How economies work · 8 of 9
Credit cycles
Explain why debt makes cycles larger, and name what turns a boom.
American households were spending 13.2% of their after-tax income on debt payments at the end of 2007. Thirteen years later that figure was 8.3%, the lowest since the series began in 1980. Nothing between those two dates made Americans thriftier. They spent the intervening years paying down, defaulting on, and writing off the debt they had accumulated through the boom, and the years it took to do that are what a deleveraging is.
One identity
Spending equals income plus net new borrowing. That is what borrowing means: you spent more than you earned and somebody lent you the difference.
Everything uncomfortable follows from taking that seriously across several years. Your spending is somebody else's income. If credit is expanding, total spending exceeds total income, which means measured income is rising faster than production. Everyone feels richer, and the feeling is real in the sense that the money is real. It is also borrowed, and the borrowing has a schedule attached.
An economy with four rules
Below is a whole economy in about a hundred lines. Lenders lend more when borrowers have room to pay and when recent years have been good. Every loan adds a fixed payment to future years. One person's spending is the next person's income. That is all of it.
There are no shocks. No bank fails, no war starts, no committee makes an error. Pull the dials and try to find a setting where it stops cycling.
The dial most people reach for first is the one that stops lenders extrapolating, because over-optimism sounds like the culprit. It shrinks the swing by about a third and leaves the cycle exactly where it was. Only turning lending off entirely removes it, and an economy with no credit is a different economy rather than a fixed one.
The household debt service ratio is FRED series TDSP, published quarterly by the Federal Reserve and free to view. The model figures come from the simulation on this page at its default settings, which you can reproduce by leaving the dials alone.
Why it turns
The mechanism is one asymmetry. Debt service is contractual and income is not.
A borrower who signs for $900 a month owes $900 a month whether or not the raise arrives. When income falls, the payment stays where it is, so the share of income going to debt rises at exactly the moment the borrower can least afford it. That rising share is what lenders look at, so they lend less, which lowers income further, which raises the share again.
An earlier version of this model made repayment a share of income instead. It could not produce a bust at any setting anyone tried, because a falling income automatically shrank the obligation and nothing was ever hard to pay. Making the payment fixed is what made the model behave like an economy.
ConventionHyman Minsky sorted borrowers into three kinds, and the sorting is more useful than most macro vocabulary. A hedgeborrower's income covers interest and principal. A speculativeborrower's income covers the interest, so the principal has to be refinanced. A Ponziborrower's income covers neither, and the position only works if the asset keeps rising. Late in a boom the mix shifts toward the third kind, which is why the same interest rate that was survivable in year one ends the party in year six.
What to do with this
Find the debt service ratio rather than the debt. For a country it is published. For a company it is interest cover, which is operating profit divided by interest expense. For a household it is the payment against take-home pay. That ratio tells you how much room is left, and room is what decides whether the next year can be bigger than this one.
Then ask which way the change in borrowing is moving. Rising borrowing that is rising more slowly is already a drag. And when you hear that a boom is different this time because of a new technology or a new policy, check whether debt against income is going up at the same time. If it is, the argument is about what the borrowing was spent on, which is a real question, and not about whether the payments will come due, which is settled.
Test yourself
01Borrowing in an economy is still positive, but lower than last year. What happens to spending?
It falls. Spending depends on the level of new borrowing, so growth in spending depends on the change in that level. Borrowing going from 100 to 80 is a twenty unit hole in somebody's revenue even though credit is still expanding.
This is why a recession can begin while every credit number in the newspaper is still positive. Nobody has to pay anything back for the trouble to start. They only have to borrow less than they did.
02Two countries end a boom with the same debt-to-income ratio. One lets borrowers default and banks take losses. The other extends and pretends. Which recovers faster, and what does the slow one buy with the delay?
The one that writes the debt off. A default removes the obligation, and the obligation is the thing draining spending. The country that keeps the loans alive keeps the payments alive too, so income stays committed to servicing debt for years and cannot go into anything else.
What the slow route buys is the appearance of solvency in the banking system, and sometimes that is worth having, because a bank that admits its losses may fail and take the payment system with it. The choice is between a fast painful adjustment and a slow one, and both are real choices that real countries have made. Switch defaults off in the model and watch how much longer it takes.
03Why does a boom feel like rising productivity while it is happening?
Because the arithmetic is identical from inside. Spending is up, revenue is up, hiring is up, and asset prices are up. Nothing in the experience distinguishes income that was earned from income that was borrowed into existence, and the borrowing is a stock nobody publishes weekly. The tell is not in the growth figures. It is in the debt service ratio, which rises quietly through the good years and is the reason they end.
A tutor that knows this lesson. It asks before it explains, and it will not tell you what to do with your own money.
Educational material, not investment or policy advice. Figures are cited where they come from a filing or a statistical series, and labelled as illustrative where they do not.