Monday, March 16, 2020

What Sort of Financial Panic Is This?

Well, as you know, things are changing by the day in financial markets, and central banks are moving quickly to keep up. Most central banks have taken aggressive action recently. After a between-meetings policy rate cut of 50 basis points on March 3, the Fed yesterday reduced its target range for the fed funds rate to 0-0.25%. More to the point, the interest rate on the Fed's overnight reverse repo facility is set at 0%, the interest rate on reserves is 0.10%, and the interest rate on primary credit at the discount window has been reduced to 0.25%. The interest rate target moves, in timing and magnitude, are very aggressive. This was done on a Sunday, before financial markets opened on Monday, and amounted to a 100 basis-point drop in the target range. The discount window rate (the rate at which the Fed lends to Fed member financial institutions) was cut even more, as this rate moved from 50 basis points above the top of the fed funds rate range to the top of the range.

There's more. The Fed lists actions that are intended to enhance the functioning of credit markets, though some of these "actions" are simply encouragement to banks to use the Fed's credit facilities - the discount window and intraday central bank credit - and to lend in instances where the bank has liquidity and capital in excess of regulatory mandates.

One item that stands out is the discontinuation of reserve requirements in the United States. Reserve requirements were dropped long ago in some countries, and currently Canada, the UK, New Zealand, Australia, Sweden, and Hong Kong, are without reserve requirements. In the US, reducing required reserve ratios to zero is essentially a formality. US financial institutions have figured out workarounds, including sweep accounts, that allow them to skirt the effects of reserve requirements, and the Fed's large balance sheet has made reserve requirements non-binding for most banks. The elimination of reserve requirements in the US is long overdue, but why now? Possibly what's going on is that liquidity requirements (the liquidity coverage ratio requirement, for example) exclude required reserves, so eliminating the reserve requirement does in fact relax a constraint on banks, and could encourage them to acquire other assets - loans or asset-backed securities, for example.

Closer to (my) home, the Bank of Canada has reduced its policy rate by 100 basis points, to 0.75%, in two steps, and has announced some moves to enhance market liquidity, as outlined here. But I'll focus on the US, where most of the action is.

What is the Fed responding to? Most people are aware of the downward slide in the stock market, which continues today. But there are various other signs of distress in financial markets. This has little in common with traditional banking panics, such as what the US banking system encountered in the Great Depression, or prior to the founding of the Fed, between the Civil War and 1914. The primary feature of such panic episodes as the 1907 panic was runs on retail bank deposits, and disruption of retail payments. In principle, such panics could be mitigated, or stopped entirely, through central bank lender-of-last-resort lending, for example through the discount window in the United States. Indeed, a principle function of the Fed, as the authors of the Federal Reserve Act intended, was to lend to banks during financial crises. The Fed could lend to banks, replacing the outflow of deposits, and could finance this lending by issuing more currency, thus giving consumers a safe means of payment to flee to. But what we're seeing now is certainly not a flight from bank deposits to currency.

The current panic does have something in common with the financial crisis of 2008, though there are important differences. During the 2008 financial crisis, the origin of the crisis was in the financial sector. Incentive problems in the mortgage market ultimately led to a loss in confidence in the value of a class of asset-backed securities, which fed through to drops in asset prices, a flight to safety in financial markets, and chains of defaults and potential defaults. The current crisis is rooted in the non-financial sector, but some of the same elements are in play. The reduction, or anticipated reduction, in income flows for businesses and individuals, has caused those businesses and individuals to re-evaluate their portfolios, generating a desire to trade in asset markets. But at the same time, everyone is very uncertain about outcomes, and where the risks are. So perceived risk is high, in general, market participants are fleeing to safe assets, and people also want to sell liquid assets to adjust their portfolios in ways they perceive as optimal. What's happening? First, as in the financial crisis, there's been an increase in interest rate spreads. Here's the difference between the 3 month commercial paper rate and the T-bill rate:
You can see that this spread has increased substantially, but it's nowhere near as large as at times during the 2008 financial crisis. Similarly, if we look at a risky corporate bond spread, we get the same story, more or less:
Though note that the data in the last 2 charts only runs to late last week, so things could have changed for the worse.

As well, markets in US government Treasury securities are not functioning normally. Apparently the bid-ask spread in markets for Treasury securities has widened, and volume has declined, making it more difficult to buy and sell Treasuries. The liquidity of this market is important. There has to be some asset that is easy to buy and sell, at a predictable price, and if government debt is not filling that role, then nothing is.

Which brings us to a key part of the Fed's announced intervention plan, which is the resumption of quantitative easing (QE), after a sort-of hiatus of five years or so. Over the next several months, the Fed plans to purchase at least $500 billion in Treasury securities, across maturities, and $200 billion in agency mortgage-backed securities. That's a program about the size of QE2, which ran from 2010-2011, but not as large as QE1 (2009-10) or QE3 (2012-14). This program would amount to a nominal increase of about 18% in the Fed's asset portfolio, or about 3.2% of annual GDP. Not small potatoes, but the Fed has done this on a larger scale before. What's the rationale? The Fed has already expanded its repo program. The Fed is currently lending in excess of $100 billion in the overnight repo market every day, and is also lending substantially in term repo markets. So, if you want to unload some Treasuries and are having trouble doing it, the Fed will lend you the cash. Similarly, the Fed's reverse repo facility is still in place, so if you want to buy Treasuries, but are having a hard time doing it, you can lend the cash to the Fed through the reverse repo facility, say if you're a money market mutual fund, or you can just hold interest-bearing reserves if you have a reserve account.

But, possibly there's a role for QE in purchasing on-the-run Treasury securities that financial markets are having a hard time absorbing. In this case, the purchases of Treasuries potentially make life easier for the Treasury. The Fed purchases the Treasury securities, increasing the balance in the Treasury's general account (TGA) with the Fed, and as the Treasury spends the balance, these funds end up in bank reserves. But the Treasury is currently holding a balance (as of last week) of about $380 billion in the TGA account, which is enough to fund Treasury outlays for a typical month. So, the Treasury has a substantial buffer if it has trouble selling its debt, and such difficulties aren't likely to last long.

So, why are we into another QE episode? Principally, the FOMC thinks this is an accommodative policy - it's supposed to increase real GDP and inflation. Of course there's no evidence for that - from experience in the US or elsewhere. And QE could actually be harmful, in that it's fundamentally a swap of less useful assets - reserves - for more useful assets - Treasuries and MBS. It's possible that more reserves and less Treasuries and MBS in the market right now would be a good idea. But why commit to this asset purchase program over the next several months? The Fed could at least hold off decisions about continuation of the program until we have more information.

Here's what I'm afraid of. The Fed's actions may be appropriate in the moment, but I'm worried about how we get back to some semblance of normal again. The Fed had a set of normalization plans as early as 2011, to undo policies that were put in place during the financial crisis. But the FOMC never followed through on that normalization plan, and ultimately chose to stick with its large balance sheet. Interest rates were never normalized, in the sense of attaining a level that would sustain inflation at 2% indefinitely. Now we're back at the zero lower bound, with plans for a substantial Fed balance sheet increase, no solid science telling us why this is a good idea, and no plans for getting out of this once this virus gets out of town.

Thursday, March 5, 2020

Coronavirus and Monetary Policy

On Tuesday, the FOMC voted to reduce the Fed's range for the fed funds rate by 50 basis points, to 1.00-1.25%, and on Wednesday, the Bank of Canada followed suit with a 50 basis point cut in its overnight interest rate target, from 1.75% to 1.25%. As you know, 50 basis point cuts by central banks are aggressive, particularly in the Fed's case where the cut occurred outside a regularly scheduled FOMC meeting. Other central banks might be doing the same - if their policy rates were not already at or close to zero, if not negative.

What's going on? Well, Jay Powell and Steve Poloz have told us why they're doing what they're doing. What do they have to say? First, in the FOMC statement that Powell and his colleagues issued, it says that the "fundamentals of the US economy remain strong," but the coronavirus "poses evolving risks." The FOMC is cutting 50 basis points "in support" of the Fed's dual mandate. There's no information beyond that. In his short press conference, Powell didn't really add to that. Basically, the cut is to lend "support," apparently, and there are no promises about where the Fed goes from here. And, by the way, this has nothing to do with what Donald Trump wants, according to Powell.

So, that doesn't enlighten me much. An extreme action has been taken, but it's not based on anything that's actually showing up in the data yet. This is based solely on speculation about the effects on economic activity of a contagious virus that is going around. About which we know more than we did a couple of months ago, but much less than we would like to know. So, from all evidence, the Fed is taking a shot in the dark.

But maybe Steve Poloz knows something that Powell doesn't know? The Bank of Canada statement certainly has more information in it than does the FOMC statement, in part because the Bank of Canada made its decision in a regularly-scheduled policy meeting. There's a review of current conditions, which basically looks like more of the same. Current inflation (headline CPI) is above target at 2.4%, the labor market looks good, but there is some weakness in 4th quarter 2019 real GDP growth. Otherwise, like the US, the fundamentals are strong, but the statement seems to go out of its way to suggest that there's some weakness, presumably to justify the big reduction in the target rate. With regard to the coronavirus, the statement says:
...business activity in some regions [of the world] has fallen sharply and supply chains have been disrupted. This has pulled down commodity prices and the Canadian dollar has depreciated. Global markets are reacting to the spread of the virus by repricing risk across a broad set of assets, making financial conditions less accommodative. It is likely that as the virus spreads, business and consumer confidence will deteriorate, further depressing activity.
So, I think that's really the guts of the information that's supporting the policy decision. We know there's something going on with respect to the real side of the economy. Anecdotal evidence tells us that supply chain disruptions, originating mainly in China, could be a serious problem for domestic economic activity. We're also seeing large asset price movements. Stock prices are down, and prices of government debt are up. We're expecting to see some negative information in soft data - business and consumer confidence - but we're not measuring that as yet.

But, we still don't have so much to go on. No hard data, really. Asset prices have been known to move by large amounts in the past, without any central bank response. But what's this about financial conditions becoming "less accommodative." I've heard that comment elsewhere, and I'm not sure what it means. The idea seems to be that financial markets get "tight" on their own, and then it's the job of the central bank to loosen things up. Maybe I can see some sign of "tightening" financial conditions in stress indices, or financial conditions indices. Here's the St. Louis Fed's financial stress index:
And here's the Chicago Fed's financial conditions index:
Those two charts show basically the same thing. Not much going on. Maybe people have something else in mind? There certainly have been large movements in the prices of Treasury securities in the US:
The chart shows the 3-month T-bill rate, and the 10-year bond yield, both of which have dropped even more than the drop in the Fed's policy rate. The Fed moved down 50 basis points, but the whole Treasury yield curve has shifted down by much more than that. But why would I call that financial tightening? Looks like easing, don't you think?

Another interesting phenomenon in US financial markets is what's going on in overnight markets. Since last fall, the Fed has been purchasing Treasury bills, and now has a stock of about $280 billion in T-bills. As well, to maintain its overnight fed funds target, the Fed has been intervening in the overnight repo market, on the lending side. On February 27, there was about $140 billion in repos outstanding (overnight and term). However, on Tuesday and Wednesday of this week, the Fed's lending on the repo market expanded substantially. In fact, demand for overnight repos apparently exceeded the Fed's $100 billion cap. So, overnight Fed repos on Tuesday and Wednesday were $100 billion, where last week overnight repos were running roughly $30-$50 billion per day. So, something is causing high demand for overnight credit, and upward pressure on the overnight repo rates, while Treasury yields at all maturities are dropping like rocks. So, two issues here. First, why is the Fed capping its overnight repo lending? If it's so important to hold down short rates, the Fed should lift the cap. Second, the drop in Treasury yields in part reflects a high demand for safe assets. Why is the Fed continuing to buy T-bills? The market wants them, so the Fed should be selling T-bills or, short of that, stopping its purchases.

What does history, and theory, tell us about what central banks should be doing now, in response to the coronavirus? History doesn't give us a lot to go on. In modern times, we haven't had to deal with a global pandemic of this type, with this level of potential widespread economic disruption. But, is this like dealing with other types of large shocks? First, is this like a financial crisis? No. The financial crisis originated as disruption in the financial sector. As such, it had historical precedents, and there is a wealth of experience concerning how a central bank should deal with a financial panic. In a financial crisis, the standard responses involve central bank lending, and bailouts. There is of course plenty of judgement that comes into play concerning who to lend to and how, who to bail out and how, and what the costs and benefits are. But, in general, this is all relatively well-understood. I know policymakers got plenty of flack in the last recession, but in general they were following a time-worn script. Not so much after the recession was over, but that's another story.

Second, for policy purposes, is the coronavirus anything like a war? No. Wars involve a reallocation of labor and capital across sectors, as human beings go into the armed forces, peacetime production goes away, and wartime production ramps up. There are important questions concerning how the expenditure on the war machine is financed. Little of that is going on here, as what policymakers are dealing with is a general cutback in across-the-board economic activity, and a reduction in time devoted to market activities. Public spending comes into play, but it's not of the magnitude that we observe in a war. So, I don't think we can learn anything from wars in this case.

Finally, what about natural disasters? Is that similar? Maybe. Floods and hurricanes, for example, can destroy aggregate economic activity temporarily, so that looks similar. And public spending is required to deal with the disaster. But if we take that seriously, this might tell us that monetary policy should do nothing and central bankers should sit on the sidelines. You may remember some monetary policy response to a hurricane or flood, but I don't.

But what would theory tell us? There are of course plenty of Keynesian models - both new and old - in which output is demand-determined, and therefore output contractions occur because "demand" falls. I guess what people mean when they say the coronavirus is a "supply" shock and not a "demand" shock is that Keynesian models are not equipped to help us understand what is going on in this instance, so let's forget about that stuff. Some people -Roger Farmer for example - like to think about multiple equilibrium models. In a multiple equilibrium world, what policymakers say, or what they commit to, can make a difference. So, perhaps the large moves by the Fed and the Bank of Canada this week could be seen as bolstering confidence. But, those aggressive moves could also cause people to think that a disaster is looming. Maybe the central bankers know something we don't, and we should panic. Go buy toilet paper, instead of that new car.

But, there are regular declines in economic activity that we deal with on a regular basis, apparently without any discretionary response from the central bank. Those regular declines are due to weekends, and seasonals. Every weekend a lot of people produce no market output. Every winter, some types of economic activity become more costly to undertake. Every summer a lot of people go on vacation. But, we've all become so accustomed to staring at seasonally adjusted data, that we forget a lot of that. There's a substantial amount of seasonal variation in a many macroeconomic time series. For example, here's US employment, unadjusted:
Typically, there's a big dip in employment in January, and a smaller one in August. The drop in employment from the busiest month to the least busy month isn't as large as the drop in employment during the last recession, but it's about 25% of that drop, typically, and it happens every year. And, importantly, the central bank is responding to that. Part of what the central bank does is to smooth out seasonal variation in interest rates, and also to smooth out bumps in the payments system, that is predictable disturbances that have to do with the month, the day of the week, or the season. An important part of central bank activity (though increasingly less important) is to supply currency elastically on weekends, in the holiday season, etc., to accommodate demand.

So, that's all just happening automatically, as a result of standard central banking procedure - pegging an overnight nominal interest rate. And that will happen while the coronavirus is making us sick. The key question is whether the central bank should be doing more. Should the central bankers have done what they did - aggressively cut interest rate targets as if they were dealing with a financial crisis, or a large drop in (seasonally adjusted) employment at the onset of a recession?

Jay Powell says that the interest rate cut will "support" the economy. Steve Poloz says that the interest rate cut will help consumer and business confidence. But how? Central bankers are always telling us about Friedman's "long and variable lags." The usual claim is that it takes a long time - maybe up to a couple of years - for monetary policy to have its effects. I don't know whether that's true or not, but it's a standard part of central banking lore. But, if the effects of the coronavirus are going to be temporary - a few months - why do anything if it's not going to matter within the relevant time frame? Maybe the interest rate cuts are supposed to inspire confidence. Typically, we think that interest rate cuts are going to tip the balance for consumers thinking about buying consumer durables and houses, or for firms making investment decisions. But, if the message I'm getting from the central bank is that things could get really bad, I'm going to be inclined to hunker down and postpone big spending decisions. So, it's hard to see what got accomplished this week on the monetary policy front.

Finally, there's the issue of leaving policy space to deal with whatever future shocks might be thrown at us. Central bankers find it very easy to cut interest rate targets, and very hard to increase them. It's clear that, now these rate cuts are in place, central bankers will be slow to move up again, after the coronavirus passes. Non-neutralities of money are temporary - we don't get permanently lower output by having low nominal interest rates forever. So, once we cut rates now, we can't cut them later, and the real effects, such as they are, will dissipate.

So, I'm inclined to think that the Fed, and the Bank of Canada, did the wrong things this week. The potential benefits are too small, and the potential costs too large. At the minimum, central bankers could have waited for more information, or cut less aggressively.