给我信贷

Memo to:

Oaktree Clients

From:

Howard Marks

Re:

Gimme Credit

The questions I get from clients enable me to understand in real time what’s on their minds. At various points in the last ten years, the most frequently asked question was “when will the Fed raise/cut rates?” During crises, it’s usually “what inning are we in?” For a year or two, it’s been “can we talk about private credit?” And in the last few months, it’s “what about spreads?” Ever since interest rates got up off the floor in 2022, there’s been increased interest in credit, and that’s why I’m devoting this memo to it. It’ll come a little closer than usual to “talking my book,” but I think the subject justifies that. Most of my references will be to high yield bonds, where I have the most experience, there’s the most data, and the fixed coupon rates make the explanations most straightforward. But the points I’ll make are applicable to credit in general. While I’m setting the stage, I want to get one thing out of the way. When people ask me, “can we talk about private credit?” my answer is always the same: “can we talk about credit?” I see no reason why investors should blithely skip over public credit instruments and go straight to private credit. For that reason, I’m going to address both here. Last year was a great one for credit, illustrated by the 8.2% return on the ICE BofA US High Yield Bond Index. That followed even better results in 2023, when the benchmark returned 13.5%. What’s been behind these returns, and where do they leave the credit sector? Background As everyone knows, promised yields on credit instruments were meager in the low-interest-rate period I’ve discussed so much: 2009-21. At the beginning of 2022, before the Fed embarked on its program of interest rate hikes, high yield bonds yielded in the 4% range, with issuance taking place in the 3s and one bond issued in the 2s! I described Oaktree’s challenge at that time as “investing in a low-return world.” The ultra-low bond yields were unhelpful for most institutional investors, and many got out of the habit of investing in fixed income. There was, however, good interest in private credit, where yields in the area of 6% were being levered up to 9% or so. In 2022, investors who feared the Fed’s rate increases would bring on a recession caused the average high yield bond price to incorporate risk protection in the form of a yield spread of more than 4%, taking the overall yield to roughly 9½%. I argued at the time that these promised returns were (a) high in the absolute, (b) relatively safe because of their contractual nature, and (c) well in excess of the returns most institutions targeted. For these reasons, I urged that credit should be weighted significantly in portfolios. These high-single-digit yields alone would have given holders healthy returns. However, investors began to buy because they saw there was good value in credit, and they anticipated rate cuts that would make bonds with high coupons more desirable. Over time, investors also became less worried about a possible recession, and this led to reduced insistence on generous risk protection via credit spreads. Increased demand, lower interest rates, and reduced insistence on risk protection in the form of higher spreads is a perfect formula for price appreciation, and it ensued. This caused the bonds’ total returns to exceed the

2025 Oaktree Capital Management, L.P.

All Rights Reserved

promised yields, and as a result, the high yield bond market delivered an annualized return of 10.8% over the two-year period 2023-24. The flip side of a rising price, of course, is a declining prospective return. As a result of the developments described above, the yield to maturity on the average high yield bond now stands just above 7%, down from 9½%. Just as rising fear and risk aversion cause investments to offer higher prospective returns, rising optimism and risk tolerance lead to lower ones, incorporating reduced yield spreads. (The reduced yield is also attributable to 100 basis points of cuts in the base interest rate.) What Is a Yield Spread? Why would someone lend money to a risky borrower when there are plenty of safe borrowers to lend to? The answer is that risky borrowers pay more for their money, and if you can charge a risky borrower an interest rate that’s high enough to produce a return above that available on safe debt, even after allowing for expected credit losses, it could be worth taking the risk. That was precisely the theory that underpinned Michael Milken’s popularization of high yield bonds in the late ’70s, as well as my career. The differential between the promised yield on risky debt and the yield on a less risky comparator is called a “yield spread,” “credit spread,” or just plain “spread.” It’s also called a “risk premium,” which is what it is: the incremental return you’re offered to accept incremental default risk. Thus, it’s the equivalent of an insurance premium: what policyholders pay to get auto insurers to shoulder the risk that they’ll crash their cars. Yield spreads primarily fluctuate with trends in, and investor psychology regarding, defaults. When more companies are defaulting and investors expect elevated defaults in the future, they’ll demand more protection in the form of wider spreads. They’ll do so to a lesser degree when they’re optimistic about creditworthiness. Thus, the spread is a good barometer of investor psychology, or a “fear gauge.” It’s worth noting the obvious: the spread doesn’t tell you what the actual default rate will be, as some mistakenly say. It tells you what investors think the default rate will be. The thoughtful investor has to evaluate that expression of opinion against what the reality is likely to be and assess whether investors are being too optimistic or too pessimistic. Are Today’s Yield Spreads Adequate? This is the question of the day. Let’s say high yield bonds yield 8% and a Treasury note of the same maturity offers 5%, for a yield spread of 3%, or 300 basis points. Which is the better deal? It all depends on the likelihood of default. If high yield bonds have a 4% chance of defaulting each year and you’re likely to lose three-quarters of your money in a default, your expected annual credit loss is 3% (4% x 75%). If those estimates are accurate, you should be indifferent between the two. Or (holding constant the 75% loss in case of default), you should prefer the Treasury note if high yield bonds are more than 4% likely to default or high yield bonds if they’re less than 4% likely to default. When I managed high yield bonds, I considered the normal range for spreads to be 350-550 basis points. More recently, I think this has been revised to 400-600 bps. Today, however, the yield spread is around 290 bps, one of the narrowest spreads on record since high yield bonds began to be issued in 1977-78. Does that mean investors shouldn’t hold them here? That’s what people mean when they ask me, “can we talk about spreads?”

2025 Oaktree Capital Management, L.P.

2

All Rights Reserved

It’s essential to note that the “normal” spreads mentioned above have proved far more than adequate. We know this because the unmanaged high yield bond indices – even with their defaults and credit losses – have significantly outperformed no-risk Treasurys. Data from Barclays shows that from 1986 through 2024, the 39-year period covered by Oaktree’s record, the annualized return on high yield bonds was 7.83%, compared to 5.14% on 10-year Treasurys. The fact that the average high yield bond gave investors 269 bps more return per year than Treasurys tells us the historical spread was considerably more than sufficient to offset credit losses. Thus, the historical spread shouldn’t necessarily be the standard for adequacy, and investors might intelligently opt for high yield bonds over Treasurys even at spreads below the historical average. Thus, the key question isn’t whether today’s spread is historically narrow or not. It’s whether today’s spread is sufficient to offset the credit losses that will occur. This takes us back to the calculation discussed three paragraphs above. Over the course of Oaktree’s 39-year track record in high yield bonds, from 1986 through 2024, the high yield bond universe’s default rate has averaged 3.5%, and defaulting bonds have cost investors about 2/3 of the money they had at stake, meaning annual credit losses have amounted to about 230 bps (two-thirds of 3.5%). This suggests today’s historically narrow spread of about 290 bps would have been enough to offset the defaults that occurred in the past. Before that’s accepted as the appropriate conclusion on the subject, however, there are caveats to be considered: •

The average default rate of 3.5% overstates the typical experience. That 3.5% average is far from the norm. Out of the 39 years covered by Oaktree’s track record, there were only 14 years when the universe’s default rate was at or above 3.5%, and 25 when it was below. The average was pulled up by double-digit default rates during crises in 1990-91 and 2001-02. If you took out those four years (along with the four best years, in which defaults were 1.0% or less), the average for the remaining 31 years was just 3.0%. Further, the median default rate for the 39 years (the midpoint of the annual observations) was even lower, at 2.7%.

•

The historical default rate might not be relevant to the future. In the Global Financial Crisis of 2008-09 and the Covid-19 pandemic of 2020, central banks and national treasuries showed that they’ve developed tools with which to counter recessions and credit crunches. As a result, the default experiences associated with those events were well below those in the earlier crises, even though the GFC and pandemic were much more serious in a macro sense. Thus, it can be argued that the macro environment has become safer, meaning the historical spreads are no longer called for.

•

The average high yield bond’s credit rating (supposedly an indicator of quality) has risen substantially. Mainly because companies are less concerned about ratings these days, large numbers of investment grade triple-B-rated companies have opted to increase their use of leverage and allow their rating to slip to double-B, the upper tier of the high yield bond universe. The following table shows the change in the ratings profile of the high yield bond universe over the last 25 years:

BB B CCC and below

December 31, 1999

December 31, 2024

32.7% 54.6 12.7

52.6% 33.7 13.7

Data from ICE

2025 Oaktree Capital Management, L.P.

3

All Rights Reserved

Research from Barclays indicates that since the average high yield bond is now higher in creditworthiness, today’s average yield spread provides a good bit more compensation per unit of credit risk today than it did at the “all-time tight” of 2007. •

Active credit managers strive to reduce (a) the incidence of default in their portfolios and (b) the percentage of capital lost when defaults occur. Since the historical spreads have been adequate to protect against average credit losses in the past, that means they’ve proved more than adequate for investors with superior credit discernment. For high yield bond managers with the ability to reduce credit losses through active management, there’s a greater likelihood that spreads will prove sufficient to offset future credit losses.

For all these reasons plus one more, I believe the concern about historically narrow spreads is very much overblown. My additional point is that spread widening is a short-term phenomenon, analogous to volatility in stocks. If the yield spread widens, increasing the demanded yield, that results in a price decline for bondholders. But the price decline is temporary, whereas the higher interest payments are received every year . . . and then the bond eventually returns to par at maturity (assuming it performs). I did some research with Oaktree’s Nicole Adrien to test this thesis. We identified the all-time lowest yield spread on our usual high yield bond benchmark and looked to see how we would’ve fared if we’d bought bonds that day. The lowest spread was 241 bps, reached in June 2007, just prior to the onset of the Global Financial Crisis. Here are the results for high yield bonds and some comparative indices if you chose that time to invest: Annualized Returns Following All-Time Tight U.S. High Yield Bond Spread

1 year 3 years 5 years 10 years 15 years

ICE BofA U.S. High Yield Index

ICE BofA U.S. Treasury Index

Bloomberg U.S. Aggregate Index

-1.13% 5.29 7.26 7.35 6.01

10.19% 7.27 7.20 4.15 3.03

7.54% 6.99 6.83 4.47 3.34

Source: ICE, Bloomberg Note: BofA U.S. High Yield Index all-time tight gov’t OAS spread (241 bps) recorded on June 1, 2007

The one-year return on high yield bonds shows, unsurprisingly, that if you buy a risky asset at the height of its popularity and immediately encounter one of the worst financial crises the world has seen, your initial experience won’t be good. Thus, in the first year following the purchase at the low on spreads, high yield bonds underperformed Treasurys by 11.3 percentage points and the U.S. Aggregate Bond Index by 8.7 percentage points. But note that the high yield bond investor still lost very little money, thanks to the receipt of interest! (At Oaktree, we call this “the power of the coupon.”) High yield bonds didn’t pull ahead of Treasurys and the Aggregate until the five-year mark, but over the 10- and 15-year periods, they outperformed those indices by about 3 percentage points per year despite having been bought at the worst possible moment spread-wise. Of course, managers able to navigate defaults in the high yield universe would have achieved even better returns. As the above data shows, narrow spreads at purchase are far from synonymous with sub-par performance in the medium-tolong term.

2025 Oaktree Capital Management, L.P.

4

All Rights Reserved

The bottom line for me – as I tell anyone who asks – is that you can’t eat spread, or spend spread, or pay pension benefits with spread. For those things, you need returns. Spreads have to be assessed to ensure they’ll be adequate to offset credit losses, but in the end, it’s the total return that matters. Contractual Returns A good part of the reason for the ability of high yield bonds to perform even when spreads have been historically tight, as shown above – and for investors’ ability to ignore the spread tightening that has taken place to date, as I argue – stems from the contractual nature of bond returns. You buy a bond at a given yield to maturity, which could incorporate an anemic yield spread. And if investors decide later to demand increased default protection, the spread will widen and – all else being equal – the price of the bond will decline. But as long as the issuer pays interest and principal as promised, the price decline brought on by spread widening has only a temporary effect. When you’re repaid at par, you’ll have received the yield you expected, regardless of price fluctuations experienced in the meantime, including declines related to spread widening. The bottom line is one that applies to all bonds: if you hold to maturity and the bonds pay, you receive the yield you signed up for. I’ve written so much about this that I’m not going to belabor it further (see my memo Ruminating on Asset Allocation, October 2024), but I’m always available to talk. (Before the bond pros jump down my throat, I’ll admit that the foregoing is less than 100% accurate. There are three components in bond returns, not two. Everyone knows about the interest payments and the movement of price to par at maturity. But there’s a third: the interest earned from reinvesting the annual interest payments, better known as “interest on interest,” and thanks to the power of long-term compounding, this is a major matter on 20- or 30-year bonds. The standard yield-to-maturity calculation assumes interest receipts are reinvested at the yield in effect at time the calculation is performed (for example, at purchase), but that’s a simplifying assumption, and the reality may well be different. No one wants to see the price of a bond one owns decline. But the truth is that if the bond price declines, the yield rises, meaning interest payments received can be reinvested at a higher rate than was anticipated. Thus, surprisingly, interim price declines can raise the overall return earned from holding a bond to maturity.) What About Private Credit? This is today’s other FAQ, along with the one about spreads. A lot of people have questions about private credit, which makes one wonder how the sector can be seeing such strong capital inflows. My responses generally go like this: •

Like anything else, there are pros and cons. The most obvious pro is that, to compensate for the lack of liquidity, private credit offers higher yields than public credit. The second is that private credit managers are able to offer funds (and thus returns) that are levered, which isn’t true of most public credit funds. The main negative stems from the absence of a market for the loans, and thus their illiquidity and the difficulty of actively managing holdings. Further, because there’s no market, private credit can’t actually mark to market. A final negative is that the fees are higher on private credit investing than on public credit, often including an incentive fee.

•

What about the lack of marking to market, and the resulting low level of volatility? It’s obviously unrealistic to think the value of private loans doesn’t fluctuate. But on the other hand,

2025 Oaktree Capital Management, L.P.

5

All Rights Reserved

it seems many people consider the non-marking to market a plus, in that they can report that their investments didn’t go down much in a difficult environment. Private credit managers are supposed to mark their holdings to reality based on fundamentals, but that’s clearly less volatile (and less objective) than marking to a market. On the other hand, is it desirable that public asset prices reflect every up and down of investor psychology? Not marking to market may be unrealistic, but it may be welcome. (Investors in public securities could have the same experience if they refused to read the newspapers and tossed their brokerage statements in the drawer, but such behavior would be called irresponsible.) •

For me, the most important observation about private credit is that it mostly emerged since 2011 in response to banks’ reduced lending activity after the Global Financial Crisis. Since then, the economy has witnessed an unusually long string of years without a recession (if you don’t count the two-month Covid 19-related recession that flared up and was reversed in mid-2020). To paraphrase Warren Buffett, the tide has never gone out on private credit, meaning we haven’t had an opportunity to see its flaws. As far as I’m concerned, the main one is the possibility that some managers have been in such a hurry to scoop up capital and put it to work – so they could come back for more – that they relaxed their credit standards and failed to demand a sufficient margin of safety. If there’s ever another difficult period in the economy and the market, we’ll see the result. Note: this isn’t a sweeping concern about the loans themselves, just a question about the behavior of individual managers.

•

Connected to the above (and to the absence of marking to market), we don’t know what’ll happen if and when a difficult environment does arrive. Is there a limit on the ability of managers to keep marks too high? Is it right for fund returns to ignore deteriorated fundamentals? Can managers avoid recognizing credit difficulties by granting forbearances and “kicking the can down the road”? For how long? Are there ill effects on fund investors in the meantime? Since private credit managers are mostly unregulated, will the truth come out? Which truth? Questions like these also are answered only when the tide goes out.

•

Lastly, I don’t believe private credit represents a systemic risk. People have been on the lookout for systemic risk ever since the GFC, in which troubled banks brought trouble to other banks and took them down. My belief is that the risk in private credit isn’t systemic, since (a) private loan portfolios and their owners aren’t levered nearly as much as banks were in 2007-08 and (b) there isn’t the same level of interconnectedness, or “counterparty risk,” since the holders haven’t sold each other default protection and other forms of hedging, like banks did before the GFC. There are those who believe some holders of private credit have multiple layers of leverage, which could increase the risk in a downside scenario, but I have no way of knowing.

The bottom line for me is that the return premium on private credit relative to public credit seems roughly fair given the merits. Extra return is a good thing, but the downside related to the lack of liquidity and resulting difficulty in actively managing holdings is a real consideration. All else equal, I would suggest employing a combination of the two. Credit Versus Equities I’ve written about equity valuations – primarily referencing the Standard & Poor’s 500 – as recently as this January in my memo On Bubble Watch. Suffice it to say that, according to past data, from p/e ratios like today’s, the S&P has historically produced ten-year returns averaging between -2% and 2% per year,

2025 Oaktree Capital Management, L.P.

6

All Rights Reserved

and some investment banks have expressed expectations that are similarly in the low to mid-single digits. Obviously, today’s expected returns on credit are considerably higher. On January 27, an article on the front page of The Wall Street Journal said the following: “Stocks haven’t looked this unattractive, by at least one measure, since the aftermath of the dot-com era.” This wasn’t a reference to the elevated p/e ratio, but to the fact that the yield on the 10-year U.S. Treasury note is higher than the “earnings yield” on the S&P 500 stock index. (The earnings yield is the ratio of earnings to price, the inverse of the p/e ratio.) This doesn’t prove that bonds are going to beat stocks in the years ahead, but it’s one more argument. And if Treasurys are poised to out-yield the S&P 500, high yield bonds will do so to an even greater extent (assuming credit losses don’t exceed the historical experience). As I’ve written in other memos recently, the current level of offered yields implies higher returns from credit than the S&P 500, with returns that are contractual and thus subject to much less variability and uncertainty. This is true despite the return contraction that has been brought on by the swing from pessimism to optimism over the last two years, and even given today’s narrow spreads. *

*

*

The bottom line is that credit presently offers a better deal than equities (to the extent the S&P 500 is representative of equities), even at today’s spreads. Credit isn’t a giveaway today, but it offers healthy absolute returns and is fairly priced in relative terms. This is true despite the narrowness of yield spreads. These observations aren’t limited to high yield bonds. They also apply to senior loans, mezzanine debt, asset-backed loans, CLOs, and private lending. We’d rather buy at higher yields and wider spreads, and we may get a chance to do so . . . or not. But that preference in itself isn’t a reason for not increasing allocations to credit today. March 6, 2025

2025 Oaktree Capital Management, L.P.

7

All Rights Reserved