Investing Classics: Bonds & RatesThe Gravity Behind Every Asset Price

Duration turns the abstract "interest-rate risk" into a computable number: a bond with duration 8 falls roughly 8% in price if yields rise 1%. The longer the maturity and the lower the coupon, the larger the duration. Convexity says the relationship isn't a straight line — when yields fall sharply, price rises a bit more; when they rise, it falls a bit less. Convexity itself favors the holder, but very long bonds magnify both convexity and volatility. The core insight: a zero interest rate doesn't abolish risk; it merely relocates it into duration. You think you bought a "safe bond"; you're actually holding a long lever that rates can pry.

In 2020 Austria issued a 100-year government bond with a coupon under 1% and enormous duration. As central banks hiked fast in 2022, this "century bond" lost roughly half its price within two years — a drop comparable to a stock-market crash. That same year the UK's pension LDI strategy held leveraged, very-long-duration gilts; in September gilt yields spiked within a week, triggering a "margin call → forced selling → yields rise again" death spiral that forced the Bank of England to step in and buy bonds. The Bloomberg US Aggregate Bond Index fell about 13% in 2022 — its worst year on record.

Duration isn't inherently bad — in an easing cycle, long duration is a tool for amplifying gains; the problem is always direction and leverage. Common misuses: ① equating "bond" with "safe" and never asking about duration; ② reaching for a sliver of extra coupon on a very long bond and taking on huge duration; ③ using leverage (LDI/repo) to amplify duration, turning manageable volatility into forced-liquidation risk.

• Do I know the duration of the bonds or bond funds I hold? If yields rise 1%, how much do I lose?
• For that extra bit of coupon, exactly how much duration am I taking on — is it worth it?
• Is there leverage in my bond exposure? Could I be forced to liquidate in an extreme move?
• Do I treat bonds as "risk-free," or do I understand where their price swings come from?

Normally the long end yields more than the short end, because a longer horizon demands more compensation for uncertainty (the term premium). When central-bank hikes push the short end up while the market, expecting a weakening economy and forced cuts ahead, pushes the long end down, the curve inverts. It is a strong signal precisely because it aggregates countless participants' bets on whether the central bank will be forced to reverse course. An inversion isn't the cause of a recession; it's the market's collective verdict that tightening has gone too far — a weather map drawn with real money.

In 2006–07 the 10-year/2-year spread inverted, and the 2008 global financial crisis followed. In August 2019 the curve briefly inverted, and recession arrived in 2020 (though triggered by COVID). From mid-2022 into 2024 came the deepest and longest inversion since the 1980s. But note two limits: the lag from inversion to recession can run 6–24 months, and quantitative easing has artificially suppressed the long end's term premium, making this ancient signal harder to read.

The curve is a probabilistic signal, not a timing tool. Common misuses: ① selling everything the moment it inverts — historically, stocks have often kept rising for more than a year after an inversion; ② ignoring how central-bank QE/QT distorts the long end and treating a suppressed long rate as a pure market signal; ③ treating "eight-for-eight in the past" as "guaranteed this time."

• Am I treating the curve as a "probable signal that will eventually pay off," or as a "act tomorrow" timing order?
• How much of today's long rate is distorted by central-bank buying and the term premium?
• How long has the inversion lasted? What does the historical lag say about the patience window?
• Is my decision based on the curve itself, or on my wishful reading of it?

The spread is the credit market's "fear-and-greed index." In booms, risk aversion vanishes, everyone chases yield, and spreads get squeezed razor-thin — you're bearing default risk for pitiful compensation. In panics, spreads explode, which is exactly when compensation is richest and opportunity greatest. Marks insists over and over: what decides your gains and losses is never "what you bought," but "what you paid". The same batch of high-yield bonds is a trap at a 250-basis-point spread and an opportunity at 2,000.

In 2007, US high-yield spreads were squeezed to a historic low of about 250 basis points, and Marks warned in his memo The Race to the Bottom that "risk aversion is disappearing." A year later Lehman fell, and high-yield spreads blew out to roughly 2,000 basis points by November 2008. Oaktree deployed about $6 billion into distressed debt against the panic, producing one of its best vintages ever. As the spread swung from greed to fear, Marks caught the other end of the pendulum.

A wide spread is no free lunch — spreads widen in recessions alongside genuinely rising defaults, and "cheap" can be exactly the value trap. Common misuses: ① looking only at the absolute spread level, not where the default cycle stands; ② giving up safety for a sliver of excess yield when spreads are ultra-tight; ③ turning "contrarian" into "catching a falling knife," bottom-fishing too early while fundamentals keep deteriorating.

• Is the current spread compensating me for risk, or am I paying for someone else's greed?
• Is the default rate implied by this spread reasonable? Where does it sit versus historical cycles?
• Am I buying a "cheap good asset," or a "bad asset that merely looks cheap"?
• When spreads widen, have I kept cash ready to catch the other end of the pendulum?

Any asset's value is the discounted stream of its future cash flows, and the bedrock of the discount rate is the risk-free rate. When rates fall, the denominator shrinks, the same cash flow is worth more, and stocks, bonds, real estate and growth stocks are all lifted together; when rates rise, the tide runs out. Over the past forty years (1981–2021), the US 10-year Treasury yield fell from around 15.8% all the way to about 0.5% — a century-scale tailwind that lifted nearly every asset. Howard Marks calls it a "Sea Change," and warns that this tailwind may already be over.

2022 was a lesson in gravity's return. The Fed hiked at the fastest pace in nearly 40 years, pushing the policy rate from 0 to above 5.25% in 16 months. The result was stocks and bonds falling together: the Nasdaq dropped about 33%, long-term Treasuries about 30%, and the classic 60/40 portfolio had one of its worst years in a century — because when gravity (the rate) suddenly strengthens, everything previously held aloft by zero rates sinks at once. For contrast, in the late 1970s Volcker pushed rates to nearly 20% to tame inflation, and that too weighed heavily on asset valuations.

The danger isn't rates themselves — it's mistaking the past forty years' tailwind for a permanent norm. Common misuses: ① valuing growth stocks with an ultra-low discount rate, assuming low rates forever; ② treating the high returns of 2009–2021 as a repeatable benchmark; ③ admiring "a great company" in isolation without asking "what rate does this price assume."

• In my valuation model, does the discount rate reflect today's rates, or an old habit of assumption?
• Are my return expectations built on top of one special easing-tailwind era?
• If rates stay higher for longer, which part of my portfolio is most fragile?
• Am I paying for a good business, or paying for the hidden assumption that "rates will always be low"?