The Only Mistake Henry Kravis Made Twice

Henry Kravis (unintentionally) makes the case for some investment chiefs losing their jobs.


“People don’t change?” I asked. “People don’t change much,” Henry Kravis answered.

Our hourlong conversation was winding down. Kravis, who needs no introduction, had sat for a filmed interview with Institutional Investor in Chelsea, the art gallery and brunching capital of Manhattan. We had bantered (Kravis is a world-class banterer) over his war stories — the early, heady days of Kohlberg Kravis Roberts before it became KOHLBERG KRAVIS ROBERTS — while playing gin rummy (his choice).

As filming wrapped and talk turned to his art — making companies better — he shared the Kravis method: Inherit great people, or hire new great people. The only mistake Henry Kravis has ever made twice, or so he claimed, was waiting for people to change.

The firm has paid off its institutional investors handsomely — the first, he notes, was the State of Oregon Public Employees Retirement System, whose initial $5 million investment in 1979 has blossomed into a recent $1.5 billion allocation. Private equity limited partners, like portfolio company executives, tend to act on repeat. Bad allocators remain bad, research suggests. Like Kravis, they repeat that mistake.

A 2016 paper — “Measuring Institutional Investors’ Skill from Their Investments in Private Equity,” by four academics from California State University Fullerton and Ohio State University — delivers what its title suggests. The conclusion: Picking outperforming managers is a skill, and one that persists. “Some LPs consistently invest in the top half of funds while some are consistently in the bottom half of funds,” the authors write, adding that the effect is “even larger for investments in venture capital funds.”


Their findings destroy two common defenses of underperforming LPs: variance in risk appetites and lack of access. Among investors of similar types — and, presumably, similar risk tolerances — good and bad allocators remain thus. Access is rarely an issue with first-time fundraisers, yet “higher-quality LPs tend to outperform in first-time funds by about the same amount as they do in their investments in funds from established partnerships.” In Kravisian terms, people don’t change — and neither do limited partners.

The idea that very good and very bad LPs exist was still top of mind when Britt Harris — longtime CIO of the Teacher Retirement System of Texas and, in Kravis’s opinion, a very good LP — arrived for his own video interview. (Before the cameras began rolling, Kravis lauded Harris’s “intelligence and pioneering spirit,” echoing many of Harris’s peers.) The Texan and I traded industry gossip while our mikes remained cold, primarily about his home state. The University of Texas Investment Management Co. (UTIMCO) had recently parted ways with CEO and CIO Bruce Zimmerman after nearly a decade. My sources attributed the split to average returns and friction among Zimmerman, his staff, and the board. I wondered, had David Barrett, the uber-recruiter conducting the search, called Harris for the job?

He demurred. “There seems to be a lot of turnover in the endowment world,” he said, transitioning topics to a novel fee model he’s pioneering with Albourne Partners, an alternatives consultancy. (For those wondering about this “1 and 30” model, a sizable majority of the hedge funds approached have accepted it, and predictably so given the buying power of Texas Teachers.)

On turnover, Harris was right: Nonprofits underwent significant leadership churn over the past year. According to an Institutional Investor study, 23 percent of top-level CIOs had left their roles in the twelve months ended in February — absurdly high for an industry that rewards CIOs with near-tenured employment. (David Swensen has been on the job since the middle years of the Reagan administration.)

Were boards finally reckoning with the uncomfortable fact that they can’t all have “good” CIOs/LPs? Had those with years of mediocre returns decided it was time to be much more aggressive in their staff selection?

“It’s slightly more nuanced than that,” Barrett explained over the phone. (He would only chuckle when asked about leading candidates for the UTIMCO job, noting that he was in month two of a roughly six-month process; the search for the Dartmouth CIO, which his firm is also conducting, is more streamlined.) “It’s more that, coming out of the crisis, there is increasing scrutiny on absolute and relative performance. In that context the investment committees of boards are taking their fiduciary responsibility much more seriously, and they’re taking a much more critical look at the five- and ten-year performance figures.”

The churn occurring now is a natural phase in that timeline, he said. “When someone comes into a position, it can take 18 months to make the moves they want — and on the private side it can take even longer. Five years is really where you can begin to analyze talent.”

Boards became hyperaware of this in the 2008–’10 time frame, and so “2015 is where this has really started to come out in the wash.” And come out it has: While many CIOs are retiring, other departures are not exactly voluntary. Unsurprisingly, Barrett was unwilling to explain which were which.

Research and other industry leaders back up Kravis’s view on the immutability of talent. The same, I hope, applies to gin rummy.

This is very important to note: I beat Henry Kravis — Henry Kravis — at a hand of gin rummy. (As the gregarious victor, it’s only fair to admit that playing gin rummy while fielding interview questions on film is very difficult, and we didn’t technically finish the game.) I will never again play gin rummy. In my obituary the New York Times can write, “In his first and last game of gin rummy, he beat Henry Kravis.” And unlike LPs picking private equity managers, I don’t expect my card skills will persist.

Kip McDaniel is the Editorial Director and Chief Content Officer of Institutional Investor.