Don't Fear Rate Hikes, Fear the Cuts: 56 Years of History

Every Fed meeting is sold as the event that decides the market's fate. Rates up — «brace for a crash». Rates down — «here comes the rally». The logic sounds so obvious that almost nobody checks it.

We checked. We took every Fed rate cycle since 1970 — 15 hiking cycles and 17 cutting cycles — and measured what the market actually did. The result inverts the intuition: the market rose during hiking cycles more often than it fell, and the worst crashes in modern history began precisely when the Fed started cutting.

Below: the numbers, why this happens, and which of our indicators show this picture in real time — so you can see it instead of guessing.

Methodology

Data comes from FRED (St. Louis Fed): the federal funds rate (DFF), Nasdaq Composite (NASDAQCOM), unemployment (UNRATE), and official recession dates (USREC). A «cycle» is a sustained move of at least 0.75 percentage points in one direction. BTC returns are measured separately on Kraken weekly candles — Bitcoin has a shorter history, only two complete rate cycles.

One caveat up front: 15 and 17 cycles is a small sample. None of the numbers below is a «rule». They are a baseline that stops you from mistaking a popular opinion for a fact.

Fact 1. Markets usually rise while the Fed hikes

The most surprising result. Of 15 hiking cycles since 1970, the market finished higher in 10. Average return +13.8%, median +7.2%.

Look at 2016–2019: the Fed lifted rates from 0.41% to 2.42% and the Nasdaq gained 51.9% along the way. Or 1999–2000: rates climbing, market up 57.4%. Hiking rates does not, by itself, sink the market.

That gain comes at a price though: the average maximum drawdown inside a hiking cycle is −18.9%, and a drawdown occurred in all 15 cycles without exception. So the correct phrasing isn't «hikes are safe» — it is that hiking delivers volatility without a trend change. Painful, not fatal.

Fact 2. The decision is priced in long before the meeting

This answers the «is it already in the price?» question directly. Take the US 2-year Treasury yield and test whether its gap to the current policy rate predicts what the Fed does nine months later.

Across 7,992 daily observations since 1994, the correlation is r = 0.66 with a slope of 1.16. In plain terms: if the 2-year trades 1 pp above the current policy rate, the Fed on average really does raise rates by about 1.16 pp over the following nine months.

The bond market doesn't wait for Powell — it votes early, and the Fed mostly ratifies a price that already exists. The practical consequence: trading the «fact» of a hike is late. The meeting itself usually produces a volatility spike around a known outcome. The real information lives in the gap between what the market expected and what the Fed said.

Fact 3. A rate cut is not a gift — it is a diagnosis

Here is the core of it. Average returns after a cutting cycle begins look respectable, and that average is exactly what misleads people. Everything changes once you split cuts into two groups: those followed by a recession within twelve months, and those that were not.

The gap is enormous. When cuts came without a recession, the market was higher a year later in 8 cases out of 8 — a clean sweep, averaging +15.7%. When cuts coincided with a recession: only 3 of 9, with an average maximum drawdown of −32.8%.

The most infamous crashes are all rate-cutting episodes:

• July 2000: the Fed began cutting from 6.54%. The Nasdaq lost 46.2% over the next 12 months and 64.8% over 24, with a 66.7% drawdown.
• May 2007: cutting from 5.25%. Two years later the market sat 32.1% lower, with a 47.3% drawdown along the way.
• September 1973: cuts into the oil shock. Down 42.3% in a year.

The mechanism is simple and unpleasant. The Fed does not cut out of generosity. It cuts when the economy is already slowing — which means corporate earnings are already at risk. Rate cuts do help, but they work with a 6–12 month lag, while falling earnings hit immediately. The damage arrives first; the help arrives later.

That is why «the Fed is cutting, time to buy» is a dangerous sentence. The right reaction is «the Fed is cutting — what broke?» If the answer is «nothing, inflation simply returned to target», that is the best setup a market can get. If the answer is «employment is collapsing», no rate cut will save you.

Fact 4. What Bitcoin did

BTC has a shorter history — two complete rate cycles — so this is observation, not statistics. But it fits the same logic.

Note the contradiction: the 2016 hiking cycle coincided with BTC up 933%, while the 2021 hiking cycle coincided with BTC down 76%. Same macro signal, opposite results. That is the cleanest proof that the rate alone predicts nothing — what matters is the liquidity phase it rises into.

The 2023–2024 turn, by contrast, went by the textbook: BTC bottomed at the rate peak (August 2023), not at the first cut. The market turned on the expectation of easing, a full year before the fact.

How to track this: our indicators

Everything above reduces to three questions, and each has an answer that updates without your involvement.

1. Is the rate already priced in? → Macro Dashboard

You cannot trade the Fed funds rate directly, but you can trade its reflection. Macro Dashboard tracks the three channels through which a Fed decision reaches the market — well before the meeting:

• US 10-year yield — a move beyond 0.15 pp over 30 days flips the signal. Rising yields mean bonds start competing with risk assets.
• DXY, the dollar index — threshold ±1.5% over 30 days. A strengthening dollar pulls capital out of BTC and equities.
• Global M2 — actual liquidity: above +5% YoY is expansion, below zero is contraction.

The dashboard combines these with sentiment, BTC momentum and funding into a single verdict and — more useful — surfaces blockers and triggers: not just «bearish», but «yields rose 0.3 pp over 30 days» plus the specific condition that would flip it. That is how you see a rate priced in without waiting for the meeting.

2. Is liquidity expanding or contracting? → Business Cycle

This directly answers the puzzle from Fact 4 — why two identical hiking cycles gave +933% and −76%. Business Cycle overlays M2 money supply (year-over-year) on the BTC chart and labels the phase: expansion (M2 above +5%), neutral, or contraction (M2 below zero).

A hike during expansion and a hike during contraction are two different events with different consequences, even though the headline reads the same. The rate tells you the Fed's intent; M2 tells you what is actually happening to money. Watch the second one.

3. Did something actually break? → Market Cycles and Brain Forecast

The central question — «is this cut a normalization or an emergency?» — cannot be answered by one indicator, because recessions are declared in hindsight. What works instead is agreement across sources: Brain Forecast reads 15 analytics blocks (structure, volume, whales, liquidations, macro, sentiment, cycles) into one weighted verdict, and says «no clear edge» honestly when they disagree.

That disagreement is itself the alarm: the Fed easing while market structure, liquidity and whale positioning all point down. That is exactly what 2000 and 2007 looked like.

What to do with this

• Don't trade the meeting itself. The decision is already priced (r = 0.66 nine months ahead). What trades is the surprise, not the fact.
• During hikes, prepare for a drawdown, not a reversal. Average in-cycle drawdown is −18.9%, yet the market closed the cycle higher in 10 of 15 cases. That is a position-sizing problem, not an exit signal.
• The first cut is a question, not a buy signal. Check whether a recession is coming. Without one: 8 of 8 positive. With one: 3 of 9, and drawdowns to −66%.
• The bottom usually forms at the rate peak, not at the first cut. BTC turned in August 2023, a year before actual easing.
• Check M2, not headlines. Liquidity explains the difference between +933% and −76% where the rate explains nothing.

Bottom line

«Rates up, market down» is not supported by the data: in 10 of 15 cycles the market closed the hiking period higher. The inverse and far less popular idea does hold up: a rate cut is a signal that something in the system broke, and fixing it takes longer and hurts more than people expect.

Even that isn't automatic. What separates the outcomes is not the rate but the recession: without one, cuts are fuel (8 of 8); with one, they are the opening act of a crash (3 of 9). So watch liquidity, yields, the dollar, and whether the rest of the market agrees — which is precisely what Macro Dashboard, Business Cycle and Brain Forecast are built to collect.

Educational material, not investment advice. Historical patterns do not guarantee future results, and a sample of 15–17 cycles is small by statistical standards.