A rate-cycle rabbit hole · Oct 2026
The textbook is clean: higher rates discount far-off tech earnings harder, so the Nasdaq should be first to wilt when the Fed tightens. I split 27 years of returns by what the Fed was actually doing and checked whether the Nasdaq-100 really underperforms the S&P 500 when rates are rising.
Credit where it’s due: the logic is airtight on paper. A stock is the present value of its future cash flows, and the Nasdaq’s cash flows sit further out than the market’s — more growth, less dividend, longer duration. Raise the discount rate and the most distant dollars lose the most value, so when the Fed hikes, high-duration tech should fall hardest. It’s one of the most repeated ideas in markets, and 2022 looked like a clean proof.
But “the Nasdaq underperforms when rates rise” is a precise, testable claim. I took the 3-month Treasury yield as a stand-in for Fed policy, tagged every month since 1999 as rising, falling or flat, and compared the Nasdaq-100 to the S&P 500 in each bucket. The answer is the opposite of the theory — with one important asterisk.
Nasdaq-100 vs S&P 500 per year, while the Fed was raising
QQQ compounded at +11.7% a year against SPY’s +6.6% in rising-rate months since 1999. The Nasdaq won, it didn’t wilt.
Same edge, equal-weighted
Strip out the mega-caps and the typical Nasdaq stock still beat the S&P by the identical margin in hikes. Not a Magnificent-Seven illusion.
The 2022 asterisk (QQQ vs SPY, calendar year)
QQQ −33% against SPY −18%. When a rate shock hits expensive tech, the theory bites hard — it just hasn’t been the average.
“The Fed is raising” needs a definition that doesn’t cheat with hindsight. I used the 3-month Treasury yield — which tracks the Fed’s policy rate almost exactly — and looked at where it sat versus six months earlier. Up more than a quarter-point: raising. Down more than a quarter-point: cutting. In between: flat. That quarter-point buffer keeps the long stretches of near-zero rates from registering as noise.
It splits the 331 months since QQQ began trading into 94 raising, 162 flat and 75 cutting — and it lines up with the cycles you’d name by hand: the hikes of 1999–2000, 2004–06, 2015–18 and 2022–23. No cycle dates are hand-picked; the rule tags each month on its own.
If the duration story drove returns, the Nasdaq would lag the S&P whenever the rate line is climbing. It doesn’t. In rising-rate months the Nasdaq-100 compounded at 11.7% a year against the S&P’s 6.6% — a 2.6-point annual edge, and it beat the S&P in 53% of those months. Returns are lower in hikes than in calm markets, yes (11.7% vs 19.5%) — but they’re lower for the S&P too. The Nasdaq’s relative advantage doesn’t shrink when the Fed tightens.
| Regime | Months | Nasdaq-100 | S&P 500 | Nasdaq edge/yr |
|---|---|---|---|---|
| Raising | 94 | +11.7% | +6.6% | +2.6 |
| Flat | 162 | +19.5% | +17.4% | +1.2 |
| Cutting | 75 | +3.9% | −1.7% | +2.5 |
Annualized total return, 1999–2026
$NDX and $SPX indices back to 1985 (price only, which slightly flatters the lower-dividend Nasdaq) tells the same story: a +3 to +4-point annual edge in rising-rate months across four decades and six hiking cycles. The direction of the result doesn’t depend on the QQQ era.Here’s the tell that the rate story is a red herring: the Nasdaq’s edge over the S&P is about the same size no matter what the Fed is doing — +2.6 points a year in hikes, +1.2 in flat regimes, +2.5 in cuts. If rate direction mattered, the edge would swing from negative in hikes to large in cuts. Instead it’s a roughly flat line. The Nasdaq out-returns the S&P because its companies grew earnings faster over these 27 years, and that secular edge mostly ignores the rate cycle.
Nasdaq-100 minus S&P 500, annualized, by regime
There’s a fair objection: “the Nasdaq-100” today is really a handful of trillion-dollar names. Maybe the giants muscled through the hikes while the rank-and-file tech stocks got clobbered exactly as the theory predicts. So I rebuilt the index equal-weighted — every member the same size, rebalanced monthly, survivorship-safe — which tilts hard toward the smaller, higher-beta, longer-duration names the theory should punish most.
It barely moves the hiking-regime answer. Equal-weighted, the Nasdaq-100 still beat the S&P by +2.6 points a year while rates rose — identical to the cap-weighted figure. The mega-cap weighting flatters the Nasdaq in flat and falling-rate regimes (where the giant compounders shine), but in hikes the whole index, big and small, outperformed.
Nasdaq-100 minus S&P 500 per year, cap-weight vs equal-weight, by regime
Which brings us to the asterisk. The thesis isn’t wrong — it just needs a genuine shock to show up. In calendar 2022, the sharpest hiking year in four decades, the cap-weighted Nasdaq fell 33% against the S&P’s 18%, a 14-point rout that is the single best piece of evidence the duration story has. But look at what equal weight did that year: −24%, a full nine points better than cap weight. The pain concentrated in the biggest, most expensive names — the opposite of the usual “mega-caps are the safe part” intuition.
In a real rate shock it’s the largest, most-loved names that reprice hardest, because they’re the ones priced for the most distant growth. Equal weight was the safer version of the Nasdaq in 2022, not the riskier one.
The duration hit lands on the priciest names first.Zoom out to the four full hiking cycles and the average hides a wide spread. The Nasdaq crushed the S&P through the 1999–2000 melt-up, lost to it in the slow 2004–06 cycle, and won the 2015–18 and 2022–23 rounds. One of four went the theory’s way.
Cumulative total-return difference over each cycle, percentage points
| Cycle | Nasdaq-100 | S&P 500 | Difference |
|---|---|---|---|
| 1999–2000 | +63.1% | +11.3% | +51.9 |
| 2004–2006 | +7.7% | +17.7% | −9.9 |
| 2015–2018 | +38.3% | +26.6% | +11.6 |
| 2022–2023 | +13.5% | +9.0% | +4.5 |
So does the Nasdaq hate a rate hike? On the tape, no: since 1999 it has out-returned the S&P by about two and a half points a year while the Fed was tightening, and equal-weighting — the fairest test of “the typical Nasdaq stock” — says the same thing. The textbook duration effect is real, but it’s a second-order force that gets swamped by the first-order fact that these companies grew faster. It surfaces as relative pain only in a genuine rate shock like 2022, and even then it’s the priciest mega-caps that take the worst of it. Rate direction, on its own, is a poor reason to tilt toward or away from the Nasdaq.
QQQ (cap-weight) and a survivorship-safe equal-weight rebuild of every point-in-time member (Norgate .NASDAQ 100 Current & Past, monthly rebalance); S&P 500 via SPY. All total return.%3MTCM), its change over the trailing 6 months; >+0.25% = raising, <−0.25% = cutting, else flat. The yield tracks the Fed funds target closely.$NDX vs $SPX price indices, 1985–2026 (six cycles), as a robustness check. Price-only data slightly favors the lower-yielding Nasdaq, so treat the +3 to +4-point figure as an upper bound; the sign is what matters.None of this makes the duration textbook wrong. A dollar of far-off earnings really is worth less when rates rise, and 2022 showed that force can bite hard. One rabbit hole deeper, the lesson is just narrower than the slogan: the effect is real but small, it shows up as absolute drag more than relative underperformance, and over a quarter-century the Nasdaq’s faster earnings growth has simply mattered more than what the Fed was doing.