July 26, 2013
Can The Fed Prevent The Next Recession?
“None of the U.S. expansions of the past 40 years died in bed of old age,” MIT economics professor Rudiger Dornbusch famously observed in 1997. “Every one was murdered by the Federal Reserve.” Researchers tend to agree, as shown by numerous studies that link inverted yield curves with economic contractions. But does the history of the business cycle and monetary policy in the decades prior to the Great Recession of 2008-2009 still resonate today? In other words, what are the odds that the next recession will be a byproduct of monetary policy decisions, intentional or otherwise?
I’ve been running such questions around in my head for a while, pondering why and how the business cycle could make another trip to the dark side. There’s minimal risk on that front at the moment, as last week’s review of indicators suggest. But assuming that the business cycle hasn’t been repealed (a reasonable assumption, to say the least), we’ll see another downturn one day. But given the Fed’s extraordinary efforts of late in keeping that risk to a minimum, is the probability of a new recession lower than it might otherwise be?
It’s an interesting question for a number of reasons, starting with the fact that the central bank’s posture is somewhat unique, particularly for this stage in the business cycle. It’s been four years since the last recession ended and yet monetary policy, at least by the standards of history, is unusually accommodative. One clue is the year-over-year real (inflation-adjusted) change in the monetary base, which was higher by roughly 20% through last month vs. a year ago. That’s not all that high relative to the last several years, in the wake of the economic and financial crisis of 2008-2009. But in the grand scheme of modern Fed history it’s fair to say that Bernanke and company are pumping up the monetary base at an unusually high pace.
Looking at the chart above inspires the question: Can we have a recession if the monetary aggregates are rising at a robust clip? If the Fed is actively determined to keep the economy out of the cyclical ditch, are the odds favorable for thinking that it’ll succeed? I ran the question by Robert Dieli, an economist who publishes macro research at NoSpinForecast.com. He says that if we take the Fed out of the picture as a real and present danger for the business cycle, it’s only prudent to look for other catalysts that could trigger a recession. Not surprisingly, there’s no shortage of possibilities. But as Dieli reminds, we don’t live in an economic vacuum. Whatever scenario you can dream up for risk factors that could, in theory, push the economy over the edge, the question becomes: How would the Fed react?
Meantime, there are few, if any, smoking guns in the here and now, starting with the biggest kid on the block. “There’s a list of things you hear when the Fed is on the warpath, and we’re not seeing any of those,” Dieli says. Indeed, the yield curve isn’t inverted, nor is it anywhere close to being inverted. As of yesterday, the 10-year Treasury Note’s yield is 2.61% vs. roughly zero for a 3-month T-bill. Never say never, but for now it’s hard to fathom how short rates could move over a 10-year rate any time soon in the current climate.
Some analysts say that looking at the yield curve and related measures is no longer relevant because the Fed is “manipulating” short rates. Isn't that always true? In any case, some economists say that the reliability of the yield curve as a risk factor in business cycle analysis has been rendered null and void. But is this circular reasoning?
It’s clear that recessions in the past have been associated with tighter monetary policy. But that factor has been taken off the table for the foreseeable future, which implies that recession risk is low. Yet some analysts look at a positively sloped yield curve and say that it no longer matters, which amounts to arguing that it’s different this time.
Perhaps. But if it is, we should expect to see some trigger (or triggers), apart from tighter monetary policy, that elevate recession risk. In that case, would the Fed respond? Could it respond? If you accept that the central bank’s monetary policy can be effective when the policy rate is zero, then the answers to those two questions are likely to be “yes” and “yes.”
All of which brings us down to the real question: Can we have a recession if the Fed is keenly focused on keeping one at bay? No one really knows the answer, in part because the Fed has often been a key factor, perhaps the leading factor, in causing recessions. But this time really could be different.
Imagine that in six months we see a chain of events that scares the heck out of consumers and retail spending drops like a rock. In turn, that leads to a deep round of layoffs and unemployment starts moving higher in rapid fashion. The Fed sees this and rolls out a new and aggressive round of quantitative easing. Does that nip the recession in the bud?
There are a number of reasons for thinking that the Fed would be successful in heading off a cyclical attack. But in the realm of macro one can never really be sure. In some sense we’re in uncharted territory. The expansion, modest and unsatisfying as it is, is now 48 months old, by the NBER’s accounting. That’s still below the average expansion since World War II, which weighs in at 58 months. According to the Dornbusch doctrine, age has nothing to do with recession risk. But if the Fed isn’t a risk factor, what does that say about the business cycle for the foreseeable future?
In some sense it’s back to the drawing board. What are the causes of recessions generally? Economists have never really come up with a satisfying answer, which is probably one reason why recessions are a constant through time. Sure, we can point to the telltale signs, such as rising unemployment, lower spending, and tighter monetary policy. But what’s the source of those changes? If you spend any time looking for robust answers, you quickly realize that cause and effect in this corner of economics is a bit like walking in a house of mirrors.
This much, at least, is clear. If we do have a recession while the Fed is aggressively trying to avoid one, we’re in deep trouble.
Posted by jp at July 26, 2013 7:48 AM
So you're arguing that the real monetary base will have little or no influence as it relates to the business cycle going forward because of changes in the central bank's policy post-2008. You may be right... or not. Only time will tell, but I certainly remain open to the possibility.
More generally, I'm skeptical that any one indicator is going to tell us all that we need to know in my primary objective in these investigations: developing context for evaluating recession risk in real time. As such, I'm routinely looking at a diversified set of indicators. Indeed, as you note, there's always a good reason to question if the historical record still applies to any one data set going forward. No one really knows if, or how, it will be different the next time, and whether any differences are permanent or just a quirk of a particular historical period. That said, I'm a bit less skeptical of looking at a broad array of indicators. For example:
One other point: note that the real monetary base has briefly gone negative twice on a year-over-year basis since the recession ended in 2009. We didn't have a new recession, but the negative comparisons coincided with heightened macro risk, albeit not necessarily due to monetary conditions alone. The implication: monitoring this indicator is still productive, even if it's unlikely to be flawless, which is why broader context is always necessary.
Posted by: JP at July 29, 2013 7:04 AM
JP, the nature of base money changed in 2008. Pre-2008, the Fed created base money (at a profit) to satisfy the demand for money (a large quantity of currency and a tiny quantity of bank reserves). Post-2008, it creates base money (at a loss) to fund its discretionary long term investments. Since this base money earns a competitive rate of interest, there's no necessary relationship between the quantity and the price level.
If you look at currency only (not currency+reserves), then it's not as meaningless, because currency still doesn't earn a competitive rate of interest (unlike reserves). So nothing fundamentally changed in 2008 as far as currency is concerned.
Posted by: Max at July 28, 2013 8:50 PM
Well, yes, the monetary base is always "distorted" by Fed actions, and that's the point. We want to know how the Fed's "distorting" the base. If the focus is on business cycle analysis, the first question is how is the Fed adjusting that slice of money supply over which it has direct control. The monetary base offers an answer. MZM and other broader measures of money supply are infected with money outside the Fed's direct and immediate control, and so those measures provide less timely insight for connecting the dots in terms of Fed activity and the business cycle.
For example, going into the Great Recession of 2008-2009, the real (CPI adjusted), 1-year rolling % change for MZM was positive and rising, implying that the Fed was supplying adequate liquidity. As we know, however, that wasn't true at all, as shown by the real monetary base 1yr % change, which was negative going into the recession.
Posted by: JP at July 28, 2013 8:24 AM
The monetary base is the wrong thing to look at, since it's distorted by QE. If you must look at money, use MZM.
Posted by: Max at July 27, 2013 1:30 PM
Well, I agree... sort of. At the risk of nit-picking, you can play with the data and argue that there's more of a gray area for those four US recessions that you cite. For instance, using daily data, the curve inverted ever so briefly for a few days in 1989, based on daily constant maturities for the 10-year and 3-month T-bill via the St. Louis Fed database. But that's a year ahead of the recession. Nonetheless, the curve did move relatively close to inversion overall, a not insignificant event.
Also, if we use effective Fed funds as a proxy for the short rate, and compare it with the 10-year yield on a daily basis, then the curve inverted rather convincingly, and for most of 1989.
And if we look a rolling 1yr % changes in the real monetary base, as the chart above reminds, cause and effect looks a bit clearer.
Nonetheless, your point is well taken. Still, the Fed's fingerprints are on most if not all scenes of the crime, so to speak.
Posted by: JP at July 26, 2013 5:47 PM
James, be aware that of our last ten recessions, four of them had no inverted yield curve. Those were the 53-54, 57-58, 60-61 and 90-91 recessions. Japan's last three recessions did not have a inverted yield curve.
Interest rates are the main reason that recession occur in my opinion, but the long end of the curve can also cause the problem. Since 1947, we have had 18 large increases in long term interest rates, and 11 of them gave us a recession. Long rates have moved up enough that this is number 19. We are already seeing the impact of rising interest rates on the housing market, as Housing Starts have already dropped by over 15% from there most recent peak. We are seeing lumber prices confirming this as well.
With 2nd quarter GDP expected to be only around 1%, and we are going to have to face higher interest rates and higher gas prices, along with higher taxes, and our savings rate just about at a recovery low, I think the chances that we will be back into recession by the end of this year is very possible.
The last two times we had a large increase in long term interest rates, Dec 08-June 09 and August 2010-February 11, on a 3 quarter lag, we got a 0.1% and 0.4% GDP. If history repeats itself, we should expect close to zero GDP by the first quarter of 2014.
If a recession is coming, I do not believe that the Fed can prevent it from happening, no matter how low 3 month T-Bills are, or how much QE is out there.
Posted by: Tony Ferreira at July 26, 2013 4:56 PM