Friday Letter: Track record and its prescriptive limitations

LPs are united in their quest for top quality managers. But is there really any way to predict which GPs will rise to the top, or is it all just a guessing game?

If you ask most LPs what they look for when selecting a GP, you'll likely find that “track record” comes pretty close to the top of the list. For most it makes sense. If you've had a successful fund before then you clearly know how to pick decent investments, you clearly know how to effect operational change, and you clearly know how to create good returns on exit.
Or do you?
At PEI's Hong Kong Forum this week, Thomas Kubr, the executive chairman of Capital Dynamics, left delegates slightly taken aback when he said that track record counts for… exactly nothing. 
“The relevance of past performance is completely not there. It doesn't matter,” Kubr said. “The person that generated the past performance is not the person that is going to work your money in the future.”
It may be physically the same person, but that manager has been through some fundamental life and career changes, Kubr said. No longer the hungry first-time fund manager, he's now rich and has likely lost his lust for the day job.
“This makes this job of really finding good GPs incredibly difficult because the history that they've done is simply irrelevant to my question of will they be a good steward of my investment going forward.”

Jie Gong, a partner at fund of funds Pantheon Ventures, agreed that “past performance does not predict future success”. However, she posited some methods for determining the potential success of a given vintage in a particular market.
Gong's team uses three lenses to assess a particular pocket of a given market. The first is enduring trends; examples in Asia would be things like urbanisation and productivity gain, which she classifies as “multi-decade developments”. The second is business cyclicality, so things like government tightening and easing, interest rates and other factors that have an effect on the public markets, a key exit route for many private equity investments in Asia. 
The third variable Gong refers to as the “liquidity cyclicality”, which is dictated by the supply and demand of capital into a given market. This affects valuations, which is a “leading indicator to the performance of a given vintage”.
“The swathe of capital is always the enemy of performance, and no GP regardless of skill, is able to triumph over that if the overall market valuation is extremely, extremely elevated,” Gong said.
But for Kubr, this too is all academic. At the end of the day, you just can't predict the future.
Of Gong's lenses he said: “You can use them to look backwards and understand why a certain vintage worked better than the other, but trying to be prescriptive about this, when to invest where, we find extremely challenging.”
Kubr said if you were to look at any large data set of private equity fund performance over a 20-year period, categorised by geography, asset class, strategy or any other criteria and then ranked each vintage year from best to worst by numbers, you would discover that there's no pattern.
“It's completely random. And this randomness in past performance is a pretty good indicator of just how impossible it is to make this prescriptive,” he said. 
There's one thing that every LP will surely agree on: a good manager will make you money, even in sub-par conditions, and a bad manager will lose you money even in fair weather.
The million (or in some cases billion) dollar question is, therefore, how do you pick a good manager?
Kubr's answer is that you shouldn't try. Instead, LPs should focus on avoiding the bad. 
“If LPs would spend more time rooting out the screaming inefficiencies and deficiencies of GPs and just get rid of those, and then just randomly pick from what's left over, their performance would be naturally five percent over the average,” he said. “Frankly anyone that tells me that they can really pick the future top quartile or second quartile, they're dreaming.”
LPs are all on the hunt for quality managers, but that quality comes in many shapes and sizes, and will only reveal itself fully ten years down the road – by which time the deed is clearly done.   
So until LPs find the magic formula for predicting the future, they'll have to keep analysing – and guessing – their way to success.