Credit Carry and the Spread Premium
A corporate bond pays more yield than a government bond of the same maturity, and most of that extra yield is compensation investors collect for a default risk that, most years, never actually happens.
Prerequisites: Carry Across Asset Classes, The Merton Structural Model of Default
Lend money to a government and it pays you the risk-free rate. Lend the same amount to a corporation for the same number of years and it pays you more — sometimes a little more, sometimes a lot more. That extra yield is the credit spread, and collecting it, quarter after quarter, for as long as the borrower keeps paying, is credit carry.
Credit spread is priced to compensate a lender for expected defaults plus a premium for bearing that risk. Because actual defaults are rare in most years, credit carry behaves like an insurance premium: steady income most of the time, with occasional sharp markdowns when defaults or downgrades cluster.
Where the spread comes from
A credit spread is not pure compensation for expected loss — it is larger than that, and the gap between the spread and the expected loss it should cover is called the credit risk premium. Two components make it up: the actuarially fair price of expected defaults (probability of default times loss given default) and an extra premium investors demand for bearing a risk that is skewed — mostly small, steady gains, occasionally a sudden large loss — and that tends to show up exactly when other risky assets are also falling, which is the worst time to need the money back. This is the same reasoning behind the Merton view of credit as a bet the firm's assets stay above its debt: the equity holders effectively hold a call option on the firm's value, and bondholders are short a put.
Shift this toward a skewed shape in your head: unlike the symmetric bell curve shown, a credit portfolio's return distribution has a short right tail (spreads can only compress so much) and a long left tail (a default can wipe out most of the position at once) — that asymmetry is exactly what the spread is paid to compensate for.
Worked example
A 5-year investment-grade corporate bond trades at a spread of 120 basis points over the 5-year government bond. The bond's rating implies a historical annual default probability of about 0.4%, with a typical recovery rate of 40% on default, so loss given default is 60%.
That is 24 basis points of expected loss per year. Against a 120 basis point spread, that leaves 120 − 24 = 96 basis points as the credit risk premium — the part of the spread that is not just covering the actuarial cost of defaults, but paying the investor for bearing the risk and illiquidity of holding it. Collected every year the bond does not default, that premium compounds into a meaningful return; the one year it does default, the position can lose the majority of its value at once.
What this means in practice
Credit desks and credit-focused funds explicitly separate a bond's spread into "expected loss" and "risk premium" pieces, because the risk premium is what a manager is actually being paid to bear once ratings-implied default rates are backed out. Diversifying across many issuers reduces idiosyncratic default risk, but does little to protect against systemic credit cycles, when downgrades and defaults across an entire sector or economy rise together — exactly the environment where credit carry strategies give back years of collected spread in a matter of months.
A wide credit spread is not automatically "cheap" — spreads widen precisely when default risk is rising, so a high-carry credit trade during a weakening economy can be collecting a premium that is about to be tested rather than one that is safely banked.
Related concepts
Practice in interviews
Further reading
- Berd, ed., Lecture Notes on Credit Risk and Credit Derivatives