Money does not leave the market during a slowdown. It moves. Capital rotates out of the sectors that do badly in the coming phase and into the ones that do well, and it does so before the phase arrives, because markets price expectations rather than conditions.
Sector rotation is the attempt to be early to that movement. It has a respectable theoretical basis, a well-known model behind it, and a failure mode that ruins most people who try it.
What the strategy actually is
Sector rotation means shifting weight between parts of the equity market according to where the economy sits in its cycle, instead of holding a constant allocation.
The premise is that sectors do not respond to the cycle identically. Some are highly sensitive to growth and credit conditions; others sell things people buy regardless. That difference is stable enough to build a framework on.
The main division is between cyclicals and defensives.
Cyclical sectors — consumer discretionary, industrials, materials, financials, technology — depend on the economy expanding. Their earnings swing hard in both directions.
Defensive sectors — consumer staples, healthcare, utilities — sell necessities. Demand for food, medicine and electricity does not fall much in a recession, so their earnings are steadier and they outperform on the way down.
The Stovall model
The standard reference is Sam Stovall's framework, set out in his 1996 Standard & Poor's Guide to Sector Investing, which maps sector leadership onto the phases of the business cycle.
The stylised sequence runs roughly like this.
Early recovery. The recession is ending, rates are low, credit is loosening. Financials lead — they benefit first when lending resumes and credit losses stop rising. Consumer discretionary follows, as households start spending again. Technology and industrials pick up as capital spending returns.
Mid expansion. Growth is established, capacity is being added. Industrials, technology and materials lead as investment and production run hot.
Late expansion. Capacity is tight, inflation is rising, central banks are tightening. Energy and materials do well on commodity prices. Money starts moving toward staples and healthcare as investors position defensively.
Recession. Growth contracts. Utilities, staples and healthcare lead — relatively, which usually means falling less rather than rising. Cyclicals are worst hit.
The modern sector classification has eleven groups rather than the ten Stovall was working with — real estate was separated out in 2016 and communication services was reconstituted in 2018 — but the logic transfers.
Why this is harder than it reads
The model is clean. Applying it is not, for four reasons that compound.
The market leads the economy. This is the central difficulty. Equities anticipate the cycle by roughly six to twelve months. By the time GDP data confirms a recovery, the rotation into early-cycle sectors has already happened and the leadership has moved on. Rotating on confirmed economic data means arriving after the move.
Phases are only identifiable afterwards. Nobody rings a bell at the transition from mid to late expansion. Economists disagree about the current phase in real time and revise their views when the data is revised. You are positioning against a variable you cannot observe.
Cycles are not uniform. Expansions have run from two years to more than a decade. Some recessions are financial, some are inflationary, some are exogenous shocks, and the sector response differs each time. Pandemic-era rotation looked nothing like the model.
Sector definitions drift. The largest technology companies now derive a large share of revenue from advertising and subscription services that are anything but cyclical. Some retailers are more defensive than some staples. Classification lags business reality.
Doing it in practice
The workable version starts with the cycle but does not stop there.
Read where the cycle is from data that leads rather than lags: yield curve shape, purchasing managers' indices, credit spreads, and the direction of monetary policy. The GDP release tells you where the economy was, not where it is going.
Check what the market already believes. Relative strength — the performance of each sector against the broad index — shows where money is actually moving. If defensives have been outperforming for two months, the market has already made a judgment about the cycle, and it is usually a judgment worth taking seriously.
Move gradually. Shifting weight incrementally as evidence accumulates is more survivable than switching wholesale on a call about the phase. Being partially wrong is recoverable; being completely wrong at a turning point is not.
Use instruments with low friction. Sector ETFs exist precisely for this, and they make the strategy cheap to implement. Rotating between baskets of individual stocks adds selection risk to a strategy that is already hard enough.
The overtrading problem
Sector rotation feels productive, and that is its most dangerous quality. Constant repositioning creates the sensation of being engaged and responsive while quietly destroying returns.
Costs accumulate. Every rotation is a round trip: spread, commission, and in taxable accounts a realised gain. A strategy that turns over the portfolio six times a year needs to beat the benchmark by a meaningful margin before it breaks even.
Whipsaw is the norm. Rotating into defensives on a growth scare that reverses two weeks later means selling cyclicals at the bottom and buying them back higher. Do that three times a year and the cost exceeds anything the strategy could have earned.
Complexity hides the result. With eleven sectors and frequent changes, it becomes genuinely difficult to know whether the rotation is adding value or whether the portfolio would have done better sitting in the index. Most people who run it never measure it against that benchmark, which is the only comparison that matters.
The honest version of the strategy makes a handful of adjustments a year, not a handful a month.
What it tells a currency trader
Sector rotation is an equity strategy, and the reason it belongs on a currency trader's screen is that sector leadership is a read on risk appetite that is harder to argue with than sentiment surveys.
When defensives lead — utilities, staples, healthcare outperforming — the equity market is pricing slowing growth, and the currencies that respond are the ones with growth and commodity exposure. The Australian and Canadian dollars weaken, the dollar, yen and franc firm.
When cyclicals lead, the reverse holds. Commodity currencies do well, funding currencies underperform, and carry trades work.
The signal is most useful when it disagrees with the headlines. Equity sector leadership frequently shifts before the economic narrative does, and a rotation into defensives while everyone is still talking about a soft landing is worth more than another growth forecast.
A framework, not a system
Sector rotation describes something real. Different parts of the economy do respond differently to the cycle, that pattern has held for decades, and knowing it makes market behaviour legible.
What it does not provide is timing. The market moves before the data confirms the phase, the phase itself is disputable in real time, and every cycle has features the previous one did not. Treating the model as a schedule produces exactly the overtrading it is supposed to avoid.
Used as context — which sectors should be doing well if the current narrative is right, and what does it mean that they are not — it is genuinely informative. Used as a set of instructions for when to switch, it is an expensive way to underperform the index it is measured against.
