What’s the best way to measure separate account performance?

Private equity separate accounts often feature bespoke investment strategies and reduced fees, making them a popular option for limited partners like the California Public Employees’ Retirement System and New Jersey Division of Investment. LPs committed $127 billion to separate accounts between 2010 and 2014, according to Preqin.

But while many LPs consider separate accounts useful for enhancing returns, their complexity can confound attempts to assess their performance. For example, as of June 30, more than half of New Jersey’s separate accounts failed to generate a top-quartile return multiple for their respective vintage years, according to a Buyouts analysis of Cambridge Associates and New Jersey data.

As a matter of policy, New Jersey only selects funds with a demonstrated record of top-quartile performance. The $74 billion pension system measures its portfolio’s performance against a Cambridge Associates private equity benchmark.

“The determination of whether an investment fund or investment manager is top quartile is typically more complex than comparing the Total Value in Paid in Capital Multiple (TVPI) to a broad benchmark such as the Cambridge Global Buyout & Growth Equity,” said New Jersey spokesman Chris Santarelli in an email.

New Jersey tries to use benchmarks or indices that reflect each fund’s strategy, return objective and risk profile, he said. Santarelli did not provide detail on the benchmarks used for the pension’s private equity separate accounts.

Netting a top-quartile buyout return as dictated by a private equity benchmark may be impossible for many separate accounts. Many pensions group these accounts in their private equity allocations, even if they invest in a blend of strategies that yield a significantly smaller return.

Blackstone Group, for instance, invests New Jersey’s $330 million account with its Tactical Opportunities platform in assets like real estate, credit and hedge fund strategies, along with traditional private equity. As such, finding a proper benchmark to assess performance of many separate accounts can be problematic, sources tell Buyouts.

“It’s not an easy answer, right?” said one LP at an insurance company. “You’ve got five different PMEs [public market equivalent benchmarks], and then the peer ones are really complicated.”

To be sure, many separate accounts are strong performers. One of New Jersey’s separate accounts with TPG netted a 1.46x multiple as of June 30, which ranks the fund among Cambridge Associates’ top earners for the 2012 vintage class. Other New Jersey accounts managed by Blackstone Tactical Opportunities, Asia Alternatives and Siguler Guff also generated top-quartile returns.

Similarly, CalPERS’ $800 million commitment to Blackstone’s Tactical Opportunities platform, a 2012 vintage, netted a 16.7 percent internal rate of return as of June 30, according to the pension’s website. That IRR puts the account firmly in Cambridge Associates’ top quartile for global buyout and growth equity funds.

CalPERS’ Tactical Opportunities fund also matched the three-year return generated by the retirement system’s private equity benchmark, a blended FTSE Russell public equity index plus 3 percent. Even so, measuring the performance of a separate account against a traditional private equity benchmark may not work, sources said.

Jeffrey Bunder, global private equity leader for Ernst & Young, said he believes many LPs pull apart their separate accounts by asset class, which makes it easier to compare them to a more relevant benchmark. It’s impossible to benchmark accurately unless you know the composition of the account’s underlying assets.

One GP who manages separate accounts said his team creates a composite benchmark with underlying benchmarks for the asset class for each of its funds. The fund’s asset mix determines how much each underlying benchmark counts toward the composite. For example, if 40 percent of a separate account is in credit strategies, the firm uses a credit benchmark to account for 40 percent of the composite.

That approach is far from perfect, however, the GP said, noting that a manager can game its returns against the benchmark by blurring the lines between asset classes. For example, an equity stake in an energy asset may wind up in the separate account’s pool of infrastructure assets, which have a lower return profile. The energy asset boosts the return of the infrastructure pool, improving its overall performance against the underlying benchmark.

“It’s a bit artificial,” the GP said, adding that his team targets a net absolute return of around 15 percent. “You’re not going to see us in our letters saying, ‘We beat our blended benchmark by so many basis points.’”

LPs often judge underlying assets similarly, sources said.

“If you have sub-strategies in a product, then those component pieces should be evaluated against similar risk opportunities,” said Irwin Loud of fund-of-funds Muller & Monroe Asset Management. Absolute performance is also important to consider with separate accounts. Beating a benchmark means very little if LPs don’t clear the actuarial rate of return needed to meet their obligations. “Part of it’s relative and part of it’s absolute,” Loud said.

Opening a separate account carries a certain amount risk. Given the size and the scope of their strategies, LPs must consider how a potential account’s returns will stack up against fund investments offering similar exposure, sources said. But LPs must also take into account that fund investments often do not come with the same perks they can get with a separate account.

Discounts on fees and carried interest rate, along with improved access to co-investments, provide serious economic benefits to LPs trying to reduce the costs of their portfolios. For those reasons alone, separate accounts can be a boon to investors.

It is important for an LP to ask: “What strategic benefits am I getting beyond investment dollars put in the ground?” said a placement agent. The agent noted that discounted fees, discounted carry and access to a firm’s resources and investment team benefit LPs like public pensions in the long term. That said, “if they don’t do well, the discount on fee and carry doesn’t mean very much.”

Action Items: See Cambridge Associates’ performance data here: www.cambridgeassociates.com. Click this link to download New Jersey separate acount performance data: NJ separate accounts.

Photo of child measuring self courtesy of Shutterstock