A reader called me last week about something that triggered my BS detector: MSCI was building a ‘Private Equity replicator’ indexand Goldman Sachs packaged it into an ETF, ticker GTPE.
Here we go againI thought. Another product to democratize ‘private equity’ for the masses. Another way to collect fees for promising to replicate something that cannot be replicated. Or better: should not be replicated. I’ve seen this movie before and it usually ends with disappointing returns and a lot of explanation.
But then I caught myself. Because I’ve actually invested in other replicators, the trend following stuff from the Return Stacked guys, DBMF, and I also follow Unlimited ETFs closely. So who am I to simply dismiss this without at least kicking the tires?
The premise is not crazy
Here’s MSCI’s pitch, and honestly, it’s more thoughtful than I expected:
“While some of private equity’s historical performance advantage is attributed to the skill of general partners in sourcing, selecting and operating portfolio companies, some can be linked to more measurable exposures such as sector allocation, regional focus and factor characteristics such as size, value or leverage. These systematic elements can be approximated using liquid publicly traded equities.”
Translation: Not all of PE’s outperformance can come from David Rubenstein making brilliant calls. Some of it may just be math. Smaller companies. Supplied balance sheets. Value attributes. Sector tilts. If you have enough data, you might be able to extract the genius and bottle what’s left.
And look, this isn’t some wild new concept. People literally did this to Warren Buffett.
The Buffett Template
Remember that famous AQR paper “Buffett’s Alpha”? Frazzini, Kabiller and Pedersen took apart Berkshire Hathaway’s returns like a mechanic under the hood of a vintage Ferrari. What they found:
- Buffett’s Sharpe ratio was significant compared to the market (but nothing at 0.79). spectacular)
- Much of its performance comes from systematic exposure to factors: value stocks, quality companies with stable earnings, low beta names
- He brilliantly uses leverage through insurance floats
- Even after all these factors are taken into account, there remains some alpha… and that part is pure skill
The conclusion is that you can replicate some from Buffett. Not all of them, not the part where he gets a call about See’s Candies or negotiates the BNSF deal. But the general approach? The patterns? These can be systematized.
Private equity is the same story. Of course, you can’t replicate Apollo’s ability to first introduce Club Med or KKR’s operational playbook for portfolio companies. That’s the quirkiness, the advantage that only comes from relationships, timing and being smart.
But if PE firms consistently lean toward certain sectors, predictably increase leverage, and focus on smaller, cheaper companies with specific characteristics, well… that could be a signal. And signals can be received.
So what’s the catch?
This is from Frazzini&gang paper: “We emphasize again that the ability to explain Buffett’s returns using factors from academic papers written decades after Buffett put them into practice does not make Buffett’s success any less impressive.“
The next thing that baffles me about all of this: Do they still work when you publish the “Buffett factors”?
This is the central paradox of factor investing. Value worked fine for decades… until everyone knew about it? Suddenly it seems that being a value investor means buying the same busy trades as any other value investor, and the premium can be arbitraged away.
The question is not whether PE has systematic components; that is clearly the case. The question is whether these components provide a sustainable advantage or whether they have already been arbitraged away by the time you pack them into an ETF (or whether they weren’t there in the first place).
This is exactly what happens with Unlimited Funds ETFs. Smart people. Good methodology. Mediocre results (so far).
The model
MSCI claims that they combine top-down and bottom-up models to avoid the weaknesses of both approaches. That’s the same framework the guys at Return Stacked use for trend tracking. It’s rigorous. It is backed by research. But rigorous doesn’t always mean profitable.
The model, as I understand it, doesn’t really look at public stock market performance to build its replication. But did MSCI try fifty different models until they found one that beat the World IMI but still fell below the PE Index? Because that is the cynic’s phrase: backtest until you get the result that makes for a good sales pitch.
Or, and this is a more interesting possibility: perhaps the PE index actually provides a decent factor timing model. Country tilts. Sector rotations. A long/short factor sleeve.
And get this: no leverage (!!). The whole outperformance comes from being smart about where you are and when.
If that’s real, if that really works in the future, then this isn’t just a PE replicator. It is a more capital efficient way to allocate stocks in general and factors in particular. That would be really helpful.
The real test
If private equity outperformance was that only driven by managerial skills, survival bias and luck, then this whole exercise is pointless. You’re trying to replicate something that was never replicable in the first place.
But if there is a systematic component, if PE managers follow a process that can be identified and extracted, then this could work as a liquid, transparent and cheaper way to get that exposure.
MSCI has the data. They have decades of PE fund performance and deal-level fundamental data. If anyone can make this happen, he has the raw materials. The question is whether the end product actually delivers.
The fact that they can produce that backtest with a beta anything less than 1, I’m getting very leery of the whole pie, but who knows… At least GS is only charging 50 basis points for the ETF, perhaps the best bargain since Uncle Warren took over Berkshire 😉
What I’m reading now:
Follow me on Bluesky @nprotasoni.bsky.social
#replicate #private #equity #returns #MSCI