Well, we’ve just learned that CalPERS is afraid of something else besides the California legislature. The Wall Street Journal also rates.
The Journal ran a rather peculiar story today, ONLY ROBOTS CAN TALLY WHAT THE LARGEST U.S. PENSION FUND PAYS IN FEES. As we’ll discuss shortly, the story focuses on CalPERS but gives an odd characterization of how public pension funds are using software tools. A bit of backstory may illuminate why the story came out the way it did.
This section comes right before the close:
Some critics also say the way Calpers presents the new data can be confusing. Earlier this month, for example, Calpers showed its board a chart illustrating that management fees and expenses paid for investments had fallen to $638 million in fiscal 2016 as compared with $1.04 billion in fiscal 2011.
But that $402 million difference excluded $121 million in management fees and “partnership expenses” such as legal and auditing costs, referencing these additional charges in a footnote. Calpers in a press release last week touted the drop without mentioning the extra charges.
A spokesman said Calpers will correct the press release.
It’s not hard to infer what went on. The Journal was not buying what CalPERS was selling about its presentation and was going to give it prominent play. But CalPERS decided to retreat instead.
Why is this a reasonable inference? A mutual source said as much.1 So the reporters presumably had to reorient the piece late in the game.
Moreover, it fits with other evidence. On May 3, Sam Sutton at Buyouts Magazine wrote about CalPERS’ failure to include private equity fund-level expenses in its budget:
Unlike previous years, however, CalPERS’s proposed budget for PE-management fees in the 2017-2018 fiscal year does not factor in expenses private equity firms bill to their fund investors, including amounts for legal and auditing services. Though CalPERS still pays these expenses, they are not accounted for in its new budget proposal, CalPERS spokespeople told Buyouts.
The Sutton story is detailed. Among other things, it mentioned that the excluded fund level legal and accounting costs were $75 million in the past year. This is the same figure that CalPERS board candidate Margaret Brown focused on in her e-mail and public comment to the board, since CalPERS didn’t correct the figures historically and worse, tried to depict the exclusion of this expense as actual cost reduction, as opposed to an accounting change. But in the Sutton article, CalPERS had already marshaled board members Henry Jones and Priya Mathur to defend the exclusion.
Keep in mind that this was not the only cost removed this way. CalPERS also eliminated fund of fund fees of $46 million, bringing the total to $121 million.
CalPERS’ justification to Sutton relied heavily on the barmy claim, “We can do separate this cost now because we have our fancy new private equity computer system”:
In interviews, CalPERS spokespeople attributed the elimination of partnership expenses from the budget to a shift in how the nation’s largest public pension monitors its PE portfolio’s fees and expenses.
CalPERS launched its Private Equity Accounting and Reporting System, or PEARS, in early 2015. The new accounting system allows pension staff to segregate annual management fees from the other fund expenses that private equity GPs charge to LPs.
In the past, those expenses were often lumped together with management fees in private equity fund tax documents, CalPERS spokeswoman Megan White said. CalPERS finance staff used those tax documents to formulate the retirement system’s annual budget. CalPERS staff was unable to extract fund-related expenses from private equity management fees, which is why in previous years they were presented as a lump sum in budget proposals and other retirement-system documents, including its annual financial
This is ridiculous. The data was there before PEARS, and to the extent some general partners weren’t as forthcoming as others, CalPERS could request additional information. South Carolina, widely cited as an industry leader in getting cost information, has done so by being rigorous and insistent. Deciding what tool to use to compile the information when you have it isn’t the most difficult part of this equation, even though CalPERS would have you believe otherwise.
I spoke to Sutton last week. He had pointed out that it was inconsistent for CalPERS to include certain types of costs historically, exclude some of them this year, and then treat an accounting change if it were a bona fide cost reduction, and asked for an explanation. CalPERS ignored his query.
It looks as if the Wall Street Journal reporters went down the same path as Sutton, perhaps based on seeing his report. And it appears that CalPERS didn’t want to have the Journal calling out its misleading press release. But the Journal wanted to put down a marker to show that they had gotten CalPERS to relent.
With that as background, no wonder some contacts read the Wall Street Journal story as having some fun at CalPERS’ expense while maintaining the veneer of playing it straight.
For starters, have a look at the atypical formatting:
And better yet, “The Quants” is a ticker-style gif.
The butt of the joke is CalPERS’ regularly touted PEARS system. The Journal curiously doesn’t name PEARS, which is CalPERS Private Equity Accounting and Reporting Solution but golly gee, see what it does:
The nation’s largest pension plan has 380 people overseeing roughly $320 billion in assets. But when one of its top officials was asked during a board meeting how much in performance fees was paid to private-equity managers, he had to acknowledge no one knew.
“We can’t track it today,” said Wylie Tollette, the chief operating investment officer of the California Public Employees’ Retirement System, at the 2015 meeting. The disclosure surprised board member JJ Jelincic, who said: “If you can’t track them they are kind of hard to manage.”
Only algorithms could find the answer. A software program developed by outside firms determined at the end of 2015 that Calpers had paid $3.4 billion in performance fees over the past quarter-century to the private-equity firms that managed its money. In 2016, that number was $490 million.
As the nation’s largest public pension funds plunge deeper into complicated investments as a way of chasing returns, they are becoming more reliant on machines to make sense of it all. Some executives worry that a greater reliance on databases, coding and other quantitative tools creates the false impression that they have a better handle on their investments than they actually do.
Has tech inflation become the new grade inflation? What used to be called analytics, at least for CalPERS, has been recast as algos and robots. In fairness, I’ve never seen CalPERS make such grandiose claims for PEARS. But that may be because “We can’t yet do that in PEARS” has become a new staff excuse for blowing off board inquiries.
And as someone who grew up in the stone ages of finance, and had to prepare spreadsheets on green accounting ledger paper, finding and entering numbers from annual report and computing them with a calculator, I can attest that doing that sort of yeoman work gives one a much better understanding of finance than providing some inputs to a program and having a spreadsheet or data table pop out.
But questioning the use of databases? Banks have been using them as soon as they could get their hands on them, which is why so many are dependent on mainframes using Cobol. And the Journal makes clear soon enough to any remotely knowledgeable reader that the tools that the pension funds are deploying are far from bleeding edge. For instance:
Some public pensions say computer models can help manage their complex portfolios and predict how their assets will behave in different economic environments. Grouping investments by risk rather than type, they said, exposes dangers that might otherwise remain hidden.
“You can look at your portfolio and say ‘Oh wow, I’ve got a lot more inflation risk than I should have or a lot more credit risk than I should have,” said Robert T. Bass, a BlackRock Inc. managing director, who provides software that conducts this analysis for pension funds. “Maybe I should take that down a little.’”
In Fairfax County, Va., new software programs allow public pension fund managers to plug their assets into a computer and ask what they should buy if, say, they want better protection against inflation.
These are old sell-side analytics, and the sell side would often give it to the buy side to deepen relationships circa the early 1990s.
But despite the “gee whiz” tone of the article, the CalPERS part is all about bean counting:
In California, Calpers turned to computer models to understand its private-equity costs…
Calpers was long unable to separate one set of fees from the other, relying in part on a set of spreadsheets to keep track of the data. The information was also stored in a range of different formats, making it difficult to aggregate and analyze.
It took five years to develop a new data-collection system that relies on private-equity managers to fill out new templates describing their various fees. A data and accounting firm then compiles the information and feeds it into the software program.
What CalPERS will never confess to is that the reason it was having such a terrible time before was it had done a lousy job of organizing its data. Even in the few relatively few Public Records Act requests I have made, it’s been clear, for instance, that CalPERS too often was not using the same name for a private equity fund in various databases (it also used State Street’s Private Edge). This is a rookie level error. I’ve also found errors and inconsistencies over time in fund names in CalPERS’ Comprehensive Annual Financial Report. Not having the same names for some funds would make extracting or compiling data across funds a heavily manual task. So implementing the PEARS system hopefully involved a lot of data scrubbing, which may explain why it’s taken so long to become operational.
But the bigger point is that Sutton and the Journal were onto the same issue, which is one we separately flagged (we were a bit late to the party because we wanted to see what CalPERS staff and board actually said at the May 15 Investment Committee meeting about the documents; it looks like the keen interest in CalPERS means we’ll have to move faster).
Despite two board candidates and a respected private equity reporter alerting staff and the board that the effort to cook the books was a tad too obvious, CalPERS dug in until the big guns from the Journal showed up. CalPERS beneficiaries and California taxpayers should be concerned. This incident provides yet more proof that CalPERS thinks it is perfectly legitimate to present misleading information as long as it isn’t caught out in widely-read media.
1 No, CalPERS, it is not who you think it is.