Predicting Recessions The Easy Way: Monetarists, MMT, And The Money Stock

by: Steve Roth


Recessions are reliably predicted over the last half century by declines in real household net worth.

Stock investors leaving the market as of the decline reported in the March 6, 2008 Fed Z.1 report avoided the 50% S&P decline over the next twelve months.

This pattern reveals a modified "monetarist" view of the economy, with real household net worth representing the "money stock." When households have less money, they spend less.

If you see a year-over-year decline in real household net worth, you are just into, or about to be into, an NBER-designated recession.

That's been true in the U.S. since the late 60s (click for source):

Every time this measure went negative, a recession ensued -- with one exception: Q3 2011. It's eight for seven over the last half century. Change the criterion, "YoY decline greater than 1%," and it's seven for seven. Yes, that's a small sample size, but...

If you'd followed this indicator as an S&P 500 investor, you would have gotten out of the market on release of the March 6, 2008 Fed Z.1 release (which showed a 3.2% YoY decline in real household net worth as of Q4 2007), avoiding a 50% market decline over the next twelve months.

This has me thinking like a monetarist, but with a very different definition of the "money stock" (or the utterly incoherent term, "money 'supply'"): Real Household Net Worth (inherently including the net worth of all firms, which are ultimately owned by households).

Key emphasis here: "Net." Hold that thought.

When the quantity of money declines (absent a matching runup in velocity), spending declines. So production and employment decline. It's a fairly simple and straightforward (behavioral) economic story: When people have less money, they spend less. Big surprise. (Pace the monetarists, the proportion of that money that's stuffed in mattresses is rather immaterial to how much they choose to spend.)

If this definition of the money stock holds water, we need to think about what affects the money stock. Short answer, in strict accounting terms:

• Government deficit spending. The government spends dollars into existence, ab nihilo. Sovereign currency issuer and all that. Pure Modern Monetary Theory.

• Runups and declines in existing-asset markets. (This will raise eyebrows, including/especially among MMTers...) When the financial and real-estate markets goes up, household net worth increases, with no change in household liabilities. Again, this is ab nihilo "money creation," caused purely by animal spirits, people's changed beliefs about the value of (and future output from) existing assets. Ditto when markets decline. Money is destroyed.

• Inflation. (This rather turns monetarism on its head. Bear with me.) When price inflation is high, real household net worth grows more slowly/declines more rapidly. And vice versa.

What about private lending/borrowing? Doesn't that create money? If the money stock equals household net worth, no: not in direct accounting terms. It creates new household assets (bank deposits) and new household liabilities (loans due) in equal amounts. Change in household net worth: zero. (It obviously increases household assets, but that's ignoring the increased liabilities.)

Private lending certainly can (does) cause increases in household net worth (if the borrowing results in there being more output than there would have been otherwise). That's an economic effect. Think: proximate versus ultimate causes. But in strict accounting terms, the act of private lending/borrowing doesn't increase household net worth.

Big stocks of household debt (relative to household net worth and/or income) -- high household leverage -- can for obvious reasons make household net worth more volatile in both directions. This explain's Steve Keen's important findings that high private debt levels -- and especially rapid changes (and changes in changes) in those debt levels -- are at least predictors, and arguably causes, of booms and busts. (If we posit them as a cause, that is again an economic effect and explanation we're propounding, though it's founded on solid accounting.)

Treasury borrowing/issuing bonds (swapping them for bank deposits) and the Fed buying bonds (swapping them for reserves) likewise has no direct accounting effect on household net worth; they just change the aggregate asset mix. But of course like bank lending, these actions can/do have economic effects that can cause changes in net worth.

This leads me back to Roger Farmer's important question: did the declining stock market (Granger-)cause The Big Whatever, '07-'09?

I'd say no. The decline in real household net worth was the proximate cause. That decline was rooted (at least initially) mostly in the real-estate meltdown. The stock-market decline joined the declining-animal-spirits party, and definitely contributed. Then add the wildly leveraged financial sector (ultimately owned by households!), with all its shenanigans and mis-priced securities, amplifying the whole thing right onto household balance sheets.

Looking at the graph above, we're in no danger (that we can see) of an immediate recession. If recent days' stock-market declines continue, though, concern may be in order. The big problem, of course, is that we don't get the household net worth measure for three and a half months after the end of a quarter. For now, I'm anxiously awaiting the September 18 release of the Fed's Q2 Z.1 report, in particular its estimate of household net worth as of June 30th.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.