The PIK Fuse: How Private Equity Software Deals Like Medallia Actually Blow Up, and Exactly Who’s Next

The PIK Fuse: How These Deals Actually Blow Up, and Exactly Who’s Next

Thoma Bravo handed Medallia to its lenders in April. Blackstone, KKR, Apollo, and Antares took control of a company Thoma Bravo bought for $6.4 billion in 2021, wiping out roughly $5.1 billion in equity held by the firm and its co-investors.

The headline was the $5.1 billion. The number that actually explains it was smaller and quieter: debt service had climbed to about $300 million a year against roughly $200 million in earnings. A company earning $200 million cannot pay $300 million. That gap had existed for a while. The reason nobody saw it until the very end has a name, and it is the most important word in this whole reckoning. PIK.

Medallia is the second giant equity wipeout in 18 months, after Vista’s Pluralsight handover in 2024. It will not be the last. There are a dozen more software leaders PE bought at the peak, loaded with debt, that no longer throw off the cash to service it. What separates the ones that have already blown from the ones that are about to is timing. And the timing is more knowable than it looks, because the fuse runs the same way every time.

Here’s how the fuse works, then the deal-by-deal list of who’s already gone, who’s lit, and who’s on the clock.

How the PIK Fuse Works

Plain leverage is honest. Miss a cash interest payment and everyone knows inside a quarter. The covenant trips, the lender calls, the clock starts.

PIK does the opposite. Payment-in-kind lets the borrower defer cash interest and add it to the principal balance instead. The company looks current on its debt while paying nothing in cash. The balance compounds in the background, getting bigger every quarter at the exact moment the business is getting weaker.

A company on a PIK toggle can clear its covenant tests right up until the deferral window closes. Then it has to service a much larger balance in actual dollars it never had.

The problem does not appear at the end. It compounds the whole way through. The toggle just hides it until the window expires or the loan matures. That is why these deals look fine and then go to zero with almost no warning in between. There was warning. It was upstream, and it was quiet.

PIK does one more thing most operators miss: it degrades the lender too. BDCs, the private credit vehicles holding most of this paper, must distribute 90% of their income. PIK income counts as income. So a fund books the deferred, compounding interest as earnings, reports it in net investment income, then distributes real cash against interest it never received. That works until investors want out.

In Q1 2026, non-traded BDCs posted their first-ever quarter of net outflows. Blackstone’s flagship took ~$3.7 billion in redemption requests, near 8% of NAV, against a 5% cap. Blue Owl funds gated outright.

The Backdrop: $46.9B in Distressed Software Debt

A sense of scale before the names.

Roughly $46.9 billion of software debt is already trading at distressed levels, with about $17.7 billion of US tech loans dropping into distress over a single four-week stretch this year.

PIK now sits in roughly 12.8% of BDC loans, with toggles making up about half of those. Software is roughly 29% of total BDC assets, the most PIK-heavy and most AI-exposed category in the book at the same time. The Financial Stability Board found that PIK toggle usage is tied to a 1 to 2 point jump in the odds of a loan going delinquent the following quarter. PIK is not just a symptom. It is a measurable leading indicator.

Morgan Stanley is modeling private credit default rates reaching 8%. JPMorgan’s shorthand for the current software book is blunter: a third survive as winners, a third default, a third become zombies that exist only to service their debt.

So which deals are most at risk, and when? Sorted by where each sits on the fuse.

Tier 1: Already Detonated

Pluralsight (Vista Equity Partners, 2021, $3.5B). The canary, and the cleanest dated case study of the full fuse. Vista placed more than $1.5 billion of private credit debt on it at roughly SOFR plus 800. In Q1 2023 Vista asked lenders to loosen covenants and put in a $75 million equity injection. That was signal one, and it came a full 15 months before the end. By Q4 2023 Vista had marked the equity from $2.3 billion down to $1.5 billion. In the first half of 2024 it ran a drop-down, moving key IP into a non-guarantor subsidiary to raise $170 million of new money, $50 million of it from Vista itself, partly to fund a single interest payment. PIK and maturity extensions had masked the distress the whole way down. By Q1 2024 the equity was marked to zero. By summer the keys went to a lender group including Blue Owl, Ares, Golub, Oaktree, Benefit Street, Goldman, and BlackRock, who took roughly 85%.

One detail worth holding onto: at the end, Golub still marked the Pluralsight loan at 97 cents while Blue Owl had it in the low 80s. Same loan, 15-point spread. The marks are a lagging, inconsistent signal. The covenant request in Q1 2023 was the real one.

Medallia (Thoma Bravo, 2021, $6.4B). Bought at 9x forward revenue at the top of the market. The PIK toggle let it defer cash interest for years, which is exactly why the $300M-versus-$200M gap stayed invisible. The fuse lit at the end of 2025, when Blackstone, holding $1.5 billion of the debt, declined to extend the PIK window again. Once the largest holder refused, the rest followed within weeks. FS KKR had the debt marked at 79 cents, Apollo at 74. On April 22, 2026, the company went to its lenders and roughly $5.1 billion of equity went to zero.

Tier 2: Fuse Lit, Blast Imminent

These have already tripped a signal that historically precedes the handover by a few quarters.

Qualtrics (Silver Lake + CPP Investments, 2023, $12.5B). The next headline, probably inside 12 months. This one skipped straight to a late signal. After stacking the $6.75 billion Press Ganey acquisition on top of existing debt in October 2025, a JPMorgan-led group prepared a $5.3 billion financing in February 2026. On March 17 they pulled it, unable to place the paper amid AI-disruption fears. By May 6 that same group was looking at more than $500 million in paper losses on the deal. A pulled syndication means the market repriced the risk faster than the sponsor could restructure the stack. It is the same kind of signal that preceded Medallia, and here it tripped without the deal ever going through the slow covenant-relief stage first.

Quest Software (Clearlake). Equity already economically zero. Clearlake’s lowest-graded Quest debt trades around 25 cents on the dollar. At 25 cents there is no equity left to wipe. The only open questions are timing and structure, and it can formalize any quarter.

Cornerstone OnDemand (Clearlake, 2021). Multiple term loans underperforming, in the same corporate-learning category that already took down Pluralsight. Same sponsor as Quest and Alteryx, which matters: Clearlake is triaging 11 portfolio companies with underperforming debt at once, so its willingness to inject fresh equity into any single one is a real question.

Tier 3: Clock Running, Fuse Not Yet Lit

No covenant relief or pulled syndication here yet. The timing is set by the maturity wall, which you can read off the calendar.

Proofpoint (Thoma Bravo, 2021, $12.3B). The one to watch most closely. Debt is now around $4.67 billion after a dividend recap and the Hornetsecurity acquisition, against roughly $150 million of adjusted EBITDA, a coverage ratio that only ever worked at near-zero rates. It sits in 6 of 7 distressed private credit funds. Its 2024 refinancing term loan matures August 31, 2028. No covenant relief request has surfaced publicly yet, which means the fuse is not lit. Watch for that request to appear in 2026 or 2027, 12 to 18 months ahead of the maturity, because that is when this category of deal historically asks. When it does, that is signal one.

Cloud Software Group, Citrix plus TIBCO (Vista + Elliott, 2022, $16.5B). The largest software LBO by debt quantum, with $15 to $16 billion of leverage priced around SOFR plus 475 to 600. The threat here is platform displacement, not just AI: Citrix’s DaaS and VDI business is structurally squeezed by Microsoft Windows 365 and Azure-native alternatives. Big debt, far maturity, weakening core. This one grinds toward 2028 and 2029 rather than snapping.

Coupa (Thoma Bravo, 2022, $8B). Debt-to-equity around 65:35, with up to 30% of staff reportedly cut after the deal. Procurement and spend management is exactly the category where autonomous agents threaten seat-based pricing most directly. Thoma Bravo is reportedly pushing an agent-first re-architecture. That is a real attempt to outrun the problem, but it is a race against debt service and a 2028-vintage maturity.

New Relic (Francisco Partners + TPG, 2023, $6.5B). A $2.65 billion private credit package at SOFR plus 6.75%, roughly $300 million of annual interest against a business still mid-transition to consumption pricing. Observability is among the most AI-exposed categories, since agents increasingly write their own telemetry and Datadog and Grafana are cannibalizing the entry points. The coverage math is thin, but as a 2023 vintage the maturity is further out, which buys time the income statement does not have.

Alteryx (Clearlake + Insight, 2024, $4.4B). About $2 billion of debt at SOFR plus 650. Analytics automation, “data prep for analysts,” is arguably the single most AI-exposed seat in software. The timing risk here is less the maturity wall and more whether Clearlake, already triaging Quest and Cornerstone, has the appetite to carry a third one.

Tier 4: Leverage Loaded, Clock Further Out

Heavy debt, distant maturity, facts not yet decisively turned. The early tells show up in the secondary market before anything else.

Zendesk(Hellman & Friedman + Permira, 2022, $10.2B) carries $4.6 to $5 billion of private credit in customer support, ground zero for agents replacing seats, but reportedly generates $200M+ of ARR from its own AI. The open question is whether that outruns legacy seat shrinkage.Anaplan(Thoma Bravo, 2022, $10.7B) is the same vintage and sponsor as Coupa but executing better, with loans near par.Smartsheet(Vista + Blackstone, late 2024, $8.4B) loans have already appeared on private credit secondary bid lists, unusually soon after a buyout, which is a tell.Hyland(Thoma Bravo) loans are showing up on JPMorgan’s secondary trading lists.Finastra(Vista) placed a $2.95 billion first-lien and $500 million second-lien in July 2025, and the loans recently slipped from the high 90s toward 93 to 94.5.Avalara(Vista, $8.4B) carries about $2.5 billion led by Blue Owl in tax compliance, more defensible than most, though the leverage was still peak-vintage.

The Master Clock: The Maturity Wall

Step back from the individual deals and the timing snaps into focus. A Reuters analysis of 74 BDCs found only about $15 billion of $84 billion in assets mature in 2026. Then it steepens.

Roughly $43 billion matures in 2027, the first large-scale test of whether the market reprices software at refinancing, followed by the peak years of about $79 billion in 2028 and $83 billion in 2029.

That calendar sets the detonation dates. A deal blows when its PIK window closes early, as Medallia’s did, or when it hits maturity and cannot refinance into a market where mature software multiples have fallen from 9x revenue in 2021 to roughly 6x now. Loans maturing in 2026 that still have not been refinanced are almost by definition the underperformers, because the healthy ones were extended long ago. That hands lenders the most leverage they have had since the 2023 rate-hiking cycle, which is why amend-and-extend gets harder from here, not easier.

The Signals That Say a Deal Is About to Break

Watching a dozen of these unwind, the sequence is almost always the same. It runs in order, and each step buys the sponsor a little time at the cost of a lot more risk.

**The sponsor asks lenders for covenant relief.**The earliest tell. Vista did it on Pluralsight a full 18 months before the handover.**The PIK toggle gets activated or extended.**The one that matters most, because it is the mechanism that hides everything downstream. Every quarter on PIK is a quarter the principal grows and the real picture stays buried.**Lenders mark the debt below 80 cents, then below 75.**Once a syndicate marks below 75, the equity is economically worthless. It is no longer whether, only when.**Banks pull a planned syndication.**That is what hit Qualtrics on March 17. It means the market repriced the risk faster than the sponsor could restructure the stack.

For Proofpoint specifically, none of these has surfaced publicly yet. That absence is exactly what to watch, because step one is where the fuse lights.

What This Means for B2B

Three things matter for operators, not just observers.

PE has gotten weaker as an exit, not stronger. If sponsors cannot refinance the companies they already own, they are not buying new ones at prices that clear. The buyers who paid 8x to 10x are the ones now handing companies to their lenders. If you were modeling a PE exit in 2027 or 2028, rebuild it at 4x to 6x and see if the business still works.

Your PE-backed competitor with a 2027 or 2028 maturity is managing to a debt number right now, not a roadmap. They are cutting, not investing, and the closer they get to the wall the more true that becomes. That is an opening for well-capitalized AI-native challengers, and the window is widest in the 18 months before a competitor’s maturity. They are already taking it.

Venture debt has not had its real stress test yet. The private credit funds under pressure are cousins of the venture debt providers. Lines get renegotiated rather than renewed for anything that looks shaky. Model a case where growth drops 10 to 15 points and look at what it does to your covenants. Have the conversation with your lender now, while you still have the leverage of not having tripped one.

Have a question for Dear SaaStr? Submit it at saastr.ai/ai-mentor.