KKR/BN: Panic at the Disco!
Private credit rumblings
~18 min read.
(Edit: changed certain estimated values. Fixed grammar. Main conclusions unchanged.)
“It’s only when the tide goes out do you see who’s been swimming naked.”
— Warren Buffett
Background:
You know we have nearly 40% (at market) of the portfolio in two alternative asset managers. They manage a lot of private equity (PE) and private credit (PC) money.
The market hates private credit right now, and some private equity.
This is mainly because of software investments, among others. The current narrative is AI will harm software business models, or at least create enough competitive pressure (higher costs from investing in product/features to stay competitive, or more price pressure from vibe-coded entrants, etc.) on these companies that their economics will deteriorate. Consider for example that AI tokens represent a new variable cost the software industry previously did not need to pay. Furthermore, privately-held software firms will likely face additional pressure as the PE guys bought them with ~35% debt — there’s interest to pay whether or not their operating margins get squeezed.
If AI is a net cost to them, the numbers imply many privately-owned software businesses face potential value destruction. I estimate that if they bought a software firm for $100, they’re doing $28 in revenue, of which $7 is cash flow, of which $3 is paying debt interest, leaving $4. ~15% of revenue is not a lot of levered cash flow to deal with these problems, so to try and maintain profit margins they’ll have to “find money somewhere” (e.g., a big cost lever is sales staff, often 10-20% of a software firm’s revenue). Even without the numbers, just think: they previously acquired these companies assuming no such threat to growth or margins existed, and now it does.
On the loan side, many sophisticated lenders believe bankruptcy recovery rates will be 20-40 cents on the dollar, as there are no tangible assets to sell in recovery.
There’s the backdrop.
Lending is such a strange business, man.
Everyone buys in when it’s going well, then panics when it isn’t. It’s a microcosm of market behavior overall. What is strange is that lending has not changed for centuries, and has always had this cycle. So you can just look at what happens historically. You don’t need to guess or be surprised. There are tons of books, reports, and data sets. I spent a bunch of time reading and doing some math before investing in a bank. Now I’m pretty comfortable investing in well-run banks despite knowing it’s basically certain that eventually my bank will go through a loss cycle.
Many people don’t think about rising defaults and the loan loss cycle. Then, when it happens, they do (too late). A recession or disruptive event in one or more sectors usually starts it. AI might do it this time. We’ll see. If we’re going to go through a loss cycle — and we don’t know if we are — then we’re at the beginning. There’s decades of lending data to show what happens, though. It shows everyone shares the pain, but the pain is not evenly distributed. Lenders who maintained their discipline felt some pain but did OK. Stupid lenders took far more of the pain and usually had less of a buffer to absorb said pain, often resulting in death, or big, irreparable hits to their balance sheets and business value.
(E.g., do any Americans still bank with Wachovia or Countrywide? No, the ‘08 crisis killed them. But people still bank with Wells Fargo and JPMorgan Chase, and those banks are much larger today than they were in ‘07. Yet all of them booked a lot of loan losses in ‘08. The difference between them was discipline and defensiveness.)
When things are going badly, people panic-sell everything because they can’t tell the two classes of lenders apart. It’s easier to see a deteriorating credit/macro environment than to discern which lenders followed a disciplined, defensive process vs. which were dancing while the music played.
This is what people are reading and thinking today. Notice:
Morgan Stanley had to limit redemptions in one of its private credit funds.
Blackstone saw nearly 8% redemptions in BCRED, its non-traded credit fund for high-net-worth individuals (HNWIs), sold through investment advisors. The limit is 5%/quarter to prevent forced liquidation. Still, Blackstone honoured requests using its own balance sheet to buy units, plus employees who voluntarily invested. BCRED is an industry bellwether as it’s the largest such fund.
Other managers like Blue Owl are facing stiffer redemption requests, honouring them by selling some of the underlying portfolio to institutions and paying investors out with the cash.
JPMorgan quietly marked down certain private credit fund and software loans. That implies other institutions have, too. An Apollo executive said they have. It suggests others like KKR have, since KKR has private equity & credit investments in software.1 The mark-downs reduce how much these fund managers can borrow and invest alongside clients’ equity, kind of like margin limits in a regular investment account. So it puts the debt cycle into reverse by forcing guys to reduce or unwind positions.
Blackstone CEO Jon Gray has been out there on CNBC, sounding calm and collected. Think “everything is fine, guys!” but in more sophisticated language. However, there’s no incentive for people to take him at his word since he can’t say the opposite.
All this is causing fear. “Where there’s smoke, there’s fire,” naive market participants say, sounding wise while running away because they didn’t do their homework.
Sources differ on the actual underlying default rates and the trends. It also depends which kinds of loans you count and which ones you don’t. Fitch Ratings thinks the ones they count are going up. Apollo, with better data but worse incentives, says down.
Conflicting, difficult-to-understand information, combined with heightened emotions! Isn’t investing fun?
Why am I so chill?
Obviously, like everyone else, I prefer “number goes up” over “number goes down.” My KKR is down >40%. But volatility is inherent to this game. Find me a stock that doesn’t gyrate and just goes up 20%/yr, so I can buy it.
We’ve invested in financials since 2013, making very large bets on banks and alt. asset managers in ~2020 and ~2023. If you buy into these thinking they have nice, stable profits and growth, you are silly. Insurers, banks, and alternative asset managers are all capital allocators — they gather capital from policyholders, depositors, or clients, then invest it. Not every investment works. In a recession, more of the investments go wrong and fewer go right, at least temporarily. E.g., bank loan losses rise. Private equity businesses struggle or even go bust, like KKR’s surgery center business — supposedly very stable — did during COVID. This makes them cyclical.
The way to think about this situation clearly is: if a loan loss wave occurs, such as fallout from AI/software, will KKR be permanently screwed, or just temporarily?
(And: how much of these future problems are priced in today? Because if the market already expects an institution to lose a bunch of money — and prices it as such — then you are already being compensated for underwriting this risk. And so you “can’t lose”… kind of. There’s more to it than this but you get my point.)
The Game of Risk
When analyzing a financial (or any) company, many people get the key questions and framework wrong, in my opinion. The question is not: “are these guys taking risks?” All businesses involve risk. You should find risks. Because you’re a biased human, you’ll also find more “risks” when conditions deteriorate, and fewer when they’re good. If your frame is “risk is bad,” you’ll panic in bad times, as you suddenly see risk everywhere.
I’m not saying don’t look for risk. I’m saying the real question is: since all businesses involve risk, do the managers understand — and are they properly managing — the risks they are taking? And are shareholders being well-compensated for said risks? Buying into a business isn’t about “I have so much to gain!” it’s, “what risks am I underwriting, do I understand them, and am I being paid handsomely to assume them?”
This would help people understand private credit better. But they don’t do this.
Which brings us to math time.
KKR: What’s the Impact?
Let’s first divide the impacts to KKR into short- and long-term, then second we will quickly do a base-case valuation using a reasonable shorthand approach.
Short-term financial impact:
Fees: 7% of KKR’s private equity and credit fund AUM is invested in software. At the upper bound, this is ~$40 billion in fee-paying AUM, and $26 billion in carried-interest-eligible AUM. PE drawdown funds charge fee revenue largely on their initial cost basis, not on the marked-to-market values, so if the investments go to zero, there’s no revenue impact, at least until those funds start getting liquidated and returned to investors 3-5 years from now. By then, KKR will have raised the next vintage of funds, filling the hole. Thus, most likely, we’d see flattish growth for a 1-4 years. Credit funds earn fees based on their current net asset value, marked-to-market, so losses impact revenue. If you assume half the software exposure is in private credit and the other half was equity funds, then they’d lose ~$10-15 billion gross (~$9 billion net of recoveries) if every software loan defaults. This is implausible for several reasons, but let’s go with it.2 The upper bound fee rate on that is 1%, or $150 million, which is 3.5% of revenue. For a business (a) we think can grow ~10%/yr (+/- 10%), and where (b) we paid for exactly none of that growth at $45 & $55 per share in 2023. In fact, the price implied revenue & profit declined. See below.
So, who cares?
3.5% of revenue is noise. Double it. Still noise.
Carry: call it $30, not $26 billion. The lost profit from carried interest would be: ~$30 x 10%/yr return rate x 20% carry rate x 25% carry split to us x (1 - 20% tax rate) = $120 million/yr in lost profit. Overall steady-state annual carry at KKR is very roughly: $400 billion carry-eligible AUM x 13% average fund return / yr x 20% carry rate x 0.25% carry allocation to KKR shareholders x (1 - 0.2% tax rate) = $2.1 billion/yr. So we lose 6% of potential carry. Not material.
In a good year, this is less than 1 year of growth at KKR. E.g., last year, they raised $130 billion in new client capital, ten times the kind of credit losses we are talking about. It depends on the business cycle, but we’d be losing 6-36 months of growth, which is not material if you agree we originally paid ~half of intrinsic value when we bought the stock in 2023. It is well within the “noise” of the range of outcomes. And it’s unlikely it will even play out this badly since some of KKR’s software investments are already deploying AI without deterioration to their economics (they charge the customers for these new add-on modules, etc.)
Long-term impact — the brand: more importantly, if KKR loses money on software investments, will clients still give them money in the future? Two ways that happens, right? Either (a) investors stop allocating to this industry as they sour on it and go back to stocks and bonds (“substitution” in Porter’s 5 Forces), or (b) they still like alt. assets but they choose KKR’s competitors instead because those guys didn’t lose money but KKR did (“rivalry” in Porter’s 5 Forces).
Consider the facts:
One, performance magnitude. Again, software is only 7% of AUM. Losing 7% of a private equity fund will result in slightly worse performance over a 5-10 year fund lifetime, but it is within the normal variance for these funds. There’s no single, precise way to do the math, but try this. Imagine a $100 fund. Assume on Day 1, 7% of the assets go to $0, leaving you with $93 invested, which then grows over the investment horizon based on KKR’s typical track record. Say we compound at 15% for 5 years. You end with $187, which against $100 is 13.3% compounded. So, our bad software bet reduces fund performance by 1.7%/yr for 1-2 fund vintages. 1.7%/yr is low vs. the typical variance in private equity funds’ performance. Look at fund-by-fund performance differences page 117 of the 2025 annual report. It’s not unusual for a fund vintage to do single-digit returns because it was raised in an ebullient market followed by a crash, only for the next fund vintage (raised at the lows) to be banging out 18% annual returns. This variance won’t impact clients’ views long-term. Further, because it’s low, there are reasonable odds the funds still do OK, because KKR constructed the portfolio well in the first place (7% software exposure, not 50%; there are competing funds holding ~25% software). What if another holding in the fund turns out to be a 3x instead of a 1.5x? So this is well within the noise of normal investing successes/mistakes within a reasonably-constructed portfolio.
Two, competitor impacts. Consider that all competitors will be impacted similarly since (a) most large PE firms own some software investments, and (b) AI will almost certainly affect most software businesses, and in multiple ways. KKR will not look materially different than others, so everyone will look like they made this mistake. There is no reason for clients to single them out. Where will they go instead?
Three, market structure’s unchanged. In private credit, there’s a significant “illiquidity premium” vs. publicly-traded bonds, around ~1.5 percentage points on each loan, even for like-for-like credit risk. The Dodd-Frank Act made it harder and more economically expensive for banks to lend to mid-sized companies, but these borrowers are too small to access bond markets. Private credit stepped in to fill the hole via “regulatory arbitrage” since they have don’t have (a) capital requirements or (b) restrictions on how they can structure loans, like banks do. At worst, institutions sour on these funds for a few years industry-wide, but the economics will force them to get interested again because there is no better alternative. Say you are the CIO at an insurance company, buying bonds and loans. You can get 1.5% more on your private credit loans than you can by buying high-yield bonds for exactly the same credit risk. Why will you stick to high-yield bonds? If you don’t switch, your insurance competitors will. Then they’ll undercut you on insurance policy prices because they can make up the profits on the other side of their balance sheet. You will lose market share to them and the CEO will fire you.
So you have no choice but to get back into some KKR or Blackstone private credit vehicle.
Finally, fourth, this is only a subset of funds. Within that, it’s also mostly the funds to retail clients, KKR’s smallest source of new client money. The real estate funds, infrastructure funds, asset-based lending funds, energy funds, and many others, don’t have software holdings. In fact, AI is helping infrastructure & power, where there’s huge demand for upgrading global electrical grids, etc. I also estimate that funds to HNWIs are <10% of KKR’s inflows. BCRED and BREIT matter to Blackstone, but KKR’s equivalent doesn’t matter to KKR. They want to be a top player with HNWIs and investment advisors, but they currently aren’t, so they’re less impacted. This is a Blackstone problem. A Blue Owl problem. Not a KKR problem.
In all, there’s little reason to believe KKR’s value has been permanently impacted. It will be temporary at worst, and even then the math indicates it won’t be huge.
For an analogous situation, look back 3 years to Brookfield, which is a big investor in retail and office real estate. During the post-COVID retail & office Armageddon, BAM struggled to raise money for new real estate funds because clients were jittery. Today, they aren’t, and inflows returned. Performance is improving. And other BAM funds were unaffected. BAM is much larger now vs. 2023, owing to the facts they (a) invested smartly and didn’t lose much during Armageddon and (b) have a diversified product/fund portfolio, the pieces of which offset each other — some funds well while others do poorly, and they capture inflows regardless of which asset classes the investor sentiment shifts to. There’s no “single point of failure.”
We literally bought Brookfield when it was going through a situation akin to what KKR may now face, because we felt the issue was temporary and manageable in size.
Valuing KKR & The Margin of Safety:
We paid ~8x this company’s steady-state cash earning power in 2023. We knew this industry’s growth will ebb and flow based on client sentiment over multi-year periods, and were being compensated for this. We sought a fat margin of safety, as always.
What about now? Value KKR’s 3 major pieces roughly like this:
(1) Asset management:
(a) Fees — $600b fee-paying AUM x 0.9% fee * 70% margin x (1 - 20% tax) = $3.0 billion fee-based free cash flow
(b) Carry — $380b carry-eligible AUM x 11%/yr fund return rate x 20% GP allocation x (1 - 75% employee allocation) x (1 - 20% tax) = $1.7 billion carry-based free cash flow.
(KKR takes ~20% of funds’ profits, the General Partner allocation. We split that with senior employees; it differs fund-by-fund but will be 75% employees / 25% shareholders, roughly.)
(c) In-house banking (KCM) — ~$800 million investment banking fees x 50% margin x (1 - 20% tax) = $0.320 billion. I normalized 2025’s $1.1b to $800mm for cyclicality.
Total = $3 + 1.7 + 0.3 = $5 billion free cash flow for the asset management business.
(2) Global Atlantic = 15% ROE on ~$10b book value as KKR redeploys capital from bonds into higher-yielding private credit assets = $1.5 billion steady-state free cash flow for the annuity & insurance business.
(3) Investments — Strategic Holdings is likely worth ~$10 billion+ and can do $1 billion free cash flow in a few years. Additional investments are $4.4 billion net of debt, which is mainly investments in KKR’s own funds alongside client money.
If you do a DCF, you’ll find the math comes out similar to below. I’ll use this multiples-based shorthand with you instead of Excel:
Fees = ($3 * 20x to 27x multiple) / 0.900 billion shares = $67 - 90 per share
Carry = ($1.7 * 10x to 15x ) / 0.9 = $19 - 28 per share.
Banking = $0.3 * 20 / 0.9 = $7 per share
GA = 1.5 * 14 / 0.9 = $23 per share
Investments = ($10 + 4.4) / 0.9 = $16 per share
Total value = $67 + 19 + 7 + 23 + 16 = $132 to $164 per share. The stock’s $90; 32-46% margin of safety. We added in 2023, paying $45-55. We may add now at $90 or see if the stock it comes to us. We’re evaluating 3 other competing opportunities, too.
(Note this is below our prior ~$200ish as it doesn’t include reinvestment. Historically, KKR retains ~85% of that free cash flow and reinvests it at 12%+ returns for us, depending what it does. E.g., growing Global Atlantic is a 15% return opportunity. Buying small competitors, when done right, is also 15%. This creates further value since it’s above the cost of capital. The shorthand above only captures this in Global Atlantic’s value. You might also adjust carry down by ~10-15%, as funds typically don’t get fully deployed. There’s no carry on uninvested money.)
You can also look at it like $90 stock price - $23 - $16 = $51 per share implied price for the asset manager only. It will soon do $5 billion free cash flow, 5.6 / 0.9 = $5.50 per share. That’s $51/5.50 = 9.2x free cash flow. This is an absurdly low multiple to pay for a double-digit growth business that earns 50% returns on capital and is one of top ~10 managers globally in both size and broad-based investment performance. And in an industry where all the best competitors earn very high returns on capital because they compete on investment performance, not on product prices. Could you imagine MasterCard trading for 9 times free cash flow? Alphabet/Google has inferior economics (ROICs) to KKR, and has much higher “CapEx / reinvestment risk” (datacenters), and yet trades at a far higher multiple.
Is KKR a 5% growth business next 5 years because clients feel burned by performance? Or will it continue at 20%? Who cares: our investment will earn an attractive return either way, because the price we paid implied no growth at all.
Brookfield Next
We own much more Brookfield. It’s a good time to reiterate why.
First, Brookfield has no meaningful direct software exposure, just indirect exposures through datacenter and semiconductor investments.
(Boy, does that ever make the math easy: 0 x 0 = 0. Done!)
That’s because they focus on durable businesses that form the backbone of the economy. Stuff that isn’t going away. Think power generation & transmission: nothing functions without energy available to perform “work.” Just like you need food for energy, manufacturing plants and offices need water, fuel, electricity, etc. Another Brookfield favorite is logistics. Pipelines, cell towers, freight container leasing, air/ports, railways, and toll roads, etc. Logistics can’t be displaced because goods must move in order to sell. Although logistics businesses will evolve and change form factor, they can only be displaced by teleportation technology. Brookfield stuck to what it knew. That pays off.
Second, and much more importantly, Brookfield’s organizational culture. They’re opportunistic value investors who buy cheap, then they add value by operating well. This is in their blood: they were a conglomerate running real businesses before they were an asset manager. They’re good at both buying cheap and creating operational value; their private equity & other records show this, and make even good investors look bad by comparison. They, like me, prefer to patiently look for what seems cheap, rather than chase what everyone else already likes.
This is also true in their credit business, which is why I’m not concerned with the market’s private credit concerns.
Brookfield bought Oaktree credit management business partly because its principals share the Brookfield culture. They’re defensive and opportunistic. Oaktree tends to deploy more capital during periods of market turmoil than during good times. They are patient. They wait for the fat pitches, only available when everyone is panicking. If software & other sectors are truly screwed, Oaktree (Brookfield) will be there to pick up the pieces. They’re already doing that in some areas, based on a recent interview. In an industry with ~10,000 competitors, others will make mistakes, and Oaktree will capitalize. Oaktree’s approach & culture are surprisingly easy for an outsider to research because its co-founder’s letters to clients are published publicly; here, enjoy this 1600 page compilation of memos from 1990-2025.
If the credit downturn has begun, then I am excited for Brookfield, the quiet guys who do their jobs well. When they come out the other side, they’ll have picked up good stuff for cheap. They’ll be generating good returns for clients, taking market share from guys who messed up, and minting carried interest for us.
Brookfield’s value is relatively unchanged since last we spoke, ~$65-70 USD per share (+/- depending on future execution; including reinvestment). The stock’s $40, a good margin of safety. Every year, they also bolt on ~15% more intrinsic value as the business grows and reinvests at good rates.
This is how I think about these at the moment, in light of all the news flow and fear that is out there in the industry.
— Chris
That’s not certain, though. KKR has claimed in some areas they are the one on the front foot doing the disrupting. Many of their software investments have AI features. That’s the direction every software guy is trying to go right now, in an effort to stave off value capture from new start-ups and existing incumbents. It’s a sensible competitive response if you do it smartly (not out of panic).
E.g., most software firms are adapting, and most software clients are because software is hard to rip out and replace, so it will take years for revenue/clients to slowly churn off at like ~10%/yr
How did you come up with multiples for carry? They seem very aggressive to me.