Scott Goodwin
Thanks to Paulinho, Mitch, and the Sohn Foundation for having me today. It's great to be here, as always, to support an incredible cause. We're going to talk about what the [ __ ] is going on in private credit. That's what everyone wants to know, and everyone keeps asking us, so we're going to try to set the record straight today on what's happening. A brief disclaimer.
What does Diameter do? We're a $30 billion credit-focused alternatives manager. We want businesses with situational and operational complexity. We're sector-focused. When those things come together with a playbook that allows us to invest based on what we learn from the past and bring it forward, there are opportunities for outsized returns.
1. Private Credit Gets SaaS-y
Private credit is getting SaaS-y. Public interest is really picking up—no pun intended. The retail bros who've been in these new retail funds want their money back. They're lining up to get it back. For those of you still using Google Search, search trends for private credit are off the charts.
The media is talking about different things in private credit. They're talking about cockroaches. They're talking about SaaS, and they're conflating a lot of different things. The private credit market is a $40 trillion market. What I'm going to talk about today is a $2 trillion slice of that market: leveraged finance direct lending. It's a market that dates back to the early 2000s, if not earlier.
2. Regulation Opens the Lending Void
How did we get here? Big changes came out of the GFC when it comes to leveraged lending. Regulatory capital rules changed materially for banks. Volcker, Dodd-Frank, and Basel III came in. Then the Fed and the OCC put on leveraged lending guidelines.
These guidelines said banks cannot lend to LBOs that are more than 6 times levered. What did that do? It made all the LBOs that were more than 6 times levered a big opportunity for private credit. Direct lending—leveraged finance private credit—stepped in to fill the void. This was a $200 billion market pre-GFC and is a $2 trillion market today.
Pre-GFC, these managers were lending to small companies with $25 million of EBITDA, conservative leverage, and really tight docs. Now it's a market with many multibillion-dollar deals. It has significantly outgrown the high-yield and leveraged-loan markets by multiples.
3. Fees Drive Industrial Scale
More recently, semiliquid private BDCs have been all the rage for public alternatives managers to raise. They're promising liquidity in illiquid assets. They raised $300 billion over the last 5 or 6 years. Why? Fees.
If you can do a $1 billion or $2 billion unitranche deal and put a first lien and second lien into your CLO, you'd make 40 basis points. Put it into a private BDC, and you make 6 to 7 times that number. Why? FRE, FRE, FRE. Multiples expanded dramatically for public alternatives managers, driving their stocks higher, with a huge amount of the growth coming from private credit, from leveraged finance direct lending.
M&A picks up. If you don't own a private credit business, you have to buy one. That culminated with BlackRock buying HPS, one of the best private credit businesses, last year for 30 times forward FRE. When you have that much money to deploy, that much interest in the asset class, and you have to raise more to drive your stock higher, how are you going to deploy it?
You build an industrial-scale deployment platform. Here's one of our private credit guys, really focused on the underwriting as the loans go through the machine. What do they have to do to deploy all that capital? Much larger deals—not the $50 million or $100 million loan to the widget maker in Sheboygan, Wisconsin.
They're focusing on what the private equity guys are focusing on, and that means SaaS at scale. Many billion-dollar-plus deals. Which sectors? The sectors the banks can't lend to anymore—the deals are more than 6 times levered. SaaS, healthcare IT, and business services. This will be a recurring theme.
What are all those sectors? Asset-light. What are all those sectors? Exposed to AI. What else is happening? Leverage is going up, and interest coverage is going down at the same time. People are rushing into this asset class to raise money and get their stock up.
Structural protections are getting weaker as well. Those billion-dollar-plus deals look much more like a syndicated bank loan. Maintenance covenants go away. PIK toggles increase. If you want to be in the private credit club, you need to get really SaaS-y. If you're not getting SaaS-y, you're not going to grow.
4. The SaaS Credit Thesis
Why was SaaS such a focus for private equity? A lot of these things we already know: recurring revenue, sticky customers, and operating leverage. I get the equity bet. You could make a money multiple—and many did—investing in SaaS over the last 15 years.
From a credit perspective, though, this was the main way to grow private credit direct lending. Many new deals during COVID were done only against recurring revenue, not against EBITDA. People started lending against revenues, not against cash flow. Most of it was SaaS at scale.
There were 10, 20, 30, or 40 $1 billion-plus financings a year during COVID and afterward against ARR, not cash flow, and with more leverage. Look at the sectors on the left side of the page: healthcare, IT, SaaS, commercial services, IT services, and process and professional services. A lot of things don't have hard assets in the ground, and these are the things the banks couldn't lend to.
But why were the private credit guys so comfortable? The multiples are huge. There was huge multiple expansion in the software sector over the past 15 years. A lot of the companies being lent to were at much lower multiples. Private equity wasn't buying some SaaS company for 40 times EBITDA. They were buying the ones that had less growth and doing some interesting things to them for 10, 12, 13, or 14 times EBITDA.
5. AI Comes for SaaS
This was all fine and dandy. Then, in late 2022, we're all sitting around the holiday table and ChatGPT shows up. You're making memes of your brother and sister in interesting costumes—at least, we were in my household. My kids were making fun of us and started to use it as well.
What does Diameter do? We've sought out LPs since we started in 2017 in certain sectors that we thought could make us smarter and help us think ahead of the curve, where we wouldn't have the same depth as they would. Four of those LPs are founders of big firms in Silicon Valley. We went to them and said, “What is this ChatGPT thing? What does it mean for credit?”
We're always in private credit, so you have to think about how you're going to lose, because you're really trying to get your money back. They said, “We're building AI companies to break the SaaS companies we built 15 years ago. So be careful.”
We started talking about this in our letters. In May 2023, we said we expected material disintermediation. In Q3 2023, we said some companies would be ruthlessly eliminated by AI. In Q2 2024, we said secular change was coming. Then we get to the end of 2024, and private credit guys are still lending to SaaS. They've got 30% to 40% of their portfolios in SaaS, and they're lying about the amounts in SaaS.
In 2025, AI is accelerating. We start to see bankruptcies coming. Winners and losers are going to be made in credit. Obviously, as we know, cloud code comes along 3 years later, during the holiday season of 2025. In our Q1 2026 letter, we said AI is coming for SaaS. We've been warning about it for 3 years.
So what did we do? We de-risked. We bought a lot of SaaS during COVID. After the investment-grade opportunity, one of the best things was buying security based software. Nobody was turning off their antivirus while sitting at home in their gym shorts trading bonds and stocks. Many of those companies got disrupted by cloud-security players over the last 3 years. They're already bankrupt.
We knew that when technological change comes, it can come fast. We reduced our exposure to SaaS and then went short some of the BDCs last year that had a lot of SaaS exposure. But private credit keeps lending to SaaS.
6. The AI Credit Reckoning
These are the self-reported numbers. On the next page, we're going to get into what the actual numbers are—different from what they'll tell you. The royal blue is the self-reported number. These are all public BDCs. PitchBook is on top of that; that's gray.
Then we use some AI tools to look through the portfolios with a number of prompts to figure out what the real exposure was. It's much higher, not surprisingly. Our BDCs are on the right. We limited SaaS to 5% when we started.
What did these guys miss? Everyone knows why private equity was investing in SaaS: a lot of good reasons. But technological change has caused the non-macro credit cycles that we've seen in our careers. At the beginning of the 2000s, there was a huge cycle in telecom due to technological change and the internet. In the 2010s, there was a huge technology change in fracking that caused a cycle in energy.
Now you have the largest technological change we're going to see in our investing lifetimes. It's going to make change happen faster than these companies can evolve. In private credit, you're lending for 5, 6, or 7 years. You're not meant to sell the loan. A lot of things were missed, but mostly the pace of technological change.
And you're buying a portfolio. Private credit is a leveraged asset class, back-levered. You're buying at 99. Your upside is 100, so you can make 1 point plus your carry. It's not like you have equity convexity. Upside is limited. This is a carry asset class.
Why would you ever have all your eggs in one basket, as many of the private credit guys did? It's almost criminal portfolio construction. This is probably the most important slide of the whole deck. In credit at par, you cannot have 30%, 40%, or 50% of your fund in one sector.
A lot of people who were originating these loans in private credit came from banks. The market was growing fast, so the response was, “Let's hire the guy from this bank or that bank.” When the bankers' job was to originate to distribute, nobody taught them portfolio construction. That shows up in the vintage of private credit that a lot of people own today.
So now, where are we today? What's the opportunity? AI risk is here.
Volatility is here. That equity cushion they were so excited about is popping. And AI is coming for SaaS. SaaS multiples have de-rated materially, and syndicated bank loans in the SaaS space are also down materially.
A custom AI short basket we created in the equity space, consisting of AI-exposed companies, is down by more than 50% since the beginning of GDP GPT. So where’s the acute problem? It’s in that 2021–2022 vintage of ARR-based SaaS loans. The maturities are coming due now.
What about the marks? The marks are wrong in many cases. There’s been a 6% average difference between one BDC and another over the past 6 years, up from almost nothing pre-COVID. When we were looking at shorting some of the BDCs last summer, we found situations where the marks were 40 points apart. Just using AI tools, that created some easy trades, obviously.
The regulators are going to come for this, and this will change. So, what are private credit managers saying? “Nothing to see here. LTM EBITDA is fine in SaaS, and AI can’t impact names.” All backward-looking. “We’ve been lending to SaaS for 20 years. We know what we’re doing.”
Or it’s the equity’s problem. I think that’s the worst one: “It’s not our problem; it’s the equity’s problem. Look at the multiple.” Well, that’s not the multiple anymore. Look at what the public stocks did.
So, what’s happening in BSL? Because public credit can tell you what might happen in private credit. Recoveries are going down, especially in tech. We made a mistake in an investment in 2018 and 2019, in the early days of tech lending, and had a very low recovery on something that we lent to.
From that, we learned that when you have asset-light companies and you start to lose subscribers, the recovery is very low, especially if you have long maturities. In some of the longer-maturity private credit loans, if sponsors believe the business is disrupted by AI, their incentives are to take dividends and turn you into an IO.
These portfolios, as we showed before, are full of SaaS. This is the buildup of the reported view, PitchBook, and then our view. If you add on other AI-exposed sectors—healthcare IT, business services, and so on—the numbers are more like 40% to 50% on average of a private credit portfolio. A backward-looking portfolio has an AI risk factor. Horrible portfolio construction.
What else? There are a lot of ways that private credit managers convinced LPs they were the best: sourcing, sector selection, and SaaS because it’s not cyclical. But many did it through leverage. Second-out structures, where you turn your first lien into a second lien; bank JVs; buying CLO equity in the BDCs—ways of leveraging the vehicle more to juice the returns. Given the fees, you understand why.
Now, those non-traded structures—people are starting to figure this out, and they want out. There are $8 billion of unmet redemptions in Q1. Leverage providers are growing cautious. We’re seeing banks look to sell their exposure, either directly or synthetically, reducing borrowing bases for markdowns and changing the way they lend to SaaS.
In some cases, they’re walking away or just capping their exposure. We’ve seen this cycle before: marks, leverage, margin calls, selling.
So, what’s the opportunity? This is not systemic. The media would like you to think this is a systemic problem. It is not. The sky isn’t falling. It’s 5% of the private credit market, it’s broadly distributed, and SaaS is only 1/3 of that market.
7. Forced Selling Creates Opportunity
So, what’s the opportunity? Public BDCs, buying secondary from forced selling, and new loans. Public BDCs have traded down about 30% from the highs last year. We like to look at them as a percentage of GAV, not NAV.
A lot of what you see quoted in the press and on Twitter or X is talking about NAV. But public BDCs are back-levered. So, for every $100 of loans, you've got 50 cents of debt and then 50 cents of equity. We like to think about it through the full stack.
The opportunity set is to buy some of these when they get to 0.85x GAV, or 85 cents on the dollar, at a low-teens yield. For those, a lot of the smaller BDCs that were not in hypergrowth mode and were not as focused on SaaS and the AI names are really interesting at that price. They hit that price about a month ago, so we started buying.
We like that opportunity at the right price. Not all BDCs, but some. Now, secondaries are something you’re going to hear a lot about for the next few years. There are $300 billion in retail funds with material outflows.
We expect between retail funds causing redemptions, bank margin calls, and prudent risk management by LPs, $150 billion to $200 billion of selling across the secondary private credit market. So, how do you attack that opportunity? We don’t think it’s by bidding the wrong price for large portfolios. So far, that’s what we’ve seen some secondaries funds doing.
We think it’s about knowing your credits and cherry-picking the right names. So, how do we do that? Names you already know that we’re already lending to. We’re sector-based, so names we know from the public market. And then sponsors we know well, where their loans are for sale.
We’ve done 15 trades so far over the last 2 months in all of those categories. We think that opportunity set, as I mentioned, is going to be expansive. You have to be patient.
Finally, new loans. Spreads are 50 to 75 wider on some of the new loan opportunities we're seeing. Some of the largest players are not re-upping to existing loans because they’re capital-constrained. That’s going to create a shrinking market and a better opportunity set. But you have to know the names.
So, what I want you to take away: portfolio construction first, above everything, in par credit, whether it's public or private. This is not a systemic issue. And know the names. Thank you.