Featured Post

Was it constitutional for Proposition 124 to replace PSPRS' permanent benefit increases with a capped 2% COLA?

In this blog I and multiple commenters have broached the subject of the suspect constitutionality of PSPRS' replacement of the old perma...

Wednesday, April 6, 2016

PSPRS and its self-serving vision of success (plus investment returns through January 2016)

The following table shows PSPRS' investment returns, gross of fees*, versus the Russell 3000 for January 2016, the seventh month of the current fiscal year (FY), with the fiscal year end 2014 and 2015 returns included for comparison:

Report PSPRS PSPRS Russell 3000 Russell 3000
Date Month End Fiscal YTD Month End Fiscal YTD
6/30/2014 0.78% 13.82% 2.51% 25.22%
6/30/2015 -0.73% 4.21% -1.67% 7.29%





7/31/2015 0.13% 0.13% 1.67% 1.67%
8/30/2015 -1.43% -1.31% -6.04% -4.47%
9/30/2015 -1.02% -2.31% -2.91% -7.25%
10/31/2015 1.95% -0.36% 7.33% 0.08%
11/30/2015 0.37% 0.09% 0.55% 0.63%
12/31/2015 -0.95% -0.86% -2.05% -1.43%
1/31/2016 -1.41% -2.26% -5.42% -7.00%

There is usually about a two-month lag in PSPRS reporting its investment returns. There is nothing really new for this month with, once again, a smaller loss for PSPRS when the overall market shows a loss, and we should expect a smaller gain for PSPRS in February and March 2016, when the overall market had gains.  I did want to highlight a chart provided in the Board of Trustees meeting materials from March 23, 2016.  This chart shows the results of backward-looking stress tests of PSPRS' current portfolio during nine financial market stresses over the last 19 years:


Today's PSPRS Trust
Events Portfolio Actual
Asian Crisis of 1997 3.2% 5.7%
Russian/LRCM Crisis 1998 -3.7% -5.5%
WTC Attacks - Sept 11 0.1% -11.7%
Stock Market Crash 2002 -3.0% -21.1%
August Crisis 2007 3.8% 1.6%
January Crisis 2008 0.9% -2.7%
Credit Crunch 2008 (Aug - Nov) -9.0% -23.1%
Crisis 2009 (Jan - Feb) -4.7% -12.9%
Flash Crash 2010 -2.8% -3.7%

PSPRS touts the efficacy of their current portfolio by saying:
The current Global Trust Portfolio is now able to produce positive returns in four stress scenarios.
Downside potential losses of today's portfolio have decreased significantly.
The PSPRS Trust portfolio is 70% less volatile than public stock markets, and we continue to invest with returns expectations which are similar to those markets.
As the result of successful portfolio construction, the current Trust portfolio is now able to provide more positive returns in many stress scenarios.
While all these statements are mostly true--PSPRS' current portfolio is certainly less volatile than it former equity-concentrated portfolio--as with any information that trickles out of PSPRS, we have to look at it a little more closely to get the complete story.  PSPRS is doing its stress tests in isolation. We should certainly expect to see decreased losses when an equity-concentrated portfolio, especially one that consisted of individual stocks rather than index funds, is substituted with a more risk-averse portfolio. In hindsight it would have been a great plan if I had moved all my investments into cash in 2000 and/or 2007.  However, that does not mean that I should have permanently left all my savings in cash for the obvious reason that I would never have been able to realize the market gains following the financial downturns in 2001-02 and 2008-09.

Likewise for PSPRS.  Retroactively recalculating PSPRS' returns only when there was a dramatic financial downturn gives only half the story.  PSPRS' current portfolio would have done better than the actual portfolio during eight of the nine events, but what gains did PSPRS forgo during the periods outside of those events?  Without this full picture, PSPRS is being disingenuous about its current investment strategy.  Where would PSPRS' funded level be if the current strategy had been utilized for the entire 19-year period?  Running stress tests in isolation is like taking a handful of batting or pitching slumps from the career of a Hall of Fame player and substituting an average player into those slump periods to prove that the average player (playing averagely) would have helped the team(s) more than the Hall of Famer.  It leaves out all the positive data for the Hall of Famer that would show how much he helped his team(s) over the totality of his career.

Furthermore, PSPRS also neglects to compare its strategy against baseline passive strategies like the S&P 500, Russell 3000, high-grade government and corporate bonds, or some combination of all of them.  This would be the middle ground between the old and the current strategies.  If PSPRS had just invested in the S&P 500 over the past 19 years or in a combination of 70% in the Russell 3000 and 30% in high-grade corporate bond funds or any other combination, what would its funded ratio be today?  Passive strategies avoid the hubris of the individual stock-picking of the past portfolio or the overly complicated academic experimentation of the current portfolio.  We know the track record of passive strategies, which go back decades through all kinds of market ups and downs, and if a non-passive strategy does not produce better gains than the passive strategies, there should be serious discussion as to why the non-passive strategy is being utilized.

Though it would be better to compare the old strategy versus the current strategy to see which would produce a better funded percentage for PSPRS, it would not actually be possible to do this since the old strategy was essentially the stock picks of a single individual who was running PSPRS at the time.  (Though he has been criticized since he left, it has been conveniently forgotten that for most of his career he managed PSPRS' portfolio very successfully, though a passive strategy would have achieved the same results with far less risk.)  This individual may have continued to make bad picks after the dot.com crash or he may have gone on another hot streak and brought PSPRS back into the black.  And who knows how he would have done during the Great Recession.  Regardless, this is why using a comparison against a passive strategy would be the best since the data on the passive strategy is readily available. 

I wanted to do a simple comparison with the scant data available from PSPRS, so I compared what percentage of the Russell 3000 PSPRS was achieving over the time periods available.  Here is what it shows:

Year/Period PSPRS Russell 3000 Percent
2014 13.82% 25.22% 54.80%
2015 4.21% 7.29% 57.75%
3-year 6.16% 10.55% 58.39%
5-year 5.96% 10.40% 57.31%
10-year 4.34% 6.38% 68.03%

As can be seen over the past two fiscal years PSPRS and the three and five year annualized returns, PSPRS' returns have fit into a pretty tight range between 54.80% and 58.39% of the Russell 3000's returns.  All these periods coincide with the implementation of the current strategy.  Only the ten-year annualized returns show a significant difference at 68% of the Russell 3000.  Up until June 2009, PSPRS was still heavily invested in stocks and bonds, about 55% and 19%, respectively, and only about 26% in alternative investments.  The next fiscal year PSPRS began its change to its current strategy in earnest with stocks and bonds dropping to about 60% of its portfolio and alternative investments making up almost 40% of the portfolio.  PSPRS continued this transition over the years, and its current breakdown is about 37% stocks and bonds and 63% alternative investments.

What can we make of the fact that over the 10-year period PSPRS achieved a higher percentage of the Russell 3000 than all the other more recent periods?  The period between February 2006 and January 2011 includes the horrendous downturn of 2008-09 but also the upswing after the market bottomed out.  This could mean that PSPRS missed out on a good portion of the recovery by decreasing its stock and bond portfolio and moving more into alternative investments, but it is impossible to tell from the limited data here.  It is certainly intriguing and makes it all the more interesting to know what would have happened to PSPRS if it had maintained a 55/20/25 stock/bond/alternative investment portfolio or used a passive investment strategy that included just stocks and bonds.

Ultimately, PSPRS has to develop and implement a strategy that can earn its expected rate of return (ERR).  Actual returns matter only in relation to the ERR.  A pension not realizing its ERR is actually being underfunded, even when it has positive returns, and any deficit created will have to be made up in the future through higher contribution rates.  This is the problem we are seeing with PSPRS.  If PSPRS' current investment strategy can be expected to earn only about 55-60% of the Russell 3000, the Russell 3000 will have to average annual returns between 12.50% and 13.64% for PSPRS to achieve its ERR of 7.50%.  This is a very high bar to set for financial markets over the long run.

As mentioned earlier, there is limited data available to test PSPRS' current investment strategy, but the past five years have shown a consistent pattern.  Furthermore, the last five years are not cherry-picked like PSPRS' stress tests or speculative like PSPRS' forward-looking numbers.  The five years of data represent what actually happened under the current strategy, and what it shows is not good for PSPRS and its members.  If the current pattern continues, PSPRS will not earn enough to meets its obligations, go further into deficit, and have to decrease its ERR.  This means higher contribution rates will be required to keep PSPRS fully funded.  This will especially hurt the new group of Tier 3 members who will have to split normal costs 50/50 with employers.  As employee contribution rates rise for this tier, new hires will opt for the defined contribution pension (DC) over the defined benefit (DB) pension,which will lead to a downward spiral and eventual demise of the PSPRS DB pension.

With so much at stake, is it a lot to ask that PSPRS give a full and honest assessment of its investment strategy?

* Returns, gross of fees, are used because PSPRS usually does not report returns, net of fees paid to outside agencies, except on the final report of the fiscal year.  Returns, gross of fees, are used in the table for consistency.  The past two years fees have reduced the final annual reported return by about a half percent.  Returns, net of fees, were 13.28% and 3.68% for fiscal years 2014 and 2015, respectively.