Our conversation with Carl Kaufman, Co-CIO of Osterweis Capital Management, was one of the most grounded and practical fixed income discussions we have had on Excess Returns. Carl has been managing the flagship Osterweis Strategic Income Fund for more than twenty years, through multiple recessions, liquidity crises, and interest-rate regimes
What stood out is how often Carl’s lessons run counter to the conventional wisdom that dominates fixed income today. Many investors view bonds as a boring area of the market. In reality, the structure of credit markets has changed dramatically. High yield quality is rising, investment grade quality is falling and private credit has grown from a niche strategy into a much bigger area. Investors who rely on old models of fix income will be very surprised about the market reality today.
Here are the five lessons that resonated most from our conversation.
Lesson 1: The High Yield Market Is Higher Quality Than Ever, and Investment Grade Is Lower Quality Than Ever
One of Carl’s most surprising observations is how much the credit landscape has changed over the past decade. The traditional view of bonds is simple. Investment grade is the safe part of the market. High yield is the risky part. The reality today is more complicated.
Carl walked us through the numbers. Roughly half of the investment-grade market is now BBB, the lowest quality within that category. Meanwhile the high yield market has migrated upward. More than half of high yield is now rated BB, the top tier of non-investment grade. That shift happened because weaker companies went elsewhere.
As Carl told us, “The high yield market is probably the highest quality it has ever been. And the investment grade market is the lowest quality it has ever been.”
Two forces drove the change. First, private credit absorbed many of the riskiest borrowers. Second, the leveraged loan market captured most of the next tier of credits. What is left in traditional high yield today is a more resilient set of issuers with better balance sheets than most people assume.
The implication is straightforward. The old playbook for allocating between investment grade and high yield does not reflect the current market structure.
Lesson 2: Private Credit Is Where Most of the Trouble Will Show Up
While high yield has improved, the opposite has happened in private credit. The asset class has exploded in size. It has attracted waves of capital. And it has become the preferred home for companies that can no longer access the public markets.
Carl is blunt about the risks. “Most of the defaults are going to be in private credit,” he told us. “Defaults in private credit are already running close to ten percent.”
Two issues stand out.
First, many newer private credit managers lack the sourcing networks and underwriting experience of the established players. They raised large funds and needed to put money to work. Borrowers knew this and forced lenders to compete away protections. As Carl explained, “These are typically floating-rate loans with very few covenants, and borrowers put lenders in competition for terms.”
Second, the opacity of the market hides developing problems. Loans are not marked to market. Investors are locked up. Information moves slowly. That delays the recognition of losses and makes systemic stress more likely.
Carl does not believe private credit will trigger a 2008-style crisis, because the business models are so varied. But he does think the highest risk currently lies there.
Lesson 3: Bond Indexes Are Built Backward
Equity investors often assume fixed income indexes are built the same as equity indexes. That assumption breaks down quickly once you understand how credit benchmarks work.
Index construction in bonds weight the biggest issuers just like equity indexes, but what constitutes a big issuer is very different. The biggest weights go to the companies that borrow the most, not the companies with the biggest market caps. In other words, the most levered issuers dominate the index.
Carl summed it up simply: “In fixed income you get the largest weights through excess borrowing, not success.”
This creates two problems. Benchmark-oriented managers must own large positions in the most indebted companies. And they must own them across multiple tranches. Carl estimates many traditional bond funds hold three hundred to seven hundred positions, often including bonds they would never buy if they were not benchmark constrained.
His team chooses the opposite path. They run a concentrated portfolio of about 120 companies. They do deep credit work. They speak directly with management. And they rarely own more than one or two bonds from each issuer.
They also lend only to businesses with a need to exist. As Carl put it, “If they went away tomorrow, would anyone notice?” That single question eliminates a huge portion of issuers, especially in commodity and highly cyclical industries.
When you put these pieces together, the takeaway is clear. Benchmark indexing works well in equities, but bond indexes are structurally flawed. Skill, selectivity, and fundamental research matter far more in fixed income than in stocks.
Lesson 4: Understanding Cycles is Essential
Carl’s framework is built on two cycles, not one. The interest rate cycle and the credit cycle. Most investors collapse these into a single macro view. Carl separates them.
In an expansion, credit does well and interest-rate-sensitive investment grade does poorly. In a recession or stress period, high quality investment grade and Treasuries outperform while lower-quality credit struggles.
The nuance is that today’s cycles do not behave exactly like the textbook. Because high yield is higher quality and investment grade is lower quality, the traditional shift into investment grade at the first sign of trouble has become less compelling.
But the bigger lesson is Carl’s view on duration. He uses short-term bonds and cash-like instruments as a defensive bucket, allowing the fund to build liquidity without sacrificing yield. In late 2019 and early 2020 this defensive bucket reached thirty-five percent, which allowed the team to buy credit aggressively during the COVID selloff.
They are doing something similar today. “We are up to about forty percent in short-term defensive assets,” Carl told us. That ability to pivot and adjust positioning is deeply rooted in an understanding of cycles.
Lesson 5: Not All Credit Cycles Are Created Equal, and the AI Buildout Shows Why
One of the most interesting moments in the conversation came when we asked Carl about the recent wave of debt issuance tied to AI infrastructure. Meta, Oracle, and others have raised large amounts to build data centers and GPU facilities.
Carl’s reaction was immediate: “It gives me the willies,” he said.
He drew a direct comparison to the late 1990s. Cisco was the Nvidia of its day. It sold the picks and shovels of the Internet boom. And late in the cycle, Cisco began financing its customers so they could buy more equipment. Nvidia is now doing the same thing through joint ventures.
The sheer scale of the planned spending is also concerning. “They are talking about five and a half trillion dollars of investment,” Carl noted. “Trillions do not grow on trees.”
Companies will need extraordinary revenue growth to earn acceptable returns on that level of capital spending. History shows this rarely happens. Telecom companies in 2000 and fiber companies in the mid-2000s spent heavily, only to see prices collapse when capacity exceeded demand.
Carl does not deny that AI is transformative. He does not doubt the technology’s importance. But he has seen enough cycles to recognize what happens when capital intensity outruns economic return.
For credit investors, the lesson is sharp. The early stages of a boom are often safe. The late stages are almost always dangerous.
The Bottom Line: Fixed Income Has Changed, but Fundamental Discipline Still Works
Carl Kaufman’s perspective is refreshing at a time when many investors want simple narratives about the bond market. Rates are falling or they are rising. Credit is safe or it is risky. Private credit is the future or it is a bubble. The reality is far more nuanced.
Carl’s approach is built on three pillars.
• Understand where you are in the cycle.
• Lend only to businesses that can survive bad times.
• Keep enough liquidity to act when others are forced to sell.
It is not flashy. It is not macro forecasting. And it is not indexing. It is credit work, patience, and flexibility applied consistently over decades.
If there is a unifying theme across our conversation, it is this: fixed income is not the sleepy corner of the market many assume it to be. It is dynamic, unevenly structured, and full of both opportunity and pitfalls. But with the right principles, it can be a powerful source of steady, compounding returns.
Watch the full episode here: