Why Private Equity's Software Bet Is Going to Zero | The Weekly Wrap

Summary

Steve Eisman opens with a warning that markets feel stressed as the 10-year Treasury sits near 5.3%, then turns to FICO’s sharp selloff after FHFA head Bill Pulte moved to equalize FICO and VantageScore treatment in mortgage underwriting. He argues that FICO’s decades-long mortgage monopoly was effectively broken after years of aggressive price increases and regulator backlash. The main segment is a deep technical critique of private equity’s software exposure: SaaS looked ideal for leverage because revenues were predictable and easy to model, which drove massive buyout activity concentrated among a few specialist firms. Eisman says higher rates, variable-rate debt, and the “SaaS apocalypse” have compressed valuations, and that many 2027 refinancings could leave equity worthless if enterprise values have fallen to or below debt. He closes by arguing the Fed cannot directly offset oil-driven inflation and by explaining how Moody’s securitization database gave him the conviction to short subprime in 2006.

Notes

• The episode opens on a macro stress backdrop: the 10-year Treasury is around 5.3%, a level Eisman says has not been seen since roughly 2002.
• He frames the week as one where something may be close to breaking in markets.
• FICO had a very bad week, in his view, because the stock was hit hard after a regulatory change.
• He says FICO stock was down over 25% on Tuesday and over 60% year to date.
• The core catalyst was Bill Pulte, head of the FHFA, posting that the LLPA grids for FICO and VantageScore would be combined.
• That change removes the 20-point haircut on VantageScore and improves its competitive position.
• Eisman says the episode validates the idea that one should not antagonize the regulator.
• He says FICO raised prices 1,600% over the last five years.
• He argues FICO lost its mortgage monopoly in less than a year after years of pricing power.
• The software section begins with the claim that private equity concentrated heavily in SaaS because the business model was predictable.
• He says SaaS revenues were attractive because they were predictable, high-margin, and high-growth.
• He claims 90% of tech buyouts were SaaS companies.
• He says SaaS deal volume represented about a quarter of the entire private equity buyout market.
• He says specialist funds such as Thomas Bravo, Vista Equity, and Silverlake accounted for about 60% of total SaaS buyout activity.
• From 2017 to 2023, he says private equity transaction volume ranged from about $100 billion to $400 billion a year.
• He identifies 2021 as the peak year for U.S. technology buyouts at more than $284 billion.
• Of that 2021 total, he says $256 billion was software.
• He says the SaaS model is simple to underwrite because revenue can be modeled from price times customers, then layered with growth and price assumptions.
• That modeling simplicity, in his telling, encouraged aggressive leverage.
• He says higher rates have already hit earnings and cash flow because these deals were financed with variable-rate debt.
• He says the SaaS apocalypse has already compressed valuations sharply.
• He argues the original AI story was that AI would wipe out all existing software.
• He says that narrative briefly pushed software stocks down before a later rally.
• In his updated view, the SaaS apocalypse story did not disappear; it just changed shape.
• He presents Citrix as a leveraged software buyout with a $16.5 billion purchase price.
• Citrix was financed with $8.55 billion in syndicated bank loans and junk bonds.
• He calculates that Citrix was 52% debt-funded.
• McAfee is another example: a $14 billion purchase financed with $8.96 billion in junk bonds and levered loans.
• He says McAfee was 64% debt-funded.
• Zendesk was bought for $10.2 billion and financed with $5 billion in private credit loans led by Blackstone Credit.
• He says Zendesk was about 49% debt-funded.
• Proofpoint was purchased for $12.3 billion and financed with a $4.6 billion syndicated first-lien loan.
• He says Proofpoint was 37% debt-funded.
• Anaplan was bought for $10.4 billion with $2.5 billion of debt.
• He says Anaplan was only 24% debt-funded.
• He concludes that three of the top five transactions look vulnerable to AI intrusion.
• He says the broader sample of software buyouts averaged roughly 50% debt financing.
• He argues that many software LBOs were built on a debt-heavy structure that becomes fragile when valuations fall.
• He models a 2027 refinancing problem using a software company bought for $1 billion with 50% variable-rate private credit debt.
• He says corporate debt is unlike mortgages because principal is not amortized monthly and must be refinanced at maturity.
• If enterprise value falls to the debt level, he says equity becomes worthless before the refinance.
• He thinks many of these software investments are effectively dead money.
• He says the private equity monetization cycle has stretched from 3 to 4 years to about 7 years because IPO markets have been weak.
• He warns that many investments may eventually be marked to zero.
• He cites Pluralsight as a 2024 example where lenders took control and Vista’s equity was wiped out.
• He cites Medallia as another lender takeover in August that wiped out equity.
• He uses those cases to show that lender takeovers are not limited to cybersecurity names.
• In the mailbag, he says the Fed has zero control over oil prices.
• He argues rate hikes do not change the immediate inflation impulse from oil.
• He says oil can transmit into fertilizer, food, and the broader economy, which eventually forces the Fed to respond anyway.
• He recalls using Moody’s securitization database to compare delinquency trajectories across vintages.
• He says the database contained monthly credit metrics for every securitization.
• He says 2006-vintage loans were delinquent much faster than older securitizations by summer 2006.