3. The four cycle phases and which asset leads
Macro 3: The yield curve and the business cycle
The economy does not grow in a straight line: it accelerates, overheats, slows and contracts, and then starts over. On top of that observation the industry has built one of its best-selling ideas, phase rotation, which promises to tell you what to buy at each point of the cycle. It is worth knowing because it organizes your thinking. It is worth knowing, above all, for its limits, which are enormous and almost never make it onto the slide.
The four phases and what defines them
- Recovery. Activity has bottomed out and is beginning to improve. Unemployment is still high, inflation low, the central bank is still keeping money cheap and corporate earnings are climbing off a deeply depressed floor.
- Expansion. Growth is solid, employment normalizes, earnings grow at a good clip and monetary policy starts withdrawing stimulus without any urgency. It is the longest phase of all.
- Slowdown. The economy is running up against its capacity, inflation bites, the central bank tightens and earnings growth stalls even while it remains positive. It is the most confusing phase: everything is fine and everything is getting worse at the same time.
- Contraction. Activity falls, unemployment rises, earnings collapse and the central bank cuts. It tends to be the shortest phase and the most violent.
The clock: growth and inflation instead of a calendar
A more workable way of organizing this, popularized by the so-called investment clock, replaces the phases with two questions: is growth above or below its trend? And inflation? That gives four boxes, each one historically associated with a dominant asset:
- Reflation, weak growth and inflation falling: government bonds lead, because rate cuts are on the way.
- Recovery, growth picking up and inflation still low: equities lead, the most favorable combination there is.
- Overheating, strong growth and inflation rising: commodities lead, and bonds suffer.
- Stagflation, weak growth and high inflation: almost nothing leads. Equities and bonds suffer together and cash, which pays little, ends up being the least bad option.
Within equities, the most quoted sector map orders things like this: in recovery, consumer discretionary, financials and small companies tend to pull ahead; in expansion, industrials and technology; in slowdown, energy and materials alongside quality companies; in contraction, the usual defensives, consumer staples, healthcare and utilities, plus long-dated government bonds. That ordering has economic logic behind it, and it is not pure statistics: whoever sells durable goods depends on cheap credit, whoever sells raw materials is paid by late-cycle inflation, and whoever sells food or electricity keeps on selling whatever happens.
Why this is a rough map and not a law
Here comes the important part, and it is what separates someone who understands the cycle from someone who repeats the chart.
There is no agreement even on how many phases exist. Some frameworks use three, others four, others five, and every research house feeds its own with different indicators: some look at employment, others at activity surveys, others at bank lending. Two competent analysts looking at the same Monday can place the economy in different phases and defend both positions with data on the table. A map whose current box is arguable cannot produce clean signals.
Phases are only identified with certainty after the fact. There is no bell. The body that officially dates recessions in the United States works with an enormous delay: the 2020 recession ended in April of that year and was not confirmed until July 2021, fifteen months after the event. Anyone waiting for official confirmation always acts late, and not by a little.
The market prices the next phase, not the current one. Equities tend to turn in the middle of the recession, with the data still worsening. In March 2009 the US market bottomed; the recession did not end until June and unemployment kept rising until it brushed 10% in October. Buying in the right phase means buying when the news is at its worst, not when it improves.
Risk premia are countercyclical. This is the central idea of the literature on expected returns, and Antti Ilmanen has developed it with more data than anyone: the expected compensation for taking risk is at its highest precisely when nobody wants to take it. That explains why the map is right on average and hard to execute in practice.
The sample is tiny once again. Since 1945 there have been little more than a dozen US recessions, with radically different causes: oil, banking crises, the technology bubble, mortgages, a pandemic. An average calculated over twelve heterogeneous episodes hides an enormous dispersion, and getting the phase wrong penalizes you twice: you are out of what is working and into what is not, paying trading costs at every turn as well.
A real case: 2020, the whole cycle in a single year
February 2020: US unemployment stood at 3.5%, a half-century low. By April it had jumped to 14.7%. The official recession lasted two months, the shortest ever dated, and the stock market fell around 34% from its high on February 19 to its low on March 23.
And then, with unemployment still in double digits and no good news available anywhere, the market began to rise. It regained its previous high in August 2020, with the economy still in pieces. By the time official confirmation arrived in July 2021 that the recession had ended in April of the previous year, the index was trading at close to twice its low.
That is the honest summary of this module: the phase map describes what happened rather well, and it is practically unusable as a real-time signaling system. It helps you understand why your portfolio behaves the way it does, not what to buy on Monday.