1. Where inflation comes from: demand, costs and expectations

Macro 4: Inflation, currencies and commodities

Level 1 taught you how to measure inflation and how to strip it out of your returns. This module goes one step further back: where it comes from in the first place. That is not academic trivia. The origin of an inflationary episode determines how the central bank reacts, and it is that reaction which decides what happens to the discount rate, to corporate margins and therefore to your portfolio.

Talking about "inflation" in the singular gets you nowhere

Just as there is no such thing as "the interest rate" but rather a ladder of maturities, there is no single inflationary mechanism. There are three, and they overlap:

• Demand-pull. Total spending in the economy exceeds what that economy can produce with the capacity installed and the labor available. There is too much money chasing goods that do not exist.
• Cost-push. The price of an essential input rises for reasons unrelated to demand: a war, a drought, a gridlocked port, a production cut agreed by a cartel.
• Expectations-driven. Prices and wages are set with an eye on the future. If everyone takes 5% for granted, everyone asks for and concedes 5%, and 5% duly happens.

Demand: when spending outruns what can be produced

Every economy has a short-run ceiling on output, set by its factories, its technology and its workers. The distance between what it actually produces and that ceiling is the output gap. Once the gap closes and spending keeps growing, firms can no longer respond with more units: they respond with higher prices, because the order book is full and they know the customer will pay. In parallel, workers bargain harder because there are more vacancies than candidates.

That relationship between labor market tightness and inflation is the Phillips curve. In the years before the pandemic it looked almost flat: unemployment fell and prices did not budge, and a good part of the profession came to write it off as dead. The lesson that followed was that it is not a straight line, it is a curve: in the slack zone it has barely any slope, and when the economy genuinely overheats it steepens all at once.

The advantage of demand-pull inflation is that the central bank has precisely the right tool: raising rates cools spending, which is the cause. The bill comes as a recession, but the instrument works on the problem rather than on its shadow.

Costs: the shock that arrives from the supply side

A jump in the price of oil or gas does two things at once: it pushes prices up and it reduces output. That is the definition of stagflation, and it is the worst possible scenario for a central bank, because its two objectives point in opposite directions. Raising rates does not produce a single extra barrel.

A fine distinction is worth making here: a supply shock is, in principle, a change in relative prices, not inflation. Energy becomes expensive relative to everything else and the consumer, with whatever income is left, buys less of other things. That blow drops out of the index after twelve months through the pure arithmetic of the year-on-year comparison. Economists call it the first-round effect and the standard doctrine is to look through it.

The danger lies in second-round effects: expensive energy feeds through to transport, to fertiliser and to food, workers demand compensation for the purchasing power they have lost, firms raise prices to defend margins, and a change in relative prices turns into a generalized spiral. At that point the central bank stops looking the other way.

Expectations: the part that fulfills itself

Expectations are the most valuable asset a central bank owns, and the most expensive to rebuild once lost. If firms and unions believe inflation will return to target, they write contracts as though it will and help bring it about. If they stop believing it, indexation spreads and inflation becomes self-perpetuating.

They are monitored in two ways: surveys of households and firms, and market measures such as the breakeven you already know, above all the five-year rate five years forward, which isolates the long run from immediate noise. A jump in short-term expectations, almost mechanical when gasoline prices rise, is one thing; a de-anchoring of long-term expectations is quite another, and it is the real alarm signal.

Volcker showed in 1979-1982 what correcting a de-anchoring costs: he took the policy rate above 19% and US unemployment climbed past 10%. He was not fighting that year's prices, he was buying credibility.

2021-2022: the debate settled in real time

It was all three at once, which is why it proved so hard to diagnose. There was demand: the US fiscal package of March 2021 alone put 1.9 trillion dollars to work on an economy that was already recovering, with households sitting on forced savings and consumption rotating from services into goods. There were costs: containers, semiconductors and, after the invasion of Ukraine, European gas that passed 300 euros per megawatt hour in August 2022, more than ten times its normal level.

US CPI touched 9.1% year on year in June 2022 and the euro area reached 10.6% in October. The proof that it was not only energy came from US core inflation, which hit 6.6% in September 2022, months after oil had peaked. Powell retired the word "transitory" in late 2021, once it was plain that the demand component was not going to fade on its own.

And the detail that explains the ending: long-term expectations never de-anchored. The Michigan survey's five-to-ten-year measure hovered around 3% throughout the episode. That anchoring is what allowed inflation to come down without the deep recession almost everyone had taken for granted.