is important. If we are to work together to solve the problems facing our pension, then the Board must speak with one voice and demand the truth. Only then will we start moving STRS in a direction where it can give members the pension they were promised, full retirement benefits at 30 years and a COLA.
Rudy Fichtenbaum, Trustee STRS Ohio
Dec. 18, 2023