Lesson 8 of 8
Stated so it can be wrong
The precise content of the claim, what would count against it, and the four things it is often taken to say and does not. The most important lesson here.
By Javier Sanz
Founder of ValueScreener. Built Ninety Nine, a securities services provider, and sold it to a tier-one global bank; ran operations and internationalization at Alpaca, the brokerage API.
Who writes thisEditorial processLinkedIn (opens in a new tab)X (opens in a new tab)
In one minute
- The claim has content: a credit expansion not backed by saving should be followed by a particular pattern of relative price and investment movements, in a particular order, reversing at the end.
- What would count against it: booms with no prior credit expansion, credit expansions with no reallocation towards remote stages, and busts with no prior relative-price pattern.
- What it does not claim: a date, a magnitude, that every recession is a credit cycle, or that it can predict anything at all.
A theory stated so that nothing could count against it is not a theory. This lesson writes the claim down in a form that could be wrong, and then removes the four things people commonly attach to it that it does not support.
The claim and its negation, side by side
What the theory says should follow a credit expansion
- Resources shift towards stages further from consumption.
- Producers' goods prices lead, consumers' goods lag.
- Capital goods industries swing more than consumer goods industries.
- The order reverses before the contraction begins.
- Specific capital in early stages loses value permanently.
What would count against it
- Booms that follow no credit expansion.
- Credit expansions with no shift towards remote stages.
- Contractions with no prior relative-price pattern of that shape.
- Capital goods industries swinging less than consumer goods industries.
- Early-stage capital holding its value through the bust.
Both columns have to be statable before the theory counts as one.
The claim, written out
A sustained expansion of bank credit not matched by voluntary saving should be followed by a reallocation of resources towards the stages of production furthest from consumption; by a rise in the prices of producers' goods relative to consumers' goods; by a subsequent reversal of that relative movement as incomes catch up; and by a contraction in which the specific capital committed to the early stages loses value and is not fully recovered.
Every clause in that sentence is a place the theory can be caught.
The four things it does not claim
It does not claim a date. Nothing in Mises or Hayek says how long a boom lasts, and Mises says explicitly that the only choice is whether the crisis comes sooner or later, which is the opposite of a schedule.
It does not claim a magnitude. No step in the mechanism produces a number.
It does not claim that every recession is a credit cycle. The mechanism describes one way a contraction can be generated and says nothing about contractions caused by a war, a harvest, an epidemic or a supply shock.
And it does not predict. The market-process course in this track already established why, through Lachmann's argument that expectations cannot be treated as given data: if the same prices can be read differently by reasonable people, the theory cannot say who acts, how far, or when.
The honest problems with testing it
Three, and they are serious.
The stages of production are not observable, so any test substitutes an industry classification and the substitution is a judgement.
The natural rate is not observable either, which the first course of this pair argued at length, so "credit expansion not matched by saving" has to be proxied.
And the pattern the theory predicts is consistent with other accounts too. A matching record supports the theory without singling it out, which is why the course setting the rival accounts side by side matters more than any single confirmation.
What to do with this lesson
Carry it into the next three courses. When the bust course says the damage was done during the boom, ask what evidence of the boom would have to look like. When the Depression course sets four accounts side by side, use the right-hand column above as the test you apply to this one as strictly as to the others.
Check yourself
Which of these would count against the theory?
- A credit expansion followed by a contraction
- A large credit expansion with no shift of resources towards stages further from consumption
- A boom in which consumers' goods prices rise
Show the answer
Answer: B.
The reallocation towards remote stages is the load-bearing claim. Without it the mechanism has not run, whatever else happened.What does the theory deliberately not claim?
- That credit expansion affects relative prices
- A date, a magnitude, that every recession is a credit cycle, or any predictive power
- That capital goods industries swing more
Show the answer
Answer: B.
Mises says the only choice is whether the crisis comes sooner or later, which is the opposite of a schedule, and Lachmann's argument about expectations rules out prediction from inside the tradition.Why is a matching historical record not decisive?
- Because records are unreliable before 1945
- Because the pattern is consistent with other accounts too, so a match supports this theory without singling it out
- Because the theory is about the future only
Show the answer
Answer: B.
That is why the course setting the rival accounts beside each other matters more here than any single confirming episode.Sources
- Mises, Human Action, 1949, chapter XX, section 8, for the statement that there is no means of avoiding the final collapse of a credit boom and that the only alternative is whether the crisis comes sooner or later.
- Hayek, Prices and Production, first published 1931, second revised edition 1935, lecture III, sections 7 and 8, for the sequence of relative price movements and the fate of specific capital, which are the clauses this lesson turns into tests.
- Hayek, 'Price Expectations, Monetary Disturbances and Malinvestments', a 1933 lecture reprinted as essay IV in Profits, Interest and Investment, 1939, section III, for the requirement that a cycle theory explain simultaneous error in the same direction.
- Rothbard, America's Great Depression, 1963, part one, chapter 1, for the three facts he says a cycle theory must account for, including the wider swing in capital goods industries.
- Lachmann, Capital and Its Structure, 1956, chapter 2, for the argument about expectations that rules out prediction from inside this tradition.
What this leaves out
- This is a heterodox tradition. The mainstream account of the same phenomenon differs, and nothing here is a forecast or a view on any security.
- The episode in the example is invented to show what a contrary finding looks like. It describes no country and no period.
- This lesson states the claim in testable form. It does not report what the published tests found, which belongs to the course on the case against the theory and to the course setting rival accounts side by side.
- Neither the stages of production nor the natural rate is observable, so every test of this claim rests on proxies chosen by the person testing it.