Rudy Fichtenbaum: More on STRS v. Passive (Index) Investing

 STRS v. Passive (Index) Investing

By

Rudy Fichtenbaum

            This is the last part of a three-part series. In part 1, “Why it is So Hard to Beat an Index”, I reviewed the literature on index or passive investing v. active investing. In part 2, “Does Active Investing Beat Passive Investing”, I presented evidence that clearly showed that passive or index investing outperforms active investing. This is part three, and it will deal explicitly with STRS’s performance.

            To start, I want to make it clear that a pension needs to have diversified investments. It cannot just invest in the S & P 500. A diversified approach to passive investing would invest in broadly in U.S. equities, international equities, and bonds. Second, the reason it has taken me so long to present the third and concluding part of this series, is that I have struggled with a way to try and present this information in a way that would be both valid and understandable.

What I have come up with is two different methods of comparing STRS’s performance with passive investing strategies.

            Method One is to compare STRS’s performance to a portfolio of three Vanguard Index Funds, namely 56% VFINX (S & P 500), 34% VBMFX (Total Bond Market Index Fund), and 10% VGTSX (Total International Stock Index Fund). I used Vanguard funds because they make quarterly returns for their funds publicly available; that allows for the calculation of fiscal year returns that match STRS’s fiscal years. I started my analysis with 1989, because that was the earliest time when I could find reported returns in STRS’s Annual Comprehensive Financial Reports (ACFR), formerly known as Comprehensive Annual Financial Reports. In other words, I did not cherry pick my data.

            For the very early reports from STRS, it is not always clear whether the total returns reported were gross or net. But for purposes of my analysis, I am assuming they are net. If in fact some of the early STRS returns were gross, that would just strengthen my findings. I chose the 56/34/10 mix specified above because over the 28 years (1989-2025) STRS returns had a standard deviation of 9.64%, and the 56/10/34 mix had a standard deviation of 9.65%. The standard deviation is a measure of volatility and indicates the amount of risk that is being taken. In comparing returns, it is important to compare returns relative to the level of risk that is associated with each return, because according to modern portfolio theory, greater risk should lead to greater returns. 

            To repeat what I said above, I constructed my portfolio using three Vanguard index funds. For stocks, I used VFINX (the Vanguard S&P 500 Index Fund for Investors); for bonds, I used VBMFX (the Vanguard Total Bond Market Index for Investors); and for international equities, I used VGTSX (Vanguard International Stock Index Fund for Investors). Investor funds generally have higher expenses than institutional funds, but I used investor funds because the data available for investor funds go back further in time. For example, Vanguard does have a fund for institutional investors (VFFSX) with lower expenses, but annual return data was only available starting in 2017.

Using Portfolio Visualizer Backtest Portfolio Asset Allocation, I calculated the difference in the average annual rate of return VFFSX and VFINX from 2017 to 2024: VFFSX outperformed VFINX by 0.15%, the difference due of course to greater expenses for VFINX. I also looked on Vanguard’s website and found a note stating that the fund’s annualized six-month expense ratios for that period are 0.14% for Investor Shares and 0.01% for Institutional Select Shares. That would suggest the difference was 0.13%. So, I split the difference and adjusted the returns of VFINX by adding 0.14% to them to reflect the fact that STRS, as an institutional investor, would have access to lower cost funds. The expenses for VBMFX were listed as 0.15% and 0.01% for VBMPX; the difference in returns using Portfolio Visualizer Backtest Portfolio Asset Allocation in returns from 2010 to 2024 was 0.14%. In both cases, then, the difference was 0.14%; so, I added 0.14% to the annual returns of VBMFX, again to reflect the lower cost available to STRS. In the case of VGTSX, the investor shares have an expense ratio of 0.17% and institutional plus shares have an expense ratio of 0.05%, so I adjusted the investor returns by 0.12% to reflect the fact that if a pension were purchasing this index fund it would be at a lower cost than an ordinary individual investor.

One last point about the data: I used the S & P 500 Index because Vanguard’s Total U.S. Stock Market Fund, which is probably the equivalent of the Russell 3000, the index STRS uses to benchmark its U.S. stock performance, does not go back to 1989. However, the correlation between the S & P 500 and the Russell 3000 is 0.98.

Over that period, 1989 to 2025, STRS had an average annual return of 8.52%, and my diversified Vanguard portfolio had a return of 9.01%. In a more technical version of this paper, taking into account the interaction between STRS returns and its cash flows, I estimate that STRS would have had an additional $31 billion. That would have given STRS a 95% funding ratio. Of course, $31 billion is just an estimate. But even if it is too big by a factor of two, the lesson remains: STRS would have had many billions more had it pursued index investing. 

Method Two for comparing STRS’s performance with passive investing strategies is to use “return based style analysis” (RBSA), a method developed by Nobel Prize winning economist William F. Sharpe. This is the method that Richard Ennis employed when, in his analysis of STRS returns in 2023, he looked at STRS’s performance. Using RBSA, I have updated his results so that after the update, they apply to the period from 2009 to 2026.

RBSA uses quadratic programing to pick a mix of indices that minimizes the variance between STRS’s actual performance and performance from three indices, the Russell 3000, Bloomberg Barclays U.S. Aggregate (U.S. Bond market), and MSCI ACWI ex-U.S. (Hedged). The latter is a world stock market index that excludes the U.S. The hedged version, chosen by the model, is a version that removes gains and losses from changes in currency valuations. (STRS hedges 50% of its international investments). The model constructed a portfolio that was 58% Russell 3000, 28% Bloomberg Agg, and 14% and MSCI ACWI ex-U.S. (Hedged). 

From 2009 to 2026, STRS’s annualized return was 8.02%, whereas the return from the portfolio constructed by the model from the Russell 3000, Bloomberg Agg, and MSCI-ACWI over the same period was 9.18%.

This result is similar to that obtained when I used the Vanguard Funds. Further, the correlation between RBSA (Method Two) and STRS’s actual performance was 96%. In addition, the STRS portfolio had a standard deviation of 0.11 and the index portfolio had a standard deviation of 0.11. So, the index portfolio took the same risk that STRS took with its active management.

The conclusion that I draw from these results is that our members would have been better off if STRS had used passive investments rather than pursuing a strategy of active investing.

August 31, 2026