The Great Risk to Our Prosperity Is Not the Next Recession

According to the yield curve, there is an almost 11% chance of recession in March 2019.  That’s the highest derived probability since January 2009. But it’s not really the yield curve that makes this claim, rather it is one model preferred by certain staff members at FRBNY. Using specific parts of the Treasury curve, economists have developed statistical techniques the output of which they believe can be a useful binary indication of what’s ahead twelve months from today.

This is pretty much what people tend to focus on when they think about the Treasury curve. Is it inverted? There is some good reason for hyperattention. Each and every recession has been preceded by the condition. Whenever yields on the long segment fall below those of short duration, trouble almost certainly lies ahead.

The yield curve used to do so much more than make a simple toggle. And we used to care so much more about the state of the economy overall than whether it was in recession or not. But as the yield curve has been reduced to this binomial signal, so, too, has thinking about the economy.

Everything is subject to gradation, particularly any parameters belonging to a complex system. There is a human tendency toward shortcuts that is understandable in trying to comprehend, measure, and diagnose the complex. Rules of thumb and condensed guideposts are far easier than robust study and acknowledgement.

That’s not the way it used to be, however. Economists decades ago had developed models intended to forecast economic conditions of all kinds. As an outshoot of quantity theory and econometrics (Positive Economics), there was a pervasive belief that if enough information could be collected computing power and mathematics could be useful in predicting detailed outcomes.

There remain such “leading indicators” today, but they have largely been dismissed or ignored in favor of “recession or not.” One big problem with them, or their evolution, is that nothing stays the same. The world is not static, and therefore it is entirely awkward to expect one set of equations derived from an unchanging variable set to be able to describe the world of tomorrow.

The later forecasting models had largely turned to asset prices from which to glean useful information. This stood in stark contrast to the way previous attempts had originated. The earlier calculations were done using money supply aggregates almost exclusively. The use of asset prices as a substitute came about in the late seventies and early eighties.

Why did scholars feel it necessary to shift emphasis at that specific period in history? Economists James Stock and Mark Watson (the same duo who coined the term Great “Moderation”, though their version didn’t include the quotation marks but did acknowledge there might be some reason for them in the future) researched the topic in 2003. For a paper published in the Journal of Economic Literature, Stock and Watson noted:

“This research on asset prices as leading indicators arose, at least in part, from the instability in the 1970s and early 1980s of forecasts of output and inflation based on monetary aggregates and of forecasts of inflation based on the (non-expectational) Phillips curve. One problem with using monetary aggregates for forecasting is that they require ongoing redefinition as new financial instruments are introduced.”

It's the same “proliferation of products” that Alan Greenspan used in June 2000 to justify the Federal Reserve’s monetary blindness. In this context, the same weakness is being characterized almost as some kind of virtue. Almost.

In shifting from a monetary focus to an asset price one, the reasoning was simple. If you couldn’t define money in an academic setting, merely go to those who have no choice. Markets will have to react to monetary conditions however money may ever be defined. Thus, we might realistically assume that market prices include attention and study of monetary details and inputs.

Monetary policy did something similar at the same time. Rather than base monetary policy on money, policymakers made the same assumption that they wouldn’t have to. They could rely on economic aggregates like GDP and inflation, even market prices, too, presuming that whatever the money definitions and money conditions in the real economy real economic participations would have reacted to them and settled in their outcomes.

The problem with asset price indications, anyway, was that no one could ever find a useful universal model. They end up being, as Stock and Watson noted in 2003, largely unstable.

“This empirical finding of instability is consistent with our reading of the literature on asset prices as predictors of output and inflation, in which an initial series of papers identifies what appears to be a potent predictive relation that is subsequently found to break down in the same country, not to be present in other countries, or both.”

As the authors point out, in some ways this is “surprising.” Markets are assumed, in the orthodox neo-Keynesian framework, to be efficient. These are “cornerstone ideas of macroeconomics.”  Irving Fisher’s hypothesis, as Stock and Watson point out, says that stock prices reflect only “expected future earnings” while at the same time the conventional view of bond markets is where “high interest rates lead to an economic slowdown.”

Both of those interpretations are obviously problematic, which is why the yield curve has been reduced to a binary indication and stock prices are, well, more frequently frothy.

It’s the second of these “cornerstones” that should be more troubling, however. Stocks may or may not be in a bubble, again, but interpreting the bond market is a far more fundamental proposition. Start with the realization that “bond trading” encompasses an order of magnitude more than stock trading ever will or could. The stock market often takes up all our attention, but apart from the aggregate arrangement of 401k holdings people ignore bonds because they’ve been told they are boring, play no wider role, and even when taken seriously they are so as some kind of big unsolvable conundrum.

In the big picture, stocks mean little while bonds mean everything. The movie Wall Street made stock traders into the kings of finance, but in reality, the eighties were really ruled by the bond kings. Stock and Watson’s “new financial instruments” as Greenspan’s “proliferation of products” had nothing whatsoever to do with stocks.

That didn’t change in the nineties. Rather, what happened over subsequent decades was this monetary strangeness becoming even more bizarre, and brazen. In between the eighties and nineties was the S&L crisis. This major banking event was in many ways the final trigger for the modern monetary system.

Officials had struggled with it ever since the first contours of the emergency emerged in the mid-eighties. By early 1991, the matter was injected with great urgency amidst an ongoing recession. Many policymakers were deathly afraid that the prospect of recession-level economic retrenchment combining with so many bank failures and a loan crisis could risk repeating the early thirties.

The matter ultimately rested upon the idea of capital. Not capital in the traditional, proper sense of capitalism, but a modern definition of it specific to banking. Capital is a balance sheet item, one which is supposed to define how much surplus the firm has obtained, either as a result of raising “capital” or retaining profits both to withstand firmly against any future losses. We’ve become since 2008 more intimately familiar with the inner workings of capital ratios, and this is where it all started.

As a general matter, the S&L crisis could be summed up by uncertainty over losses. Unlike in banking events of the past, this one was mostly mortgage assets (and more commercial than residential). There was no, or wasn’t enough, monetary problems with banks so much as solvency. Did they have enough capital to withstand projected future losses? Issues with deposit runs were few.

It was often the case that they didn’t have enough capital. Hundreds of banks were closed on the basis of having too little, not enough internal cushion to absorb contemporary levels of loan losses let alone any further escalation of them (it was a recession in 1990-91). The debate ultimately centered around the idea of how much capital was “enough.”

It was all very different from how banking used to operate, though entirely too familiar in what would happen less than twenty years later. Bank examiners considered capital ratios before the eighties, but deposit safety was really a matter of deposits – cash. In a capital-centric system, everything, and I mean everything, is fungible.

The later system grew out of the understandable commitment to try and anticipate the weaknesses of the prior one. In other words, if you are down to counting cash in the vault it’s too late. It would make sense to try and anticipate depositor runs by studying and measuring the factors that might contribute to them. Capital ratios were designed with that in mind, to let depositors know which bank possessed the wherewithal to stay solvent even in times of trouble, avoiding a run altogether.

One New York Times article published in February 1991 testified to the trickiness of the modern issue.

“The bankers counter by saying that the analysts' estimates [of losses] are exaggerated. Even if all those losses materialized, they would probably occur over a period of time, during which banks could build up reserves from profits, the bankers say. Citibank officials add that the analysts' estimates are wrong because they do not recognize the full extent to which the bank company has already written off troubled loans.”

Capital therefore takes on future rather than present properties. How much capital you have today only matters compared to what you expect will happen tomorrow. This shift toward forward properties introduced probability theory, balance sheet management, and all the things that might be invented to manage not the capital ratios of today but a narrowed expected range for them tomorrow. These were, of course, the “new financial instruments” and the “proliferation of products.”

What we are really talking about is the intersection of money and markets (and often money markets). Again, the New York Times in 1991:

“The steep decline in real estate markets in the last two years has led bank examiners to conclude that many bank loans backing those projects have also declined in value. They have forced banks to write off many loans as a loss and build their reserves to cover future losses.”

Bankers objected, strenuously. They complained to the Office of Comptroller of the Currency, among other government agencies state and local, and found some support. Robert Clarke, Comptroller of the Currency, told the Times, “We don't want the environment bankers face to be more difficult than it has to be.” It was suggested that examinations had become less stringent already by the time Clarke spoke to the paper.

But if market prices are market prices, and therefore definitive, then why any disagreement?  Probabilities, rather the counting cash, is a very different thing. Banks wanted to mark capital losses when they occurred, not when the market price suggested they should. Experience suggested losses weren’t nearly as bad as prices had fallen.

It had often been a problem where illiquid markets produced prices that didn’t seem to conform to real perceptions. Thus, the question became were market prices reflecting fundamental properties of loans and losses, or that of illiquidity? Illiquidity is, obviously, a monetary component.

If you don’t know much (or anything) about money, then how can you know much (or anything) about illiquidity?  And if that’s the case, then you probably won’t be able to interpret market prices correctly (assuming there is such a thing as a “correct” price).

Or, as Stock and Watson would write, “It is one thing to understand ex post why a particular predictive relation broke down; it is quite another to know whether it will ex ante.” We need to know why things act the way they do.

The reason so much time and attention had been devoted to monetary study was that its role in any non-barter economic system was crucial, central even. Karl Marx would advocate for total socialist takeover of every central bank on that same basis. In the sixties, the Samuelson/Solow position of the exploitable Phillips Curve was propositioned on the basis of money buying a controllable increase in prosperity (I’m oversimplifying). Even William Jennings Bryan as a populist was monetary first and foremost (silver).

Then, in the seventies and eighties? People stopped caring about the topic. Economists threw up their hands in the face of substantial complexity and intellectual challenge. They opted for nothing but shortcuts.

No wonder the yield curve has been left as nothing more than a binary toggle. To try and understand anything more than recession/not recession from it would be to admit there is too much “we” don’t know.

At its December 2017 meeting, the FOMC noted the serious flattening going on all across the Treasury curve. The minutes of that meeting reassure us policymakers are as yet unconcerned (mostly):

“They generally agreed that the current degree of flatness of the yield curve was not unusual by historical standards.”

This is true in the technical sense, which is to say that it is at best misleading. If you view the yield curve as only the binary indication, so be it. The real issue with it isn’t flattening by itself, but where in the nominal range it has done so.

Central bank officials are mindful of this fact, too, though they do their utmost to also downplay its significance.

Federal Reserve Bank of Chicago President Charles Evans who, along with Minneapolis Fed President Neel Kashkari, voted at that December meeting against the “rate hike”, tried to describe to reporters at a press conference in January why he did.

“I think we have to be mindful of the fact that as we have all repriced real interest rates downwards that’s going to find its way into lower long term interest rates. We’ve been increasing short-term rates; it’s natural then almost mechanically for there to be a flattening of it.”

There’s nothing mechanical about it. By claiming as much he strips it of necessary meaning, leaving it out of interpretation. In fact, the bond market as he says is agreeing with his dissent, even if Evans doesn’t really appear to understand why. The policymaker noted that the Fed has been forecasting rising inflation, and economic fortunes, for years and they haven’t panned out. Low long-term bond rates made that same determination long before the FOMC started to influence the short end in December 2015.

Tightening, as it has turned out over the last decade, doesn’t require high interest rates nor any macroeconomic cornerstones of orthodox theory. The changed focus of bank capital is central to figuring out why, at least in this current incarnation.

Both parties agree that there is the non-trivial risk the Fed doesn’t know what it’s doing. Employing instead economic aggregates as shortcuts, inflation primary among them, they don’t even take care to appreciate when these very aggregates tell them they are missing a huge part of the economic picture – just as the non-inverted but flattening yield curve has been suggesting for years now.

We’ve reached an almost Orwellian state when it comes to these things; officials admit they don’t know how to define money, so they assume asset prices and economic aggregates will show them the consequences of the real monetary conditions they can’t define or measure. Thus, they need only pay attention to those as to judgments over monetary policy effectiveness and the true monetary condition.

Except, when those asset prices and aggregates tell them they are entirely wrong, what do they do? They now say these don’t matter because the future will surely vindicate their every expectation even as those expectations derive from nothing other than total monetary ignorance. By focusing exclusively on nothing but yield curve inversion, a whole lot of necessary context is lost. So, policymakers can claim the bond market isn’t to this point against them because it isn’t yet inverted when in fact it has been against them for years already. 

The great risk to our prosperity is not the next recession, it is the last one that proved to be more than a recession. That’s what the flat yield curve proposes, and has proposed for a very long time, inverted or not. The great thing about it, and what was appreciated in simpler times when monetary conditions were studiously considered, is that it tells us not just what is wrong today but also why.