Latest / Deep Values: Value Investing Deep Dives / Howard Marks: How AI Broke the Private Credit Market
Transcript
- 0:00You know, it is it's actually wild to think about how much the
- 0:03financial universe has just completely mutated over the last
- 0:06few decades. I mean, if you teleported an
- 0:09investor from say the summer of 1968 into today's market, Oh,
- 0:14they. 'D be completely lost, they
- 0:15wouldn't even recognize the landscape.
- 0:16Exactly. Back then the investment world
- 0:19was just, you know, quaint, very provincial.
- 0:22It was basically just publicly traded stocks and like high
- 0:25geared government or corporate bonds.
- 0:27That was it. Right, very structured.
- 0:28Yeah, very visible. And today we are staring down
- 0:32this this $2 trillion gold rush in something called Direct
- 0:36Lending 2. Trillion.
- 0:37It's a staggering contrast, really.
- 0:39We've moved from that visible public market you mentioned to
- 0:41this sprawling, highly complex and honestly largely hidden
- 0:46financial ecosystem. And that's exactly our mission
- 0:49for this deep dive. Today, we're going to unpack
- 0:51this massive hidden machine. We're basing this on a really
- 0:55fascinating recently published memo by the legendary investor
- 0:58Howard Marks of Oak Tree Capital.
- 0:59Yeah, his memo is titled What's going On in Private Credit?
- 1:02And it is incredibly insightful. It really is.
- 1:05So we need to figure out how this ecosystem was built up over
- 1:09the last 15 years, how the sudden like explosion of
- 1:12artificial intelligence is currently throwing a massive
- 1:14wrench into its gears. A.
- 1:15Huge wrench. And ultimately what this all
- 1:18means for your own investments. But really quickly before we
- 1:20jump in to access more advanced researches, applications and our
- 1:24latest published personal intelligent multi agent deep
- 1:27value intelligence. Welcome to explore and join the
- 1:31Deep values dot AI. I highly recommend checking
- 1:34that. Out yeah, and to track total
- 1:35market temperature. Welcome to search and download
- 1:38the iOS app Buffett Indicator from the Apple App Store.
- 1:41Quick disclaimer, all analysis is based on publicly available
- 1:44sources and may be subject to revisions or differing
- 1:47interpretations. All contents from this channel
- 1:50are for educational and research purposes only and do not
- 1:53constitute investment advice. Conduct your own due diligence
- 1:56before making investment decisions.
- 1:58Always do your homework. Exactly.
- 2:00If you like to access wider topics including life and
- 2:02investing philosophy, please subscribe to our sister channel
- 2:05Deep Values. Don't forget to like this video.
- 2:08Much appreciated. Okay, housekeeping done.
- 2:11So to understand the current panic that's kind of rippling
- 2:15through private markets, we have to look at the anatomy of this
- 2:18space. Right.
- 2:20How did we even get here? Yeah.
- 2:21How did a relatively obscure corner of finance amass $2
- 2:25trillion in the first place? Well, to get that you have to go
- 2:29back in time and really crace the evolution of how companies
- 2:32borrow money. I mean prior to the late 1970s
- 2:36credit, which is essentially just non government debt was a
- 2:39highly, highly exclusive. Club like you had to be flawless
- 2:42to get a loan. Exactly.
- 2:43If a company didn't have a pristine investment grade
- 2:46rating, they basically couldn't issue public bonds.
- 2:48They had to go hat in hand to a traditional bank or maybe a life
- 2:50insurance company just to beg for a private loan.
- 2:53But. Then the rules of the game
- 2:54completely changed. Yeah, the mentality shifted in
- 2:57the late 70s and 80s. This idea took hold that, hey,
- 3:00maybe non investment grade companies should be able to
- 3:02issue bonds too, right? Provided the interest rate was
- 3:05high enough to, you know, compensate the lender for the
- 3:08risk of default. And that basically burst the
- 3:11high yield bond market. Right, which then fueled the
- 3:13whole leveraged buyout craze of the 1980s.
- 3:16And then Wall Street started pulling these debts, selling
- 3:19them off in tranches in the 90s, and, well, we all know the 2000s
- 3:23brought us the alternative investment boom in the subprime
- 3:25mortgage crisis. Which triggered the 2008 global
- 3:28financial crash, which honestly acts as the massive pivot point
- 3:32for everything we are seeing today.
- 3:33Oh. Absolutely.
- 3:34It's Ground Zero for this boom. Because after 2008, regulators
- 3:37cracked down hard on traditional banks, right?
- 3:40So the banks pulled back from lending to riskier companies.
- 3:43But private equity firms, they still needed billions of dollars
- 3:47to buy businesses. So this massive vacuum just
- 3:49opens up. And non bank lenders stepped
- 3:52directly into that vacuum. That is the birth of the modern
- 3:56private credit boom. And we should probably be
- 4:00precise with our terms here because people mix them up.
- 4:03Private credit is the whole universe of non bank lending,
- 4:07but direct lending is a specific subset of that.
- 4:09It's where non bank lenders make private loans directly to mid
- 4:13market companies. Usually businesses that are
- 4:15being bought out by private equity or.
- 4:17Exactly. And the scale is wild. 20 years
- 4:19ago, the entire private credit sector was what, maybe $150
- 4:22billion? Yeah, today, direct lending
- 4:25alone is roughly 2 trillion. Which is just, I mean, that kind
- 4:28of exponential growth doesn't happen just because of a
- 4:30regulatory vacuum. It requires an emotional driver.
- 4:33Human nature, right? Marks brings in this brilliant
- 4:36quote in the memo from the economic historian Charles Peak
- 4:39Kindleberger to explain it. He says there is nothing so
- 4:42disturbing to 1's well-being and judgement as to see a friend get
- 4:46rich. It's so true.
- 4:48Envy is basically the engine of every financial bubble.
- 4:50Yeah. It sounds like direct lending is
- 4:54like the trendy new neighborhood in finance. 1st, The Pioneers
- 4:58move in. They get great deals, high
- 5:01yields, strong protections on their loans.
- 5:03Because they're taking the initial risk.
- 5:04Yeah, right. But then everyone else notices
- 5:07the big money floods in, standards completely plummet,
- 5:11and suddenly people are wildly overpaying for a fixer upper.
- 5:14That's exactly it. To put all that new money to
- 5:16work, the new lenders have to compete with each other.
- 5:18So they start lowering their standards, stripping away safety
- 5:21protections just to win the deals.
- 5:23Warren Buffett has that incredibly colorful progression
- 5:26for this exact cycle, he says. First the innovator, then the
- 5:29imitator, then the idiot. It happens every single time.
- 5:33It's clockwork. But I do want to push back on
- 5:35the risk aspect for a second though, because one of the main
- 5:38selling points for these direct loans, especially now that they
- 5:41are being aggressively marketed to like individual retirement
- 5:44accounts, is that they don't experience the wild price swings
- 5:47of the public stock market. Right.
- 5:49That lack of volatility pitch, yeah.
- 5:51Doesn't that lack of daily volatility make them inherently
- 5:54safer for a portfolio? OK.
- 5:57So that is the grand illusion of private markets.
- 6:01Private credit is essentially A Schrodinger's cat situation.
- 6:04Wait. Schrodinger's Cat.
- 6:05How so? Well, because these private
- 6:08loans aren't traded on a public exchange every day, nobody is
- 6:11staring at a ticker tape watching the price jump up and
- 6:13down. The box is closed.
- 6:15But just because the box is closed doesn't mean the asset
- 6:18inside is actually stable. Oh, I see.
- 6:20The underlying business might be hemorrhaging customers, facing
- 6:23new competitors or struggling to its bills.
- 6:27The credit risk is still very much there.
- 6:29It's just hiding in the dark. So the lack of daily price
- 6:32volatility is really just an accounting feature, not an
- 6:35actual safety net. Exactly.
- 6:37It's an illusion, OK. So they have $2 trillion to
- 6:39spend, and they are willfully ignoring the risk because the
- 6:42assets aren't publicly traded. But they didn't just put that
- 6:45money anywhere. They poured an enormous chunk of
- 6:47it into a sector that lenders used to avoid like the plague.
- 6:51Right back in the early 2000s, technology companies were
- 6:54considered way too fundamentally risky to be credit worthy.
- 6:58You just couldn't lend them money.
- 6:59So how did software debt suddenly become, what, 20 to 30%
- 7:03of the entire direct lending market?
- 7:05It comes down to a fundamental shift in how software is sold.
- 7:10Think about the transition from buying a piece of software once,
- 7:13like on a CD-ROM at Best Buy, to subscribing to it monthly in the
- 7:17cloud. The recurring revenue model.
- 7:19Yes, the recurring revenue revelation.
- 7:21Private equity firms realize that essential software
- 7:24companies operated almost like utility companies.
- 7:26Once a business integrated a specific software into its daily
- 7:29operations, it was incredibly painful to switch to something
- 7:32else. So they're locked in.
- 7:34Right. And that created subscription
- 7:36model generating highly predictable endless cash flows.
- 7:39And lenders love nothing more than predictable cash flow.
- 7:42Oh, they became completely obsessed with it.
- 7:44These software companies appear to have impenetrable moats
- 7:46around their businesses, so private equity started buying
- 7:48them up and direct lenders were more than happy to supply the
- 7:51debt. They were practically throwing
- 7:53money at them. They were.
- 7:54In fact, they were so eager that they started lending money based
- 7:58on a metric called annual recurring revenue, or ARR,
- 8:02rather than actual profit. OK.
- 8:04Let's pause on that because lending against revenue instead
- 8:07of profit sounds incredibly dangerous.
- 8:10I mean, if a company brings in $100 million in revenue, but
- 8:13spends 110 million on marketing and developers, their profit is
- 8:18-10 million. Yep, there is literally no
- 8:21actual cash left to pay back alone.
- 8:23And that is the crucial vulnerability.
- 8:25Lending against revenue only works if you assume the company
- 8:28will eventually grow its way to profitability, or more likely
- 8:31that the private equity firm will just sell the company in a
- 8:34few years for a massive premium to pay off the debt.
- 8:36So it's a giant game of hot potato.
- 8:38Basically, they were acquiring software companies at massive
- 8:41multiple sometimes 20 times EBITDA, meaning the valuation
- 8:44was 20 times their operating earnings, fueled almost entirely
- 8:47by this private debt. OK, so we have this giant tower
- 8:50of debt built on the assumption that software subscriptions are
- 8:52immortal. And then enter the plot twist.
- 8:55Artificial intelligence. The timeline on this is
- 8:57staggering because it happened so incredibly fast for a couple
- 9:01of years, like 2024 into 2025. AI starts writing code, but it's
- 9:06mostly just assisting human programmers, right?
- 9:08The market doesn't really react, but then November 2025 hits.
- 9:12Right Anthropic releases a massively powerful new model
- 9:16specifically for coding. And then it just snowballs.
- 9:18Oh yeah. In late January 2026, a barrage
- 9:22of 11 automation plug insurance are released that basically take
- 9:25over complex tasks across multiple fields without human
- 9:29intervention. And by the first days of
- 9:30February 2026, a cognitive tipping point is reached.
- 9:34The market suddenly wakes up. Investors looked at these
- 9:36software companies, the legacy ones they thought had those
- 9:39impenetrable moats in guaranteed subscription revenue, and
- 9:42realized they were incredibly vulnerable to disruption.
- 9:44Because the Moat was suddenly gone.
- 9:46Exactly. If an AI can write code and
- 9:48automate tasks, Florida State, the barrier to entry for
- 9:50competitors drops to near 0. A startup with a fraction of the
- 9:54budget can build a better software tool in a matter of
- 9:56weeks. OK, wait.
- 9:57I need to understand the mechanics of the panic here
- 9:59though, because if the legacy software company hasn't actually
- 10:04defaulted yet and they're still paying their monthly interest
- 10:07bills to the direct lenders, why are the lenders panicking?
- 10:11The loan is still performing right.
- 10:13Because credit markets are entirely forward-looking, they
- 10:16rely heavily on what's called refinancing risk and the equity
- 10:19cushion. OK.
- 10:20Walk me through the equity cushion for sure.
- 10:22Imagine you buy a house for $1,000,000.
- 10:24You put down 200,000 in cash. That's your equity, and you
- 10:27borrow 800,000 from the bank. If the house's value drops by
- 10:30150,000, you take the hit, but the bank is still safe because
- 10:34the house is still worth 850,000.
- 10:36Right. So the bank's loan is covered by
- 10:38the value of the house. Exactly.
- 10:40But if the house's value drops by 300,000, suddenly the
- 10:43property is worth less than the 800,000 you owe.
- 10:45The bank's equity cushion is completely gone now.
- 10:48In our scenario, the private equity firms are the homeowners
- 10:52and the direct lenders are the bank.
- 10:54When the market realized AI could disrupt these software
- 10:57businesses, the perceived valuation of those businesses
- 11:00plummeted overnight. And that ties directly into the
- 11:03refinancing risk because these loans eventually come due.
- 11:06Right. Yes, exactly.
- 11:07If the company's valuation has collapsed, no new lender is
- 11:10going to step in and give them a new loan to pay off the old one.
- 11:13So the original lender is left holding the bag on a business
- 11:17that is suddenly worth less than its debt.
- 11:18Precisely. The panic wasn't about defaults
- 11:21happening that afternoon. It was the sudden collective
- 11:24realization that the terminal value of these companies had
- 11:27evaporated. Wow, so AI was the spark that
- 11:30ignited this panic in the software debt market.
- 11:33But a spark only causes a massive fire if the forest is
- 11:36already bone dry. And it was very, very dry.
- 11:38Right. The macroeconomic backdrop is
- 11:41what made this whole system so incredibly fragile to begin
- 11:43with. Howard Marks refers to this in
- 11:45the memo as The Sea Change. Yeah, and it is arguably the
- 11:49most important financial dynamic of our lifetimes.
- 11:52I mean, between 1980 and 2020, interest rates in the United
- 11:55States fell by 2000 basis points.
- 11:57That is a drop from 22.25% all the way down to 2.25%.
- 12:02A 40 year tailwind. Let's actually do the math on
- 12:05how that impacts these buyouts just so we can visualize it.
- 12:07Imagine you are a private equity firm buying a company for $100
- 12:11million. You put down 10 million of your
- 12:14own money and you borrow 90 million from a direct lender.
- 12:18If your borrowing cost is just 2%, your interest payment is
- 12:21just 1.8 million a year. Which is nothing.
- 12:24The company's cash flow can easily cover that.
- 12:26Right. So you hold it for a few years,
- 12:28sell it for 120 million, pay back the 90 million debt and
- 12:32walk away with 30 million. You just tripled your initial 10
- 12:35million investment. And that's why falling interest
- 12:37rates make private equity look brilliant.
- 12:39It shrinks borrowing costs and artificially inflates asset
- 12:42values. This was particularly extreme
- 12:44between like 2009 and 2021, when interest rates were effectively
- 12:480. Right.
- 12:49Everyone looked like a genius. Exactly.
- 12:51Managers were hailed as prodigies for simply riding a
- 12:53mathematical wave. But then inflation hits and the
- 12:56Federal Reserve jacks rates up to over 5%.
- 12:59So now all these direct loans that were made when money was
- 13:02basically free are resetting at much, much higher rates.
- 13:05Yeah, run your $100 million buyout math again at an 8%
- 13:08interest rate. OK, so suddenly you owe 7.2
- 13:11million a year in interest? Right.
- 13:13And the company's cash flow can't cover it.
- 13:14All the profits are eaten by the lender.
- 13:16Deals that made perfect economic sense when leverage cost 2% are
- 13:20fundamentally broken at 8%. So the companies are worth less,
- 13:24making it impossible to sell them or refinance them.
- 13:27The Gears of the entire private equity machine just grind to a
- 13:30halt. Which is terrifying.
- 13:32It is. And Howard Marks points to a
- 13:34historical parallel here that really gave me pause.
- 13:37The 1929 stock market crash. I mean, we have investors buying
- 13:41into hyped up, highly levered assets that are largely
- 13:44illiquid. And suddenly the math changes.
- 13:46Are we actually looking at a systemic collapse?
- 13:49Well, it's a valid question. The memo references author
- 13:51Andrew Ross Sorkin, who identifies 3 deadly ingredients
- 13:54that caused the 1929 crash. And honestly, the echoes today
- 13:58are deafening. What are the three ingredients?
- 14:001st, the aggressive sale of risky assets to the public
- 14:02without regard for suitability. OK second, heavy leverage,
- 14:05borrowing enormous amounts of money to amplify returns.
- 14:07And 3rd, a severe mismatch between illiquid assets and
- 14:10short term loans. Right, because when the market
- 14:13dipped in 1929, lenders called in their loans, investors
- 14:17couldn't pay, and the forced liquidations just crashed the
- 14:21entire system. Exactly.
- 14:22And we are seeing echoes of those exact three ingredients
- 14:26today in direct lending. I mean in the twenty 20s, these
- 14:29direct lending vehicles started being aggressively marketed to
- 14:33everyday retail investors. Yeah.
- 14:34People who might not fully understand the complexities of
- 14:37private credit. Right.
- 14:38And there's heavy leverage applied within these vehicles to
- 14:41boost the yield. But the mismatch is the most
- 14:44alarming part to me. They are investing in private,
- 14:47highly liquid loans to companies, but giving retail
- 14:50investors the illusion that they can withdraw their money
- 14:52whenever they want. Like a checking account?
- 14:54Yes. And that is exactly where the
- 14:56stress fractures are opening up right now.
- 14:58We are seeing public vehicles for direct loans.
- 15:01They're called business development companies, or BDC's,
- 15:04trading at wide discounts to their net asset value.
- 15:07Meaning the fund's shares are actually selling on the open
- 15:09market for less than the supposed value of the private
- 15:12loans held inside the fund. Right, the market is essentially
- 15:15calling their bluff on what the loans are actually worth.
- 15:17Wow. And more concerningly, when the
- 15:19AI panic hit the software debt market, everyday investors got
- 15:23spooked naturally and tried to pull their money out of these
- 15:27non traded direct lending funds. But they couldn't, could they?
- 15:30No, these funds have limits on redemptions.
- 15:33They essentially had to gate the funds telling this they're sorry
- 15:36you can't have your money back right now because we can't sell
- 15:39these private loans quickly enough without taking a massive
- 15:41catastrophic loss. Which ironically, just causes
- 15:44more panic. I mean, if you tell an investor
- 15:46they can't have their money, their immediate psychological
- 15:49reaction is to try and get their money even faster.
- 15:51Yeah. Humans don't like being cold.
- 15:53A door is locked. The redemption limits have
- 15:56worked so far to prevent a 1929 style fire sale liquidation, but
- 16:01it deeply exposes the inherent flaw of treating illiquid
- 16:04private credit like a liquid ATM.
- 16:07It really does. It reminds me of that infamous
- 16:092007 quote from Citigroup CEO Chuck Prince right before the
- 16:13global financial crisis. He said As long as the music is
- 16:16playing, you've got to get up and dance.
- 16:18Famously bad timing. Right.
- 16:21But so many direct lending managers did exactly that.
- 16:24They lowered their underwriting standards just to deploy the
- 16:27billions of dollars flooding into their funds, knowing full
- 16:30well the risks. Yeah, but Marx goes out of his
- 16:33way to contrast that industry wide recklessness with Oak
- 16:36Tree's disciplined approach. I mean, they explicitly limited
- 16:40their software exposure to under 15% of their total assets and
- 16:44they predominantly stuck to senior defensively structured
- 16:47loans. They chose to sit out a large
- 16:49portion of the dance while the music was blaring, which.
- 16:52Has to be hard. It requires immense
- 16:54psychological discipline. When everyone around you is
- 16:56getting rich by lowering their standards, sitting in the
- 16:58sidelines feels like losing, right?
- 17:00But true believers make the most money in manias, while skeptics
- 17:04lose the least when the mania inevitably crashes.
- 17:06That's a great way to put it. So let's synthesize this.
- 17:09What does this actually mean for your own portfolio?
- 17:12We have a 40 year tailwind of dropping interest rates that has
- 17:15violently reversed breaking the math of leverage.
- 17:18We have an AI shockwave that just evaporated the perceived
- 17:22safety of a massive chunk of the direct lending market.
- 17:25Are we watching the death of private credit?
- 17:29No, I wouldn't call it the death of private credit, but it is
- 17:32undeniably the end of easy mode Easy.
- 17:34Mode is over, yeah. For the last 15 years, a rising
- 17:38macroeconomic tide lifted all boats.
- 17:40Now the tide is going out. The market is transitioning from
- 17:44a period of blind faith to a necessary and frankly painful
- 17:47credit cycle. So who survives?
- 17:49Moving forward, the only lenders who will survive and generate
- 17:52real returns are the ones who actually did their homework, the
- 17:55ones who maintained strict underwriting standards, demanded
- 17:58a true margin of safety, and, you know, didn't blindly lend 20
- 18:02times revenue to a software company just because their
- 18:04competitors were doing. It right So the ultimate
- 18:06antidote to irrational exuberance isn't just government
- 18:09regulation, it's humility. The financial industry is
- 18:11brilliantly creative at inventing new complex packages
- 18:15to hide risk. But whenever you are promised a
- 18:17way to get high returns without volatility, especially in a
- 18:20shiny new vehicle you don't fully understand, like a retail
- 18:22direct lending funds, repticism has to be your default setting.
- 18:26Humility requires recognizing that no system is foolproof and
- 18:30the higher the collective certainty of the market that an
- 18:32asset class is safe, the heart of the eventual fall.
- 18:35That is the enduring lesson of 1929, of 2008 and whatever cycle
- 18:39we are entering right now. Well, I want to leave you with a
- 18:41final provocative thought to Mull over.
- 18:44We just spent a lot of time mapping out how an AI coding
- 18:47model released in late 2025 was enough to instantly shatter the
- 18:50perceived safety of a multi billion dollar software debt
- 18:53market. Yeah, but think about the rest
- 18:55of the direct lending universe. What happens to the billions in
- 18:58private debt backing, say healthcare administration or
- 19:01legal services or logistics when the next wave of AI automation
- 19:05hits those specific sectors? Oh wow, are we just seeing the
- 19:08very first domino fall in a much larger economy wide credit
- 19:12reckoning? It's a chilling question because
- 19:14it implies that the fundamental nature of enterprise value and
- 19:17the collateral itself is shifting faster than the debt
- 19:20markets can possibly price it in.
- 19:22It really makes that 1960 68 market of quaint public stocks
- 19:26and high grade bonds sound pretty relaxing right about now.
- 19:29I'd take it. Well, thank you for joining us
- 19:31on this deep dive into the hidden machinery of private
- 19:34credit. Keep questioning the consensus,
- 19:36stay curious, and we'll see you next time.