Sector Rotation

Today’s read posts after the close.

Money does not usually leave the market. It moves. Sector rotation is the study of where it goes and, more usefully, of what the destination implies about what participants collectively expect.

The reason this is worth learning is that rotation is often visible while the index is doing nothing at all. A flat month at the index level can contain an enormous amount of information about changing expectations, and the index will not tell you any of it.

The idea in one paragraph

Different kinds of businesses respond differently to the same economic conditions. Companies selling things people buy regardless of circumstance behave differently from companies selling things people defer. Businesses that borrow heavily respond differently to credit conditions than businesses that do not. When the collective expectation about conditions changes, capital reallocates toward whichever group the new expectation favours, and that reallocation shows up as one group outperforming another well before it shows up in any economic statistic.

The market is not forecasting perfectly when it does this. It is voting, and it is often wrong. But the vote is observable, and a vote is information even when it turns out to be a bad one.

The map, held loosely

The textbook version assigns sectors to phases of an economic cycle. Early expansion favours the groups most sensitive to credit and to discretionary demand. Late expansion favours the groups that do well when input costs and rates are rising. Contraction favours the groups people cannot stop buying from. Recovery favours the sensitive ones again.

I use this map. I also want to be honest that it is a stylised description rather than a law, that the phases are only obvious afterwards, and that plenty of periods refuse to fit it at all. The mechanism underneath the map is covered properly in how the business cycle drives sector rotation.

The map's real value is not prediction. It is that it gives you a hypothesis to test against what the tape is actually doing, which is much better than having no hypothesis and simply narrating whatever went up.

Relative strength is the instrument

Absolute performance tells you very little during rotation, because in a rising market almost everything rises and in a falling one almost everything falls. What matters is one group against another, or a group against the index.

The practical form is a ratio: divide one thing by another and watch the line. When the ratio rises, the numerator is winning. That is the entire technique, and its simplicity is a feature, because there is nothing in it to tune and therefore nothing to overfit.

What makes ratios useful is that they strip out the market-wide move and leave the part that is actually about the comparison. A sector that fell three percent in a market that fell six percent is a strong sector, and no absolute chart will ever tell you that.

What rotation tells you that the index cannot

It shows changing expectations early. Positioning shifts before economic data confirms anything, because participants act on what they think is coming.

It shows whether a rally is broad or narrow, which overlaps with market breadth and reinforces it from a different angle. A rally led by one group is a different animal from one where leadership passes around.

It shows stress before price does. When defensive groups start outperforming inside a market that is still making highs, something in the collective expectation has changed even though the index has not registered it. This is a genuine and repeatable observation.

It also, crucially, shows nothing about timing. Every sentence above describes a condition. None of them tells you when.

Where I have been wrong with this

Two ways, repeatedly.

Reading rotation as prophecy. Defensive leadership does not mean a decline is coming, it means participants are hedging. Sometimes that hedge is prescient and sometimes it is expensive insurance against a thing that never arrives. Treating the hedge as a forecast has cost me more than treating it as information.

Confusing a sector with the companies in it. Sector labels are administrative categories, and they are increasingly bad ones. Businesses get classified by history rather than by what they actually do now, and a "sector" can contain companies with almost nothing economically in common. A sector-level read that falls apart on inspection of its constituents was never a read, it was a label. Leading and lagging stocks covers how differently names inside one group can behave.

The third failure, less mine than the field's, is the seductive completeness of the cycle map. Because it explains everything after the fact, it feels predictive. Test it forward, in writing, with dates, and the feeling weakens considerably. That test is the point of keeping a scored record.

Practical use

I look at rotation to understand what the market currently believes, not to decide what to do today. It is context, and it is amplitude: it tells me what kind of environment a position lives in and where the pressure is likely to come from if I am wrong.

The routine version is simple and it takes very little time. Compare sector performance against the index over a few horizons, watch the ratios that are changing rather than the ones that are stable, and ask what a person would have to believe about the economy for that rotation to make sense. Then check whether you believe it, and note that the market's answer and your answer can differ for a long time.

The reading method in more detail is on sector rotation reading.

Educational market analysis only. Nothing here forecasts sector performance or recommends any position.

Questions traders ask

What is this Sector Rotation page?

A dated, running read on sector rotation from a trading desk that scores its own calls publicly — hits and misses both.

How often is it updated?

After the market close on trading days.

Is this investment advice?

No — educational market analysis only. Nothing here is a recommendation.