Over and over again in 2022, economists, analysts, policymakers and market observers engaged in the “long and variable lags” debate.
The Fed’s aggressive rate hikes will manifest in slower growth and, hopefully, slower inflation, but on an unpredictable delay. It’s impossible to say, with anything approaching certainty, how long that delay is. It’s also quite difficult to preemptively quantify the growth drag and/or the disinflationary impulse.
This is especially problematic in the current conjuncture for two obvious reasons:
- Inflation is unusually high. “This is not a drill,” so to speak. The longer inflation remains extraordinarily elevated, the higher the risk consumer expectations become unanchored and the more likely inflation is to become embedded in wage-setting. Inflation needs to come down sooner rather than later, so flying blind vis-à-vis the lag between tightening and inflation relief risks inadvertently persisting in an insufficiently restrictive policy stance at the risk of unmoored consumer expectations and a wage-price spiral, an economic black hole from which there’s no escape.
- The Fed hiked rates at an extremely aggressive pace this year, and because it’s impossible to know, ahead of time, what the threshold is for “sufficiently restrictive” vis-à-vis inflation, it’s possible the Fed already overshot, pushing rates higher than needed to control price growth to the extreme peril of the economy. By definition, “sufficiently restrictive” means policy that constrains growth. But accidentally overshooting the “sufficiently restrictive” threshold means adopting a deflationary policy stance rather than a disinflationary stance, which in turn puts the economy at risk of an unnecessarily deep downturn.
If you ask Goldman, the “long and variable lags” debate often confuses GDP growth with GDP levels, a crucial distinction, with potentially far-reaching ramifications.
The bank’s Joseph Briggs suggested that one reason many forecasters are overtly bearish on the economy is the notion that “long and variable” lags means the rate hikes already delivered plus planned additional hikes during the first quarter, will push the economy over the edge and into recession.
“The obvious implication… is that most forecasters anticipate that the policy tightening delivered in 2022 will create a significant drag on growth in 2023,” Briggs wrote, before suggesting that in fact, the lag isn’t very long: A mere two quarters. The premise is simple and, for many readers, familiar. Goldman wrote that,
We believe that a tightening in financial conditions starts to affect the economy when financial markets react to expected policy changes instead of when rate hikes are delivered. Most of the impact of Fed rate hikes occurs indirectly via changes in broader financial conditions, which is the main reason why we view our FCI as more informative about the effect of policy tightening on the real economy than the current Fed funds rate. Many economic commentators and forecasters confuse lags from monetary policy to GDP growth and GDP levels. It’s often missed that Milton Friedman’s seminal assessment that monetary policy acts with long and variable lags — which is often cited as the foundation for the view that monetary policy effects on growth are “long and variable” — referred to the time until the peak impact on the level of GDP, not the peak impact on growth.
To the first point, Goldman has, of course, done the math. A 100bps “surprise” in Fed hiking expectations generally translates into a 100bps tightening in financial conditions as measured by the bank’s oft-cited Financial Conditions Index. The peak drag from an FCI tightening impulse of that magnitude on GDP growth occurs after just two quarters, even as some drag persists for six quarters. So, the peak impact on the level of GDP occurs after those same six quarters.
What does that mean in the current context? Well, based on the current level of Goldman’s financial conditions gauge, Briggs estimated that “the peak drag on GDP growth from the cumulative tightening in financial conditions since the start of the year is happening now and will fade in 2023.” The figures (below) illustrate the point.
That’s the good news, as it suggests the worst for the growth drag could be nearly over, even if the total, cumulative hit to the level of GDP won’t be realized until well into 2023.
The bad news is twofold. First, the Fed needs to do more than engineer growth drag to tame inflation. Specifically, they need to cool wage growth at a time when labor is scarce and the labor force is unlikely to return to its pre-pandemic, demographic trend. The GDP drag has to translate to a reduction in labor demand, which should eventually cool wage growth and, finally, inflation. The lags along that chain of events are longer.
Goldman leveraged their own estimates of the impact of slower growth on the jobs-workers gap, the effect of changes in that gap on wage growth and, finally, the read-through for consumer price inflation. “Although 50% of the cumulative effect of a US FCI shock on the level of GDP is realized within two quarters, the impact on other indicators generally takes longer,” Briggs said, on the way to quantifying the lags.
Goldman found that “it takes 2-3 quarters for half of the impact on the jobs-workers gap to be realized, 3-4 quarters for wage growth, and a significant additional lag for wage-driven moves in inflation.” The figure (above) illustrates those lags.
The other bad news is simply that if the growth drag has already peaked, then the Fed will probably need to work harder in 2023, particularly given that the lags for labor, wages and inflation are longer.
“The fading drags from both monetary and fiscal policy tightening and a rebound in real income growth may lead growth to surprise to the upside [meaning] the Fed might need to raise the funds rate above our current forecast of 5-5.25% in order to keep growth below potential and remain on track to tame inflation,” Briggs went on to say.
The quotes below are Friedman’s.
“There is much evidence that monetary changes have their effect only after a considerable lag and over a long period and that the lag is rather variable. In the National Bureau study on which I have been collaborating with Mrs. [Anna] Schwartz, we have found that, on the average of 18 cycles, peaks in the rate of change in the stock of money tend to precede peaks in general business by about 16 months and troughs in the rate of change in the stock of money to precede troughs in general business by about 12 months…. For individual cycles, the recorded lead has varied between 6 and 29 months at peaks and between 4 and 22 months at troughs.”
“Comparison of [changes in monetary policy] with the level of business is likely to be more meaningful for cyclical analysis than either of the comparisons Culbertson and others prefer. By the level of business we generally refer to a flow: The number of dollars of expenditures per year; man-hours of employment or unemployment per year; cars produced per year-all magnitudes having the dimensions of dollars or physical units per unit of time.”
Briggs was unequivocal in his assessment. “As correctly interpreted, Friedman’s 12-16 month estimate of the time until changes in monetary policy have their peak impact on the level of GDP is remarkably consistent with our estimate of a peak growth impact after 2 quarters but a peak level drag after 6 quarters,” he said, adding that “the commonly held view that monetary policy changes have a very lagged impact on economic growth therefore seems to largely reflect a gross misinterpretation of Friedman’s original comments from 60 years ago.
Goldman also looked for empirical evidence to support the view that policy tightening affects economic growth with long (as opposed to shorter) lags. Briggs didn’t find much.
The irony in all of this is that Friedman would probably suggest we’re all missing the point: We didn’t talk enough about the money supply while calibrating subsequent rounds of fiscal and monetary support following the initial pandemic emergency response, and to this day, we’re still not talking enough about the money supply when we discuss how to curtail inflation.
Home prices are too high. Rents are too high. Food prices are too high. Drug prices are too high. The cost of health insurance is too high. Tutition is too high. The U.S. economy needs a dose of deflation. And with the labor market as strong as it is, now is the time to let some air out of the Bubble.
I have been keeping an eye on the velocity M2, historically low. From my education in the 70s it was one of the things to keep an eye on about inflation. It is also why I have not gotten too worried about this inflationary occurrence. Pandemic labor shortages, and a hot war in Europe. Bassman and some of the other writers you’ve referred to seem to resonate with me.
Heavy cash and see what it looks like in March.
“Heavy cash and see what it looks like in March.” That sounds about right. I put new income to work when the opportunity arose — bought six new insured, call protected 5% AA munis. Rates on these same kinds of issues have recently dropped to 4.25% so now we wait. I haven’t sold anything to raise cash but I’m now letting my income pile up. Should be a nice pile in March or April.
Just on the Goldman Financial Conditions index, looking at a historical time series suggests that cycle peaks (i.e. peak ‘tightening’) tended to occur when the Fed was actually cutting rates or policy rate expectations were falling (i.e. during recessions and major growth downturns). This suggests that the credit cycle has a much bigger impact on such indices than the policy rate cycle (to the degree they have a bigger impact on credit spreads and risk asset valuations, implying that during a true recession the Fed, at least initially, cannot cut rates fast enough to drive an easing in the FCI). Makes you wonder if such indices are leading indicators to the real economy or are driven by them.
Back to Friedman, I do think we spend too much time talking about the “price” of money and not enough about the “quantity” given it’s the latter that matters far more for the real economy. Perhaps it’s because defining the money supply or quantifying it is so difficult in 2022 that we take the shortcut of using interest rates as defining whether monetary conditions or ‘tight’ or ‘loose’ (Friedman would refer us to the “interest rate fallacy” most likely).
If I correctly understand GS’ charts and this post, the FCI tightening delivered to date may produce peak deceleration of PCE inflation in mid 2023 with the remaining deceleration during 2H23, while peak deceleration in GDP growth is occurring right about now and the remaining deceleration is to come in 1H23.
This seems inconsistent with investors’ apparent belief that the deceleration in GDP growth will compel the Fed to start cutting rates sometime in 2023. If we’ve already seen half of the hit to GDP, the full hit won’t be big enough to pressure the Fed to start easing before reaping the full inflation impact of its labors.
In other words, “higher for longer”.
In other other words, what the FOMC is saying it will do might indeed be what it does.