Few recent trends in private markets have attracted as much attention as GP stake sales. As shown in the table below, of the 67 private stake sales and seven IPOs completed since 1998, 64% have taken place since 2015 and 47% since 2017. All signs suggest activity will continue apace, and the three largest investors in such deals are in the market this year seeking a combined $17 billion.1 No wonder people are talking!
# of Transactions by Year As of 12/31/18
The investment strategy has generated much interest and debate. Is there enough supply left to support current fundraising? Can stake buyers exit these investments? For now, we will let those debates rage on. Another point of interest, though, is how these transactions may influence GP behavior. Are these watershed moments, triggering aggressive fundraising and deployment so that GPs can generate larger streams of revenue for their new partners? Are these GPs “selling out” and going big to the detriment of the LPs who got them to where they are?
We have heard many GPs respond to such questions, and while this may not be an exact quote, they sound an awfully lot like the immortal words of Jennifer Lopez (or whoever actually wrote the song):
Don't be fooled by the rocks that I got
I'm still, I'm still Jenny from the block
Used to have a little, now I have a lot
No matter where I go, I know where I came from
Are these GPs still “Jenny from the block?” What does change, if anything, when a GP sells a stake to private investors or in an IPO? If you know us, then you know what’s coming next: data.
Despite all the chatter on this topic, it is rare to hear perspectives grounded in quantitative analysis. It is just hard to get robust data on something like this. Due to the scale of our platform, though, we have data for nearly every GP that has sold a stake. This puts us in a unique position to explore the issue.
Scaling the Platform
Let’s dig in further. We have some interesting data on GP fundraising behavior before and after these transactions—the chart below compares average buyout fund size step-ups in a given year (orange line) with the step-ups of all funds that came either before (dark blue) or after (light blue) a stake sale. As shown, selling GPs tended to achieve above-average fund size increases before the transaction—an attribute likely valued by GP-stake investors. However, the post-transaction picture is less clear. Megafund GPs that went public around the Great Recession had material step-downs in 2011-2014, while post-transaction step-ups in 2015-2018 have been in line with or above market.
Buyout Fund Step-Up by Year of Fund Close % Change from Prior Fund (Median)
Another way to assess fundraising behavior is time between funds, and this tells a similar story. The chart below shows that GPs’ pre-transaction funds were often raised at shorter intervals than those of their peers—again, a favorable attribute to GP-stake investors. And once again, the pattern is less clear for post-transaction funds. The 2011-2014 period featured public mega GPs taking a long time to deploy large pre-crisis funds, while the differences from peers in 2015-2018 are mixed.
Time Between Buyout Funds by Year of Fund Close Median Time for Previous Fund in Years
Does deployment tell the same story? If GPs are ramping up fund size and frequency before a stake sale, then we should see increasing investment to support that. The chart below depicting year-over-year change in flagship deployment shows that, indeed, GPs increased flagship fund deployment significantly beginning two years before their stake sale, with year-over-year increases until the year following the stake sale. The chart below illustrating flagship deployment change vs. peers shows that this is not just a function of market timing in our data set; the pattern remains even compared to peers’ deployment in the same years. Interestingly, deployment trailed off in years two and three after the transaction, in both absolute and relative terms.2
YOY CHANGE IN FLAGSHIP DEPLOYMENT MEDIAN FOR THE YEARS BEFORE AND AFTER TRANSACTION
Flagship Deployment Change Versus Peers Median YoY change in excess of respective asset class
Of course, ramping up fundraising and deployment for a given strategy is not the only way to boost assets under management—far from it. Through April 2019, 50% of stake-selling GPs had added strategies to their platform post-transaction. This is high considering that many of the stake sales happened so recently; the percentage is likely to rise materially as platform expansion plays out for the 2015-2018 transactions. Credit has been by far the most popular new product, representing more than 40% of all strategies added. Others include small/mid-cap buyout (such as after a flagship buyout fund has moved up-market), venture/growth, real estate and other real assets.
Another interesting observation involves the timing of stake sales. Most stake sales (62%) have occurred within two years of the last flagship fund’s final close. This is a win-win for both the stake sellers and buyers: GPs avoid fielding uncomfortable questions about the stake sale on the fundraising trail, while investors avoid fundraising uncertainty in the first few years of their underwriting case.
And now to the question on everyone’s mind: How do stake sales relate to performance? Not surprisingly, investors in these deals have favored GPs whose pre-transaction funds outperformed peers: 37% of such funds were top quartile in our database, while 27% were second quartile. In other words, nearly two-thirds of these GPs’ prior funds were better than average.3
Of course, the bigger question is whether performance deteriorates after the third-party capital comes in. We hate to disappoint, but the truth is that it is too early to tell. Since the majority of stake sales have occurred since 2015, most post-transaction funds are not seasoned enough for performance to be meaningful. Those post-transaction funds that are seasoned have so far exhibited a more average distribution across quartiles, but again, the sample size is small. Also, it is heavily weighted toward the funds of mega GP franchises that went public around a decade ago – a different profile than many of today’s stake-sellers. So, just like J. Lo’s highly anticipated halftime performance this winter, we will simply need to stay tuned.
Where does this leave us? By various metrics – fund sizes, fund frequency, deployment – it is clear these GPs have left behind their “Jenny from the block” days. But the idea of GP stake sales being the watershed moment is not supported by our data. The noteworthy ramp-up in deployment and assets under management seems to be occurring before the stake sale, not after.
Should this be surprising? Perhaps not. GP-stake investors would prefer firms that have already started demonstrating their ability to scale. This helps to explain why so many of these transactions happen soon after above-average step-ups in flagship fund size. Rather than causing GPs to change, it may be that stake sales follow changes that are already in motion.
Mind you, we are not saying these transactions have no influence at all on GP behavior. As noted, the jury is still out on performance post-transaction. And even though the pace of scaling may be leveling off post-transaction for flagship strategies, product expansion can be a powerful driver of AUM growth. It will be interesting to see how this plays out for the 2015-2018 stake-sale cohort, which is still too fresh to draw conclusions.
Even if GPs have already left their proverbial Bronx beginnings for a Manhattan high-rise by the time they sell a stake, the question remains whether an LP should stick with them. Just as with Jenny becoming J. Lo, LPs can debate whether the world is better off for these GPs having made it to the “big leagues.” However, the answer very often depends on the LP and GP involved. But whatever one’s opinion on the evolution, they shouldn’t assume a stake sale was the primary catalyst for it.
Endnotes1Source: “Raising the GP Stakes,” PitchBook, June 12, 2019.
2These same general patterns are evident using firm-wide deployment (rather than just flagship), but restricting to flagship is more reliable due to noise with ramp-up and wind-down of non-core strategies.
3Source: Hamilton Lane Data (June 2019)