FORTUNE — When private equity firms raise a new fund, they usually tell prospective investors to look at the interim returns of their most recent fund. After all, what better indicator of future performance could there be than the recent performance of the current team?
Unfortunately, however, it seems that most of them are using the wrong numbers to tell their predictive story.
That’s the finding of Landmark Partners, a private equity and secondaries fund manager, in a new research paper titled Searching for Outperformers.
In short, Landmark finds that fund managers typically offer up two basic pieces of data: Internal rate of return (IRR) and total value to paid in (TVPI), and then benchmark them against other funds raised the same year (known in private equity circles as “vintage”). Unfortunately, those metrics “are not strong predictors of final values and performance of the current fund.”
For buyout funds, Landmark determined that it takes until the first close of the successor fund before the interim IRR of the current fund has a statistically significant relationship to its final value, while for venture capital funds it takes until two quarters into the successor fund’s life.
Part of the problem is the inherently imprecise nature of interim valuations, while another factor is that not all funds of the same vintage have the same percentage of public market exposure. There also have been allegations of book-cooking timed to fundraising, but Landmark doesn’t address the issue.
So what should investors do, other than try to get their commitment in as late as possible?
Landmark suggests using a Public Market Equivalent (PME) metric. This continues to utilize both NAV and cash flow, but also adds public market returns into the mix.
“The idea of PME is that it looks at what happened in the public equity market between the times of drawdowns and distributions, and normalizes a fund’s returns to remove the effect of these general market movements,” Landmark explains.
The interim PME shows a 95% correlation to a buyout fund’s final value beginning one year before the start of a new fund, and hits 99% beginning two quarters before the new fund (and not regressing).
It is worth noting, however, that a similar analysis doesn’t really work for VC funds. There is no systemic significance until two quarters after the start of a new VC fund, at which point it hits 90%: “The most likely reason is the high level of idiosyncratic, or company-level, risk carried by venture investments and, perhaps, to a certain extent the higher rate of unsuccessful investments that randomly surface in the later stage of a venture fund’s life.”
Sign up for my daily email newsletter on deals and deal-makers: GetTermSheet.com