A trio of CalPERS investment-committee members this month objected to a plan to create two captive private equity funds, and the experience of sister state pension Florida State Board of Administration suggests they’re right to have concerns. Such programs don’t always work out as advertised.
Under the proposed plan, approved in concept by the committee in a 10-3 vote, the $359 billion California Public Employees’ Retirement System would commit up to $20 billion to two captive vehicles. One would be earmarked for late-stage venture and growth investments, the other for buy-and-hold stakes in mature companies. According to CIO Ben Meng, CalPERS seeks more control over private investments at a lower cost and at a bigger scale than it gets with conventional funds today.
One of the board members opposed to the concept, California Controller Betty Yee, questioned whether the program would meaningfully lift the pension’s returns. Florida SBA’s experience with a series of seven captive funds managed by New York buyout shop Liberty Partners suggests that concern is valid. The Florida SBA experience also shows the importance of negotiating favorable terms and conditions with the investment teams.
The Liberty Partners program, crafted in the early 1990s by Florida SBA Executive Director and CIO Ashbel C. Williams Jr, started out promisingly enough. Liberty Partners Pool I, to which the state committed just over $205 million in 1993, generated a 2.35x investment multiple and 20.7 percent net IRR, a third-quarter 2018 investment report shows. The successor, launched the following year with a $360 million commitment, generated a 1.61x multiple and 10.7 percent net IRR. From there the Liberty Partners team’s performance deteriorated rapidly. Three of five success pools lost money for the state:
- Liberty Partners Pool III, launched in 1996 with a $506 million commitment, achieved a just-over-break-even multiple of 1.02x and a modest net IRR of 0.4 percent;
- Liberty Partners Pool IV, launched in 1998 with a $195 million commitment, lost a third of invested capital in generating a -19.2 percent net IRR;
- Liberty Partners Pools V, VI and VII, to which Florida committed more than $1.2 billion from 1999 to 2005, posted respective net IRRs of 2.7 percent, -6.6 percent and -7.3 percent.
Because Liberty Partners could recycle some of its distributions into fresh investments, the total dollars involved are prodigious. Florida SBA committed just under $2.5 billion to the seven funds, but Liberty Partners actually drew down $3.2 billion over more than a decade.
All told, between the $3.5 billion distributed and the $9 million in remaining NAV, Liberty Partners ended up generating a 1.08x investment multiple overall. And that slim gain drags on the overall performance of the $11 billion-plus PE portfolio. In the 15 years ended June 30, the portfolio generated an annual return of 10.4 percent, making it Florida SBA’s best-performing asset class, though still underperforming its benchmark by 1.87 percentage points.
So what accounts for the lackluster performance? I wasn’t able to reach the original principals of Liberty Partners, founded in 1992 and led early on by three former executives of Merrill Lynch Interfunding — Peter Bennett, Paul Huston and Michael Kluger (who went on to co-found healthcare specialist Altaris Capital Partners) — and one from Lehman Brothers, Carl Ring.
A Florida SBA spokesperson was unable to answer a set of emailed questions, although he wrote that the “SBA’s investment … was a one-off agreement, not an implementation of an internal program, and using the Liberty investment as a ‘cautionary tale’ for CalPERS falsely suggests the two are similar.” A spokeswoman for CalPERS emailed that the pension fund declined comment on the Liberty Partners program.
A 2010 feature story by Kris Hundley in the St Petersburg Times, now called the Tampa Bay Times, describes investment missteps by Liberty, including a 1998 investment in Norwood Promotional Products, which later filed for bankruptcy protection. The firm also acquired charter-school company Edison Schools in a controversial 2003 deal that stirred up resentment among members of the state teachers union.
A source familiar with the captive program said that on at least two of the early pools Liberty Partners charged management fees and carried interest far below the standard 2 and 20. That would suggest the program made good on one big promise of captive programs: that they reduce costs.
But not all the terms favored the state. A 2003 report commissioned by pension consultant Alignment Capital (a copy of which was obtained by Buyouts) shows the executives of Liberty took profits on good deals but — unlike most PE funds — didn’t have to offset those with losses on bad ones. That term cost Florida SBA $61 million in lost proceeds assuming 15 percent carried interest and an 8 percent preferred return, according to the report. Executives also kept all fees charged to portfolio companies. According to the St Petersburg Times article, Florida SBA did renegotiate more favorable terms for Liberty Partners investments made post-2005.
To be sure, Florida SBA struck its deal with Liberty more than 25 years ago. A savvy CalPERS would presumably not agree to such investment-team-friendly terms today. CalPERS also has plenty of more modern examples of captive programs to draw upon, including a much more successful co-investment program started in 1998 for Florida SBA by Lexington Partners. (Lexington later opened the program to other investors.)
Still, CalPERS would be wise to get to the bottom of why Liberty Partners performed so poorly for Florida SBA. It’s a story that has yet to be fully told.