Why Long-Duration Stocks Are Rate Sensitive
A growth stock is a long bond with an earnings option attached. Where the present value lives on the timeline — not the sector label — determines how hard a change in yields hits the price.
Most investors know that growth stocks are rate sensitive, and most of them hold the knowledge the way one holds a slogan — true, repeated, and unusable. Asked why, the common answer is some version of higher rates hurt valuations, which explains nothing: every asset is discounted, so why does the same yield move re-rate one equity by a rounding error and cut another by a third? The difference is not quality, sector, or narrative. It is where the present value lives on the timeline — and that single variable can be reasoned about with almost bond-like precision if you are willing to take the discounting arithmetic seriously.
The framing we use: a growth stock is a long bond with an earnings option attached. The bond part is the schedule of expected cash flows and the rate they are discounted at; the option part is the possibility that the schedule itself improves. Rate sensitivity — duration, in fixed-income language — belongs to the bond part, and it is governed by a simple question: how far in the future does the value sit? An equity earning most of its value from cash flows beyond year ten has more duration than any Treasury bond an investor can buy, because unlike the bond it has no maturity, no fixed coupon, and no pull to par. What it has instead is the option — and the interplay between the bond part and the option part is the entire subject of this piece.
The arithmetic: why distance amplifies
Discounting is compounding run in reverse, and compounding is not linear in time. A cash flow one year out barely notices a change in the discount rate; a cash flow twenty years out is divided by that rate compounded twenty times, so a small change in the rate produces a large change in what the distant flow is worth in present terms. Sum a company's cash flows and the sensitivity of the total is a weighted average of the sensitivity of each flow, weighted by how much of the value it carries. That weighted average is equity duration. Two businesses with identical total expected cash flows can therefore have wildly different rate sensitivity purely because of when the cash arrives: front-loaded value is short duration, back-loaded value is long duration, and no amount of narrative changes the arithmetic.
This is why the label growth stock is really a duration label wearing a style costume. A company reinvesting everything today in exchange for large cash flows later has, by construction, moved its value out the timeline. The faster the assumed growth and the smaller the current free cash flow, the larger the share of present value sitting in the terminal years — for a young company priced on distant profitability, effectively all of it. Meanwhile a mature business converting revenue to cash now is the equity equivalent of a short-dated bond: most of its value is near, so the compounding lever the discount rate works with is short. Profitability, in this framework, is not a quality signal — it is a duration signal. Near-term free cash flow is the equity's coupon, and coupons shorten duration.
Three layers of why
That final step is where this framework earns its place in a portfolio process rather than a textbook. Sterling's Embedded Intelligence treats equity duration as a portfolio-level property to be measured, not a stock-level anecdote: when long real yields move, Sterling's first question is not which holdings are growth stocks but where the portfolio's aggregate present value sits on the timeline — because that is the number the yield move actually multiplies against.
The illustration: the same businesses at two discount rates
The cleanest modern teaching case is the round trip of 2020 through 2022, stated as history. In 2020, policy rates went to zero and long-term real yields fell to deeply negative levels. The longest-duration equity — unprofitable software, speculative platforms, businesses priced entirely on cash flows a decade or more away — re-rated spectacularly, far beyond what any near-term fundamental change could justify. The arithmetic explains what the euphoria narrative does not need to: when the real discount rate approaches zero, the present value of very distant cash flows expands enormously, and it expands most for the assets whose value is most distant. The multiple did the work.
Then the discount rate came back. Through 2022, long real yields rose from deeply negative territory to solidly positive levels, and the same long-duration cohort de-rated as violently as it had re-rated — drawdowns that in many cases dwarfed anything happening to the underlying businesses. That is the tell worth internalizing: for many of these companies, revenues kept growing right through the de-rating. The cash-flow schedule changed far less than the price, because the price move was never mostly about the schedule. It was the denominator repricing a numerator that sat very far away. Investors who attributed the 2020 gains to business quality and the 2022 losses to sentiment were reading duration effects as fundamentals in both directions — the same error, symmetric, and entirely predictable from the timeline of the cash flows.
The strongest case against this framework
The serious objection is empirical: over long stretches of history, growth equity has performed well through periods of rising rates, and the mechanical duration story would have kept an investor out of some of the great compounders of the modern era. If long-duration stocks are simply short bonds in disguise, the objection runs, why has the disguise so often outperformed while yields climbed?
The resolution is the most important sentence in this piece: equity duration is a partial derivative, and the market only ever delivers total derivatives. The duration framework tells you what happens to price when the discount rate moves holding the cash-flow schedule fixed. In live markets the two move together, and the correlation depends on why rates moved. When yields rise because real growth is accelerating, the same force that lifts the discount rate lifts the expected cash flows — the earnings option strikes — and the numerator effect can fully offset the denominator effect. When yields rise because of policy restriction or a rising premium on duration itself, there is no cash-flow offset, and the raw duration math bites at full force. The objection is not evidence against the framework; it is evidence that the framework must be applied to decomposed rate moves, never to headline yield changes. Growth-driven yield rises and policy-driven yield rises are opposite events for long-duration equity, and the historical record mixes them together.
A second objection deserves brief standing: duration cannot be computed precisely for an equity, since the cash-flow schedule is itself an estimate. True — and not a defect. The framework's output is an ordering, not a decimal: this portfolio carries more timeline risk than that one, this holding's value sits mostly beyond year ten, this one's mostly inside year five. Orderings are enough to size and hedge with. False precision is not required, and pretending equities have bond-grade duration statistics would fail our own epistemic standards.
Auditing a portfolio for hidden duration
The practical output is a duration audit, done at the level of honesty rather than spreadsheet theater. For each material holding, ask where the value lives: roughly how much of what you are paying is covered by cash generation over the next five years, and how much depends on flows beyond that? Holdings that fail the five-year coverage question are long duration regardless of sector, and a portfolio full of them has a single dominant factor exposure — the long real yield — sitting underneath all its idiosyncratic theses. The point of the audit is not to shed that exposure; long-duration equity is where much of the market's innovation and compounding lives. The point is to own it knowingly, size it as the rates position it is, and stop being surprised by the mechanism twice a cycle.
How to apply this framework
- Ask where the present value lives before asking anything else about rate risk. Value covered by near-term cash generation is short duration; value dependent on flows beyond five to ten years is long. The label growth or tech is a costume — the timeline is the exposure.
- Price long-duration equity off the long real yield, not the policy rate. Distant cash flows are discounted at term, and for businesses with pricing power the inflation component largely passes through. The ten-year real yield is the single most relevant market price for the long-duration part of a portfolio.
- Decompose every yield move before mapping it to equities. Growth-driven rises come with an earnings offset and have historically been survivable or better for growth equity; policy-driven or premium-driven rises have no offset and deliver the raw duration math. Same headline move, opposite implication.
- Aggregate duration at the portfolio level. Twenty independent stock theses can sum to one rates trade. If the aggregate present value of the portfolio sits far out the timeline, it is short the long real yield as surely as any bond position — and should be sized and stress-tested as one.