Tech valuations and the AI reality check
For several years, the private equity industry allocated significant capital to high-multiple Software-as-a-Service (SaaS) and technology companies, highly valued for their recurring revenue models. However, rapid advancements in artificial intelligence (AI) have introduced new variables, prompting a reassessment of long-term growth and valuation metrics.
“Recurring revenue is not what it used to be. If you paid higher multiples on a tech company thinking that its revenue stream is consistent and growing rapidly, the problem you are having now is that with AI it’s turned that model completely on its head. Many of your customers may no longer renew with you, or if they do, they may do it for a lower price so what you have in your portfolio is harder to value,” says Nitin.
This environment has led to longer holding periods, as managers wait for market conditions to clarify. “You will have many private equity funds that delved into tech having longer holds. More importantly, it’s become harder to ascertain if the valuation of their portfolio is what they say it is until they actually test the market,” Nitin observes. “You actually need a buyer out there who is willing to take a risk on a tech company today, and there is just so much fear ingrained from AI and potential disruption that it is harder to sell those companies and harder to believe that they would be sold for the same multiple as what was paid at entry.”
At Flexstone, we’ve always shied away from tech – our exposure is under 10% and even that is focused primarily on services rather than pure software. “It wasn’t because we saw this AI risk coming – it’s not some crystal ball – but mostly just because we have shied away from paying higher multiples and simply chasing the next hot sector,” Nitin explains.
Assessing performance standards in private equity
With overall distributions and liquidity slowing across the private equity landscape, some large-cap managers have used financial tools – such as dividend recaps or subscription lines – to optimise cash flows and distributions to Limited Partners (LPs).
Nitin emphasises that the lower mid-market relies on a different set of levers. “In the lower end of the market, you are still having to generate a lot of returns through increasing the EBITDA of these businesses and adding more value to these companies,” he says.
This direct correlation to operations makes mid-market performance easier to benchmark. “At the larger end, you are seeing greater leverage and financial engineering, whether that is a bit more on dividend recaps or potentially leveraging those companies even more. Whereas on the mid-market, leverage tends to be a bit more conservative, so you are not seeing as much IRR (Internal Rate of Return) manipulation,” Nitin notes.
He points out that while some large buyout deals feature leverage multiples of six or seven times EBITDA, the lower mid-market generally remains disciplined at four to five times. “IRRs can be manipulated by funds, whether it is by credit lines or by selling some stuff early and then holding on to the rest. But if you look at actual cash in and out and actual cash return, that is much harder to manipulate,” Nitin explains. “That is where we are seeing investors increasingly benchmarking funds based on TVPI and DPI*, which I think is the right measure: because it is harder to manipulate.”
The scale of tech infrastructure spending
The divergence between speculative growth and operational investing is visible in the massive capital expenditure (CapEx) currently being deployed by the “Magnificent Seven” and other tech leaders to build out AI infrastructure and data centers.
Nitin shares a striking historical comparison to contextualise this spend: “If you stack up all the CapEx (infrastructure and investment) spend as a percent of US GDP through the years including the 1930s public works, Manhattan Project, electrification, Apollo Project, Broadband, interstate highways, and you add it all up, it is barely above the CapEx that was spent last year alone by the tech firms.”
Despite this historic allocation of capital, the ultimate productivity gains remain to be fully realised. “To date, very few, if any, companies have been actually able to quantify what AI has added to their earnings per share. It is still all in the hypothetical – that it is going to come, it is going to come.”
On data centres specifically, rather than investing directly in their development, Flexstone has participated in this growth through the physical supply chain. “We have some companies that are electrical engineering firms. They provide engineering services, low-voltage engineering services that are going into commercial buildings, hospitals, and more infrastructure-related projects. We also have companies that build wire cages to house different products that go into data centres, so aspects of it are touched upon by different companies,” Nitin explains.
“But that is still a much less speculative bet, because 100% of their business is not data centres; it is just a component of it.”
Succession trends and operational improvement in the mid-market
While potentially speculative technology sectors undergo a period of adjustment, Nitin believes the US lower middle market presents highly attractive, fundamental opportunities – particularly through succession planning in family-owned businesses.
“There are a lot more family-owned businesses (available to buy) today because of the succession issues they have generally,” says Nitin. “Baby boomers are getting older and many of their kids do not want to take over those businesses because they would rather go invest in crypto and other more exciting stuff.”
These robust, established businesses have frequently underinvested in systems, processes, and management and benefit from institutional expertise and digital modernisation, offering private equity managers a clear path to driving genuine value.
“In these smaller, mid-market, family-owned businesses, they are quite behind the curve in terms of implementing AI, developing business KPIs, and internal infrastructure,” Nitin concludes. “They really need a partner to help them steer the business in the right direction and position it for growth particularly in the current challenging macro environment with so much uncertainty. As a result there is greater opportunity from day one to start to add value.”
