Pension Performance Reporting Tricks

Pension Performance Reporting Tricks

Richard Branson is the founder of Virgin Group, a serial entrepreneur and a multi-billionaire.  Back in 2012, one of his Twitter followers asked him if there was a talent that came to him late in life that he wishes he had possessed earlier.  His response caught my attention:

In a prior post, I questioned whether it was appropriate for the Massachusetts state pension fund to report its performance on a gross of fees basis (Complex Public Pensions Have Lost Sense of Purpose).  I argued that reporting gross of fees was(is) meaningless.  For example, an executive could brag that his company’s gross profit was $100 million dollars but would it really matter if the company had over $150 million in expenses?  Gross is the total, but net is what really matters.

Last week, the man in charge of the Massachusetts Pension fund, Michael Trotsky, was given a $35,000 raise and a full 40% bonus.  A Boston Globe article quoted a member of the fund’s compensation committee who linked Trotsky’s pay-hike to the fund’s recent performance (State pension fund head gets a $35,000 pay raise — again).

The article went on to cite gaudy, benchmark busting 5-year return numbers that a casual reader would assume were indicative of great performance by Trotsky and his organization.  Do the numbers tell the full story, though?  Do they reflect actual returns or do they prey on those who suffer from the same lack of education that Sir Richard Branson once did?  Or, are there even deeper issues hidden in the reported numbers?  Let’s dig in…

Asset Management Sleight of Hand

On a monthly basis, the Pension Reserve Investment Management board of Massachusetts (PRIM) releases an investment performance report for its Pension Reserves Investment Trust fund (PRIT).  As I’ve previously mentioned, PRIM reports returns gross of fees.  This practice in itself should raise eyebrows and reporting “Value Added” (chart below) without accounting for fees is insulting.

Unfortunately, reporting on a gross basis isn’t the only PRIM standard worthy of questioning.  Although the board reports fund performance inclusive of private equity (PE) returns, that performance is compared to a benchmark (“Interim Benchmark”) that excludes private equity.   Wait…what?

screenshot-2016-12-13-10-04-18

Source: PRIM Performance Summary, September 2016

According to PRIM, the Interim Benchmark is used because “private equity is a long-term asset class that does not have a relevant 1-year benchmark”[1].  Fair enough.  I’d argue that for pensions, all asset classes should be considered “long-term” but that’s a discussion for another day.  The time horizon debate is a philosophical one.  The issue I want to tackle here is a more technical, and obvious problem.

For benchmarking to have any significance, you have to compare apples to apples.  If PRIM wants to compare fund performance to a benchmark that excludes private equity, then logically, it should compare the benchmark to fund returns that also exclude private equity.  With some quick, back of the napkin math, we learn a lot.

Deciphering the Returns

For the past five years, the private equity target allocation for PRIM has been 10% [2].  The actual allocation has been as high as 11.8%, and never lower than 10%.  I’ll conservatively estimate that PRIM’s average actual weighting to private equity over the past five years has been 10.5%.  As of September 30, 2016, the 5-year return for PRIM’s private equity portfolio was 16%.

  • Total 5-year return attributed to PE: 10.5% * 16% = 1.7%
  • 5-Year PRIT Total Fund Returns: 10%
  • 5-Year PRIT Total Fund Returns (ex. PE): 10% – 1.7% = 8.3%

An “apples to apples” 5-year benchmark comparison looks a lot different than what shows up in the monthly performance report:

  • PRIT Total Return (ex. PE) =  ~8.3% (gross)
  • Benchmark Total Return (ex. PE) = 9.3%

If we want to take it a step further and net out fees (~54bps)[3]:

  • PRIT Total Return (ex. PE) = ~7.76%
  • Benchmark Total Return (ex. PE) = 9.3%

By glancing at the published report, one would assume the PRIT fund had outpaced its benchmark by 70bps (10% vs 9.3%).  Pretty darn good!  In reality, though, the fund underperformed the benchmark by 100bps (or 154bps if you care about fees…which you should!).  There was no “Value Added”, only “Value Subtracted”.

These calculations tell us that if you remove private equity from the equation, PRIM has underperformed its benchmark by quite a large margin.  Considering 90% of the fund is outside of private equity, that matters.

It wouldn’t be fair to ignore private equity, though, so let’s see if we can figure out a better way to analyze the total fund performance including PE.

First, we have to come up with an estimate for private equity expected returns and build it into the benchmark.  Based on published reports, median private equity returns over the past 5 years were around 12%.  Using that number, we can attempt to construct our benchmark that is inclusive of private equity*.  Not entirely scientific, but should be a reasonable estimate.

  • Private Equity Return: 10.5% weight * 12% benchmark return = 1.26% of total return
  • Non-PE Asset Classes Return: 89.5% weight * 9.3% benchmark return = 8.32% of total return
  • Benchmark Total Return including PE = 8.32% + 1.26% = 9.58%

5-Year Benchmark Comparison (gross):

  • PRIT Total Return (including PE) = 10.00%
  • Benchmark Total Return (including PE) = 9.58%

5-Year Benchmark Comparison (net – what matters):

  • PRIT Total Return (including PE) = 9.46%
  • Benchmark Total Return (including PE) = 9.58%

Update (12/29/2016): I heard back from PRIM and it appears the Interim Benchmark DOES include PE, but uses actual returns as the benchmark, thus eliminating any excess returns.  This method is more reasonable than what I originally postulated. From PRIM…

“While this is not precise because of compounding etc, if you take the 5 yr return of interim, 9.3%, and subtract the 5 yr PE component (10% * 16.03%) you will get 7.7%. Then if you add back in the Core BM PE component (10% * 10.86%), you will get 8.79% which is relatively close to the 5-year core bm return of 8.77%.” 

Based on this calculation, PRIM has actually performed better than its published benchmark (net) (albeit a very confusing way to report):

5-Year Benchmark Comparision (net):

  • PRIT Total Return (including PE) = 9.46%
  • Benchmark Total Return (including PE) = 8.77%

I’ve been told that the core benchmark for PE is the 7-year annualized return of the Russell 3000 + 3%.  Although I understand how this is calculated for a point-in-time comparison, I’m having difficulty computing a comparable number for a 5-year performance benchmark.  My guess is that they are using an average of this number over the past 5 years which would come close to the 10.86%.  Until 2015, the benchmark was anchored a bit by 2008-2009 returns.  Now, as you can see in the chart below, the benchmark rate is rising rapidly.

Private Equity Benchmark

The explanation from PRIM is satisfying, but the report was certainly confusing.  This still doesn’t address the net vs gross issue and it will be interesting to see how the board communicates private equity returns in the future, with the benchmark being set so high!

Questions Worth Asking

Publishing a report that compares fund performance inclusive of private equity returns to a benchmark that excludes private equity returns doesn’t make much sense.  Compound the issue by ignoring fees entirely and the report is at best, meaningless, at worst, misleading.  This type of asset management sleight of hand can lead people to all sorts of false conclusions.

When I read about PRIM board members praising the fund’s excellent performance, I can’t help but wonder if they have any idea what the actual performance even is.  Do they understand that performance reports and comparisons published by PRIM for public consumption don’t tell the full story?  Are they aware that 90% of the fund has lagged its benchmark over the past 5 years?  Do they believe outsized private equity returns are likely to continue in the future?  Are they asking the right questions or are they accepting the data presented to them as complete truth?

I certainly don’t have all the answers.  My hope is that by continuing to raise these types of questions, the right people will start listening.  Making good decisions is very difficult without good information.  Increased transparency and clarity surrounding performance will undoubtedly improve the odds of long-term success for the Massachusetts state pension fund.

Notes:

* On the last page of the monthly performance reports, PRIM publishes a table that includes Interim Benchmark (ex. Private Equity) returns AND Total Core Benchmark (incl. Private Equity).  In the September 30, 2016 report, the 5-year annual return for the Interim Benchmark was 9.3%.  The 5-year return for the Total Core Benchmark was 8.77%.  More back of the napkin math would imply that, based on a 10.5% PE portfolio weighting, private equity returned 4.25% annually over the 5-year period.  This number is significantly lower than any returns I’ve seen.  I’ve asked PRIM for more information about how the Core Benchmark is being calculated and will update this post when I receive a response.

[1] PRIM 2015 CAFR, pg 7.

[2] PRIM CAFR Reports, 2011-2015

[3] PRIM 2015 CAFR


Tim Brennan

Comments

  1. […] Reporting returns, pre-fee, is kind of ridiculous. (ariadnewa) […]

  2. […] Reporting returns, pre-fee, is kind of ridiculous. (ariadnewa.com) […]