The SaaSpocalypse shows private markets need risk models

Investors have little idea how bad the losses in private credit are going to be

Markets in recent months have offered up a reminder of the difficulty of modelling risk in private assets.

Advances in artificial intelligence  that triggered a selloff in software companies in January and February – the so-called SaaSpocalypse – are likely to hurt private funds more than many investors expected.

Pension funds, sovereign wealth funds and endowments have invested heavily in private assets in recent years. Total allocations grew from around $14 trillion in 2020 to about $24 trillion in 2025, according to estimates from McKinsey & Company.

These investors are now wondering “exactly how much bad news is coming their way”, says Peter Shepard, who heads research and development in MSCI’s analytics business.

It could be months before anyone really knows. 

Private funds report performance only quarterly and often with up to a 60-day lag. Information that reaches investors in May might reflect asset marks from the previous December. Those valuations, in turn, could be based on company financials from even earlier.

Reported performance also depends on valuation assumptions that are unavoidably subjective. Since the early 2000s investors have known that reporting tends to be smoothed, which is to say it often takes time to fully reflect changes in the economy. The smoothing upends the calculation of correlations to other investments in an investor’s portfolio. 

Traditionally you would look at the private companies and think they have similar factor exposures and correlations to large tech companies

Brad Dunstan, New Zealand Superannuation Fund

Data that tells investors about the underlying risks in private credit has been notoriously difficult to get hold of, and equally difficult to interpret. 

Investors know the same broad drivers that move public stock markets move private equity – growth, interest rates, industry advances, and so on. The same factors that affect loan markets and private credit. 

And yet questions as straightforward as the beta of a private investment to the market or to the rest of a portfolio become “extremely difficult” to answer, says Shepard. 

A common solution has been to match private investments to public proxies to gain a more contemporaneous picture of the risks a fund owns. Funds that do this worry, however, that private investments and proxies might turn out a bad match. 

Brad Dunstan, co-chief investment officer of the $90 billion New Zealand Superannuation Fund, which allocates less than 5% to private equity, describes the challenge. “Traditionally you would look at the private companies and think they have similar factor exposures and correlations to large tech companies that are now booming because of AI,” Dunstan says. “But one group sinks through the floor and the other shoots to the moon.”

That seems likely to be the case with software companies. 

MSCI’s private capital data shows that private equity buyout funds are heavily tilted towards software investments compared with the public market. Software, within the IT sector, represents nearly a fifth of these funds’ holding-level net asset value compared with roughly 5% of the market capitalisation of the MSCI world index. 

Related

The race to model private market risks 

BlackRock maps holdings to risk factors; competitors aim to get the best from statistical methods

What’s more, the private software companies in MSCI’s data have borrowed about five-times Ebitda on average versus four-times for non-software companies. Since the start of 2023, the average company in the group has paid more than half its Ebitda in interest payments.

The prices funds pay for these companies have been rising too. Median entry valuation multiples for the cohort climbed from about 13-times Ebitda in the years before 2019 to 21-times Ebitda in 2025.

High multiples, high leverage and low interest coverage mean funds might find it harder to exit such investments going forward, dragging down distributions. 

Allocations to tech-heavy private equity in favour of public stocks might start to look like a mistake.

That all points to the need for better data and analytics on private fund investments, and explains why BlackRock, MSCI, SimCorp and Venn are racing to advance private asset risk models. 

The SaaSpocalypse seems likely to help make the case for improvement. 

Editing by Kris Devasabai