The central bank is not the root cause of the boom-bust
cycle. The root cause is fractional reserve banking (the ability of banks to
create money and credit out of nothing). The central bank’s effect on the
cycle is to extend the booms, make the busts more severe and prevent the
investment errors of the boom from being fully corrected prior to the start
of the next cycle. Consequently, there are some important relationships between
interest rates and the performance of the economy that would hold with or
without a central bank, provided that the practice of fractional reserve
banking was widespread. One of these relationships is the link between a
reversal in the yield curve from flattening to steepening and the start of an
economic recession/depression.
Unfortunately, the data we have at our disposal doesn’t
go back anywhere near as far as we’d like, where “as far as we’d like” in
this case means 150 years or more. For example, the data we have for the
10year-2year spread, which is our favourite indicator of the US yield curve,
only goes back to the mid-1970s.
For a longer-term look at the performance of the US yield
curve the best we can do on short notice is use the Fed’s data for the
10year-3month spread, which goes back to the early-1960s. However, going back
to the early-1960s is good enough for government work and is still
satisfactory for the private sector.
As explained in many previous commentaries, the boom
phase of the cycle is characterised by borrowing short-term to lend/invest
long-term in order to take advantage of the artificial abundance of cheap
financing enabled by the creation of money and credit out of nothing. This
puts upward pressure on short-term interest rates relative to long-term
interest rates, meaning that it causes the yield curve to flatten.
At some point, usually after the boom has been in
progress for several years, it becomes apparent that some of the investments
that were incentivised by the money/credit inflation were ill-conceived.
Losses start being realised, the quantity of loan defaults begins to rise,
and the opportunities to profit from short-term leverage become scarcer. At
this point everything still seems fine to casual observers, central bankers,
the average economist and the vast majority of commentators on the financial
markets, but the telltale sign that the cycle has begun the transition from
boom to bust is a trend reversal in the yield curve. Short-term interest
rates begin to fall relative to long-term interest rates, that is, the yield
curve begins to steepen.
The following monthly chart of the 10year-3month spread
illustrates the process described above. On this chart, the boom periods
roughly coincide with the major downward trends (the yield-curve
‘flattenings’) and the bust periods roughly coincide with the major upward
trends (the yield-curve ‘steepenings’). The shaded areas are the periods when
the US economy was officially in recession.
The black arrows on the chart mark the major trend
reversals from flattening to steepening. With two exceptions, such a reversal
occurred shortly before the start of every recession.
The first exception occurred in the mid-1960s, when a
reversal in the yield spread from a depressed level was not followed by a
recession. It seems that something happened at that time to suddenly and
temporarily elevate the 10year yield relative to the 3month yield.
The second exception was associated with the first part
of the famous double-dip recession of 1980-1982. Thanks to the extreme
interest-rate volatility of the period, the yield spread reversed from down
to up shortly before the start of the recession in 1980, which is typical,
but during the first month of the recession it plunged to a new low before
making a sustained reversal.
Due to the downward pressure being maintained on
short-term interest rates by the Fed, the yield curve reversal from
flattening to steepening that signals an imminent end to the current boom
probably will happen with the above-charted yield spread at an unusually high
level. We can’t know at what level or exactly when it will happen, but it
hasn’t happened yet.
[This post is an excerpt from a TSI commentary
published on 6th December]
