Should central banks do emergency lending?

A barnstorming new paper from the Richmond Fed, written by its President Jeffrey Lacker and staff economist Renee Halter, argues that the Federal Reserve has drifted into doing too much credit policy to the detriment of its traditional goal of overall macroeconomic stabilisation.

In its 100-year history, many of the Federal Reserve’s actions in the nameof financial stability have come through emergency lending once financial crises are underway. It is not obvious that the Fed should be involved in emergency lending, however, since expectations of such lending can increase the likelihood of crises. Arguments in favor of this role often misread history. Instead, history and experience suggest that the Fed’s balance sheet activities should be restricted to the conduct of monetary policy.

The first step in their case is attacking the idea that the Fed was created to be a lender to specific troubled institutions or sectors:

Congress created the Fed to “furnish an elastic currency.”…In other words, the Fed was created to achieve what can be best described as monetary stability. The Fed was designed to smoothly accommodate swings in currency demand, thereby dampening seasonal interest rate movements. The Fed’s design also was intended to eliminate bank panics by assuring the public that solvent banks would be able to satisfy mass requests to convert one monetary instrument (deposits) into another (currency). Preventing bank panics would solve a monetary instability problem.The Fed’s original monetary function is distinct from credit allocation, which is when policymakers choose certain firms or markets to receive credit over others.

They go on to explain further the difference between monetary policy (providing overall nominal stability; making sure that shocks to money demand do not lead to macroeconomic instability & recessions) and credit policy (choosing specific firms to receive support and funds—effectively a form of microeconomic central planning):

Monetary policy consists of the central bank’s actions that expand or contract its monetary liabilities. By contrast, a central bank’s actions constitute credit policy if they alter the composition of its portfolio—by lending, for example—without affecting the outstanding amount of monetary liabilities. To be sure, lending directly to a firm can accomplish both. But in the Fed’s modern monetary policy procedures, the banking system reserves that result from Fed lending are automatically drained through off setting open market operations to avoid driving the federal funds rate below target.

The lending is, thus, effec-tively “sterilized,” and the Fed can be thought of as selling Treasury securities and lending the proceeds to the borrower, an action that is functionally equivalent to fiscal policy.

They go on to explain why Walter Bagehot provides “scant support” for the creditist approach to crisis management, while the facts of the Great Depression do not fit with the creditist story.

Finally, they note that even if there are inherent instabilities in the financial system—something far from proven—many of these are made substantially worse by central bank intervention in credit markets.

Financial institutions don’t have to fund themselves with short-term, demand-able debt. If they choose to, they can include provisions to make contracts more resilient, reducing the incentive for runs. Many of these safeguards already exist: contracts often include limits on risk-taking, liquidity requirements, overcollateralization, and other mechanisms.

Moreover, contractual provisions can explicitly limit investors’ abilities to flee suddenly, for example, by requiring advance notice of withdrawals or allowing borrowers to restrict investor liquidations. Indeed, many financial entities outside the banking sector, such as hedge funds, avoided financial stress by adopting such measures prior to the crisis.Yet, leading up to the crisis, many financial institutions chose funding structures that left them vulnerable to sudden mass withdrawals. Why?

Arguably, precedents established by the government convinced market participants of an implicit government commitment to provide backstop liquidity. Since the 1970s, the government has rescued increasingly large fi nancial institutions and markets in distress. This encourages large, interconnected fi nancial fi rms to take greater risks, including the choice of more fragile and often more profi table funding structures. For example, larger financial firms relied to a greater extent on the short-term credit markets that ended up receiving government support during the crisis. This is the well-known “too big to fail” problem.

I apologise for the length of the quotation, but the paper really is excellent. Do read the whole thing.

Voxplainer on Scott Sumner & market monetarism

I have to admit that I usually dislike Vox. The twitter parody account Vaux News gets it kinda right in my opinion—they manage to turn anything into a centre-left talking point—and from the very beginning traded on their supposedly neutral image to write unbelievably loaded “explainer” articles in many areas. They have also written complete nonsense.

But they have some really smart and talented authors, and one of those is Timothy B. Lee, who has just written an explainer of all things market monetarism, Prof. Scott Sumner, and nominal GDP targeting. Blog readers may remember that only a few weeks ago Scott gave a barnstorming Adam Smith Lecture (see it on youtube here). Readers may also know that I am rather obsessed with this particular issue myself.*

So I’m extremely happy to say that the article is great. Some excerpts:

Market monetarism builds on monetarism, a school of thought that emerged in the 20th century. Its most famous advocate was Nobel prize winner Milton Friedman. Market monetarists and classic monetarists agree that monetary policy is extremely powerful. Friedman famously argued that excessively tight monetary policy caused the Great Depression. Sumner makes the same argument about the Great Recession. Market monetarists have borrowed many monetarist ideas and see themselves as heirs to the monetarist tradition.

But Sumner placed a much greater emphasis than Friedman on the importance of market expectations — the “market” part of market monetarism. Friedman thought central banks should expand the money supply at a pre-determined rate and do little else. In contrast, Sumner and other market monetarists argue that the Fed should set a target for long-term growth of national output and commit to do whatever it takes to keep the economy on that trajectory. In Sumner’s view, what a central bank says about its future actions is just as important as what it does.

And:

In 2011, the concept of nominal GDP targeting attracted a wave of influential endorsements:

Michael Woodford, a widely respected monetary economist who wrote a leading monetary economics textbook, endorsed NGDP targeting at a monetary policy conference in September.

The next month, Christina Romer wrote a New York Times op-ed calling for the Fed to “begin targeting the path of nominal gross domestic product.” Romer is widely respected in the economics profession and chaired President Obama’s Council of Economic Advisors during the first two years of his administration.

Also in October, Jan Hatzius, the chief economist of Goldman Sachs, endorsed NGDP targeting. He wrote that the effectiveness of the policy “depends critically on the credibility of the Fed’s commitment” — a key part of Sumner’s argument.

But read the whole thing, as they say.

*[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16]

The eurozone is in dire need of nominal income targeting

It may well be that, in the US and UK, nominal GDP is growing in line with long-term market expectations.* It may well be that, though we will not bring aggregate demand back to its pre-recession trend, most of the big costs of this policy have been paid. And so it may be that my pet policy: nominal income/GDP targeting, is only a small improvement over the current framework here in the UK or in the US. But there is one place that direly needs my medicine.

As a whole, the Eurozone is currently seeing very low inflation, but plenty of periphery countries are already suffering from deflation. And this is not the Good Deflation of productivity improvements (can be identified because it comes at the same time as real output growth) but the Bad Deflation of demand dislocation. The European Central Bank could deal with a lot of these problems simply by adopting a nominal GDP target.

When it comes to macroeconomics, the best analysis we really have is complicated econometric models on the one side, and highly stylised theoretical models on the other. Both are useful, and both can tell us something, but they rely on suspending quite a substantial amount of disbelief and making a lot of simplifying assumptions. You lose a lot of people on the way to a detailed theoretical argument, while the empirical evidence we have is really insufficient to conclusively answer the sort of questions I’m posing.

In general, I think that very complex models help us make sense of detailed specifics, but that “workhorse” basic theoretical models can essentially tell us what’s going on here. Unemployment is a real variable, not one directly controlled by a central bank, and a bad thing for the central bank to target. But in the absence of major changes in exogenous productivity, labour regulation, cultural norms around labour, migration and so on, there is a pretty strong relationship between aggregate demand and unemployment. Demand dislocation is almost always the reason for short-run employment fluctuations.

Unemployment rose everywhere in 2008-9. But it nudged down only marginally post-crisis in the Eurozone, whereas in the UK and US it soon began to steadily fall toward its pre-crisis rate (the red line, though not on this graph, has tracked the green one very closely). In the meantime the Eurozone rate has risen up to 12%. This is not at all surprising, given the almost complete flattening off of aggregate demand in the Eurozone—this means a constantly-widening gap with the pre-recession trend (something like 20% below it now).

Although intuitively we’d expect expectations to steadily adjust to the new likely schedule, three factors mean this takes a while: firstly the ECB is very unclear about what it is going to do (and perhaps unsure itself), secondly some plans are set over long horizons, and thirdly the lacklustre central-bank response to the 2007-8 financial crisis is unprecedented in the post-war period.

1. We have a huge literature on the costs of policy uncertainty—the variance of expected outcomes has an effect on firms’ willingness to hire, invest, produce, independent of the mean expected outcome.

2. Many firms invest over long horizons. It may have become clear at some point in 2011, when the ECB raised interest rates despite the ongoing stagnation and weak recovery, that the macro planners, in their wisdom, were aiming for a lower overall growth path and perhaps a lower overall growth rate in nominal variables. And so, after 2011 firm plans started to adjust to this new reality. But many plans will have been predicated on an entirely different 2009, 2010, 2011, 2012, 2013, 2014, and so on. And as mentioned before, the gulf between what was expected for the mid-2010s back in 2007 and what actually happened is actually widening.

3. Thirdly, and finally, the period 2008-2010 is unprecedented and will have slowed down firm adjustment substantially. As mentioned above, even if firms set plans with a fairly short-term horizon (a few years) they wouldn’t have been able to adjust to the new normal in 2008, 2009 and 2010 unless they really expected the ECB’s policy of not only not returning to trend level, but not even return to trend rate!

All of these three issues are convincingly resolved by nominal income targeting. It’s very certain—indeed the best version would have some sort of very-hard-to-stop computer doing it. It promises to keep up to trend. And it is very stable over long horizons.

Recent evidence reinforces the view, implicit in our models, that (unconventional) monetary policy is highly effective at the zero lower bound, even through the real interest rate channel (!) All the ECB needs to do is announce a nominal income target.

*This reminds me: isn’t it about time we had an NGDP futures market so we could make claims here with any kind of confidence?

QE boosts equities by boosting fundamentals

Many people suggest that the recovery in equity prices since 2009-2010, seen around round the world but particularly in the American NasdaqS&P500 and DJIA, does not represent a general economic improvement. Instead, they believe that these numbers are simply being buoyed by new money pumped into the system. I don’t think this argument holds, and I will attempt to explain why.

First let’s consider why we think people hold equities. Essentially, people hold equities because they expect a given real return for a given risk profile. In our simplest model, people hold portfolios of assets based on their risk tolerance, their subjective judgements over probabilities, and their preferences. Adding in banks, insurers, pension funds and so on makes the overall picture more realistic, but doesn’t change our theory much. People pick financial intermediaries that hold the assets according to our preferences—the intermediaries add value through scale, or through providing a payments system and settling accounts.

Why might electronic money printing (which we call “quantitative easing” or QE) affect equity prices?

Well, firstly, we might not expect an effect from quantitative easing under one circumstance. QE increases the amount of narrow money we have—that is the number of notes, coins and bank reserves in the system. Generally we think broad money—which includes bank accounts people can debit or write checks on, and is much, much larger—is what interacts directly with the real economy. The ratio of broad money to narrow money is called the money multiplier, and usually a rise in narrow money leads to an even bigger rise in broad money—but this multiplier is not stable. It’s at least possible (although not historically typical) that a rise in narrow money could be completely counteracted by a fall in the money multiplier.

But assuming this doesn’t happen, there are three reasons why QE might boost equity prices. First would be because it increases inflation and the future price level. If prices rise, cash is worth less, so relative to a given nominal amount of cash, all things being equal a given equity is worth more. In other terms, the firms’ nominal expected returns would rise.

The second reason is that in a depressed economy monetary easing like QE may boost real growth, which we would expect to raise any given company’s expected real returns. It might also reduce the risk of very bad economic outcomes. Since equities are riskier than bonds, gilts and cash they pay a risk premium to those who hold them—a higher return (lower price) to compensate for this. If risky outcomes in general become less likely, these risk premia might narrow, making equities more desirable and expensive.

The third reason QE might raise stock prices is because it increases overall social wealth, and thus may lead to greater risk-tolerance overall, if people are willing to bear more risk as they get wealthier, and thus shift towards riskier assets like equities.

In each of these three, the jump in equity prices comes from fundamental factors. One could certainly drive up stocks by creating lots of inflation, but we can easily check if that’s what’s happening by looking at inflation. Any real/relative growth in equities would refute that explanation. In contrast, real growth, reduced risk and shifted preferences due to extra wealth are all legitimate reasons for higher equity prices.

Accounts of why QE buoys stocks without improving fundamentals (and hence part of the argument that stock indices are not good proxies for economic health) tend to rely on a narrative that QE “flows into equities”. But as explained, people try to hold their wealth in the portfolio that fits closest to their preferences. If QE money did “flow into equities” then people would now be holding more of their wealth in stocks than they wanted to—they would rapidly rebalance their portfolio. Typically people needn’t even do this themselves, because their pension fund will do so for them. QE has to improve the fundamental factors in order to boost equities.