Thursday, June 26, 2014

Addicted to central bank painkillers?

Claudio Borio (BIS) and Piti Disyatat write about the dangers of low interest rates at VoxEU. Using an argument that has been put forward many times by the BIS (Claudio Borio and co-authors), low interest rates during booms and expansions can create bubbles and financial instability. Central banks need to be aware of the costs of low interest rates.

The authors, while accepting the idea that low real interest rates might be the outcome of low growth and secular stagnation,  argue that central banks cannot simply be seen as passive agents adapting their policies to the macroeconomic environment; they are responsible for low interest rates. In their words "money and finance are not neutral". Quoting from the article:

"Not only can financial factors – especially leverage – amplify cyclical fluctuations, but they can also propel the economy away from a sustainable growth path. By influencing decisions to invest, variations in financial conditions affect the evolution of the capital stock, and hence, future economic fundamentals. An expanding capital stock during booms may help to constrain inflation and obviate the perceived need for monetary-policy tightening. At the same time, large changes in relative prices that typically occur in financial booms divert resources into surging sectors in ways that are not easily reversible. The long-lasting impact of the financial cycle becomes especially evident in the bust phase. The cumulative build-up in debt and associated resource misallocations – especially the overhang of capital – leave a legacy that takes time to resolve."

To support their claim, the authors produce a chart that shows how debt (public and private) has increased dramatically during the years where interest rate were coming down.

Their conclusion:

"More stimulus may boost output in the short run, but it can also exacerbate the problem, thus compelling even larger dosages over time. An unhealthy dependence on painkillers can be avoided, but only if we recognise the risk in time."

I have no objection to the idea that money and finance are not neutral and that central banks can have an important role in financial markets. But the analysis above is too simplistic and potentially misleading.

The logic it puts forward is that arbitrarily-low interest rates set by the central bank generate an unsustainable behavior in terms of accumulation of debt that is behind the bubbles we built in the good years and the crisis that resulted from the bursting of those bubbles.

What is always missing in this analysis is the fact that the world is a closed system. The debt that appears in the chart above has to be bought by someone. Those liabilities are assets for someone else. Who are the buyers? And who "forces" them to buy those assets at that price/yield?

Before we continue any further let's rule out the hypothesis that it is the central bank who is buying those assets. The easiest way to understand that it cannot be the central bank is that the chart above starts a lot earlier than the time when central banks' balance sheets started to increase (the second reason is that for every asset that the central bank buys it issues a liability but this will get us to a more complicated argument).

There is a simpler way to explain the chart above. A shift in the supply of saving by some agents/countries resulted in a decrease of interest rates and an increase in borrowing by the rest of the world. Some of this happens within countries, some happens across countries. This could still be an unsustainable development as borrowers go too far and lenders do not understand the risk involved but it is not simply be caused by the irresponsible policies of the central bank.

It could also be that, in addition, we have seen a significant increase in gross flows of assets and liabilities that do not result in a change in equity or net wealth but that they lead to an increase in the size of balance sheets across agents (i.e. increased leverage). Simplest example is households buying real estate with mortgages but it can also be financial institutions increasing leverage. This increases debt but it increases assets as well. This, once again, can generate instability but the borrowing that we see must come from somewhere else in the economy (not the central bank). Understanding that side of the balance sheet is important to have a complete story of what caused the crisis and what it takes to get out of it.

Antonio Fatás

Monday, June 9, 2014

Is liquidity stuck at banks?

Last week the ECB announced new monetary policy actions to help restore growth in the Euro area and bring inflation closer to its 2% target. Interest rates were reduced and further provision of loans to commercial banks were announced. In addition, there is a plan to implement purchases of asset based securities.

The effectiveness of recent monetary policy actions by central banks has been met with some skepticism because it does not deliver the necessary increase in lending to the private sector. While liquidity is introduced, it seems to get stuck in the accounts that the commercial banks hold at the central bank (reserves). Because of this, both the Bank of England and now the ECB are implementing injections of liquidity that are linked to increased lending to the private sector by the financial institutions that are borrowing that liquidity.

The role that reserves play in the recent monetary policy actions by the ECB leads some times to confusion. Some seem to think that the high level of reserves that banks hold is a measure of the failure of central banks to generate additional loans to the private sector. The logic is that reserves stay high because of the lack of willingness to lend. This is the wrong view of reserves, they cannot simply be seen as resources that are waiting to be provided as loans to the private sector.

Reserves are a liability in the central bank balance sheet that it is created when the central bank decides to allocate more loans to commercial banks or when it decides to buy securities. If a commercial bank decides to give a loan to one of its customer (household or business), the reserves do not disappear. Once that customer uses its loan for a purchase, these reserves are transferred from one commercial bank to another, but the level of reserves remains constant.

How can reserves go down? In the case of the ECB, most of the injections in liquidity have been done via loans to commercial banks. Reserves will only go down when commercial banks pay back their loans with the central bank (or when the central bank decides to reduce the amount of loans it provides as some of the outstanding ones are repaid). So any increase in the provision of loans by the ECB will lead to an increase in reserves. Below are the two series: loans to commercial banks (an asset for the ECB) and reserves of commercial banks at the ECB (a liability at the ECB). Both series are in Billions of Euros (Source: ECB).

The evolution of both series is identical. The large increase in loans (LTROs) that started in the Fall of 2011 led to a large increase in reserves. Since then, both series have been coming down as Euro commercial banks have been repaying their loans (voluntarily). So the balance sheet of the ECB has been shrinking dramatically over the last months. The new wave of loans announced by Mario Draghi is likely to increase both series again (although by how much will depend on how Euro financial institutions feel the need to tap into additional ECB funding).

The fact that the two series move together does not mean that the actions of the central bank are ineffective. It is possible that the availability of funding for some banks leads them to provide more loans to the private sector. What one cannot do is judge the success of these actions by the level of reserves in the financial system. The level of reserves will not change when the private loans are given.

As a point of comparison, the profile of the series above for the US Federal Reserve is very different: they keep trending up, no decrease at all. The reason is that the US central bank has increased its balance sheet by buying securities. So reserves are created agains the purchase of those securities. In this case, the level of reserves is even more directly linked to the actions of the central bank. It is only when the central bank decides to sell those securities that the level of reserves will come down. And this is the new step that Draghi has promised in his press conference last week, the ECB is willing to engage in true quantitative easing via the purchase of asset based securities.

Antonio Fatás

Thursday, May 22, 2014

The US labor market is not working.

In a recent post Paul Krugman looks at the dismal performance of US labor markets over the last decade. To make his point, he compares the employment to population ratio for all individuals aged 25-54 for the US and France. The punch line: even the French work harder than the Americans! And this is indeed a new phenomenon, it was not like that 13 years ago [Just to be clear, there are other dimensions where the French are not working as hard: they retire earlier, they take longer vacations,... but the behavior of the 25-54 year old population is indeed a strong indicator of how a society engages its citizens in the labor market. ]

So are the French the exception? Not quite. Among OECD economies, the US stands towards the bottom of the table when it comes to employment to population ratio for this cohort (#24 out of 34 countries).

What is interesting is that most of the countries of the top of the list are countries with a large welfare state and very high taxes (including on labor). So the negative correlation between the welfare state and taxes and the ability to motivate people to work (and create jobs) that some bring back all the time does not seem to be present in the data.

What is interesting is that the US looked much better 13 years ago (see numbers for 2000 below, the US was 10 out of 34).

The US has gone through a major crisis after 2008 with devastating effects on the labor market but so have other countries. In fact, most European countries have done much worse than the US in terms of GDP growth during the last 6 years. In fact, with the exception of Portugal, Greece and Ireland, the US is the country with the worst labor market record for this age group if we compare the 2012 to the 2000 figures.

Antonio Fatás

Monday, May 12, 2014

Groundhog day (ECB).

For the last months the press conferences of Mario Draghi at the ECB have felt very repetitive. The argument has always been the same: inflation is below target and this might be a risk. But there is uncertainty and there are other risks so let's wait for more data. But when more data arrives, confirming that inflation is below the target, there is no action being triggered and we simply start a new period of waiting for yet more data. Here is my quick search for this pattern in the speeches and Q&A from the last eight months press conferences.

October 2013:
... and are ready to consider all available instruments.
November 2013
... but there are a whole range of instruments that we can activate, if needed.
December 2013
... and are ready to consider all available instruments.
January 2014
... and to take further decisive action if required.
February 2014
... and to take further decisive action if required.
March 2014
... and to take further decisive action if required.
April 2014
... and act swiftly if required.
May 2014
... and act swiftly, if required.
So it was back in October when the ECB moved from the (forward guidance) statement of interest rates remaining low for a long period of time to explicitly mentioning the possibility of further actions where all available instruments would be considered. Since then, only the words have changed: notice that in the last two months they are willing to act swiftly while before they were willing to take further decisive actions (in both cases only if required).

It is hard to know how much the wait-and-see attitude of the ECB is a sign of a compromise to acknowledge the threat of low inflation even if there is no consensus on how to deal with it or a truly cautious approach to dealing with challenging economic times. In either case, the actions and even language of the ECB stand in sharp contrast with those of the US Fed.

Antonio Fatás

Thursday, May 8, 2014

The UK makes the Euro area look good.

A quick chart-of-the-day post motivated by some articles I was reading today about differences in country performances during the global financial crisis. Which economic policies worked best? How bad (or good) membership in the Euro area was to fight back the crisis? These are important questions to understand the effectiveness of different economic policies (monetary, fiscal, exchange rate).

When comparing performance across countries it is quite common to use a variety of indicators: GDP growth, unemployment, productivity,... They all tend to move together but they can sometimes provide a quite different view of the economic performance during a number of years. I decided to look at GDP growth but adjusting is by changes in demographics: GDP divided by working-age population (between 15 and 64 years old, as it is measured by the OECD). What I do is to compare the 2013 number with the 2007 number (which I use as the beginning of the crisis). [Click on the chart for a larger image]

What I find interesting (and surprising) is the similarities across countries, despite the differences in policies. With the exception of Greece (and possibly Italy) all the other countries are very close to each other. The three countries that originally opted out of the Euro do not look too different from the Euro countries. Yes, Sweden has done great but so has Germany. The UK has grown less than the Euro area (of 18 countries), less than France or the Netherlands and at a rate which is very similar to that of Spain. Same for Denmark. Among the small countries that are still outside of the Euro area some have done quite well, others not so well and, surprisingly, some of these countries manage to do well with a currency pegged to the Euro (Bulgaria and Latvia).

[Note on data: let me stress that I am using GDP divided by working-age population and this makes a difference for some economies. For example, Latvia's GDP in 2013 is still lower than in 2007 but its working-age population has been declining sharply over these years. Dividing by working-age population allows us to remove potential demographic changes during these years.]

So despite the stubbornness of the ECB and the constraints of a common currency, economic performance in the Euro area has not been too different from those of the other European countries that are outside. This might not really be good news. It might simply be the case that the anti-inflation obsession of the Swedish central bank and the fiscal policy austerity of the UK government have helped to make the Euro-area performance look not too bad.

Antonio Fatás

Tuesday, May 6, 2014

When all asset prices are too damn high.

Increases in stock prices over the last years combined with bond prices that remain high (yields are low) have raised the possibility of mispricing in assets, potential bubbles and future crashes. Are all assets too expensive? Some think so and refer to the current situation as a "gigantic financial asset bubble" where all assets (bonds, stocks, commodities,...) are priced too high. Others see trouble just in stock markets where valuations seem to be growing much faster than the real economy. But there are also those who think the stock market still offers a good return.

Here are two perspectives that can hopefully help understand such diverging views on asset prices:

1. How can it be that all asset prices are overvalued? When all asset prices look too high, we are making a statement about the disappointing returns these assets offer. The key question is whether we are really taking about mispricing or simply about surprisingly low (equilibrium) returns that saving is offered these days? Martin Wolf in today's FT offers many arguments on why low interest rates are here to stay because in a world of abundant saving, returns will be low and asset prices will be very high (I have written about this before). So maybe all asset prices are not too high, it is just that returns are not as high as they used to be.

2. How do you define a bubble? Before answering the question, it is good to get a perspective on the data. Neil Irwin at the New York Times has a great summary of the US stock market in six charts. What do we learn? Stock prices compared to the current level or earnings are high by historical standards. In other words, if you buy the stock market today, you should expect returns that are lower than typical returns. Is this a bubble? Maybe not if those low returns are consistent with the low returns that are offered anywhere else in the economy (back to the argument that "all asset prices are high"). If you do that comparison (see Irwin's article) and calculate the difference between returns one would expect from current stock prices and the returns that bonds offer, the difference is still positive and consistent with historical values (this is the same point that Brad DeLong makes). So stock prices look high but so do every other asset price. Once again, get used to low returns in a world where everyone wants to save.

So does it mean that everything is fine? No, it all depends on what are the expectations of current investors. A bubble in the stock market is not about how high stock prices are or about how low expected returns are. A bubble is about expected returns that are inconsistent with the current stock prices and their relationship to the fundamentals of the economy. If investors are buying stocks today having as a reference the returns that we have witnessed in the last years, then we are in a bubble. But if investors are buying stocks today as an investment that offers a low but consistent return with any other form of saving, then we are fine. From Irwin's article at the New York Times:

"Add it all up, and and it leads you to a conclusion.. Stocks may not be wildly overvalued relative to fundamentals. But for them to rise much from here, a lot of things will have to go just right for investors."

Correct. Stock are not a bargain like they were two years ago (when risk aversion was very high). Their prices are back to levels that are consistent with fundamentals and those fundamentals can deliver returns that are reasonable given other investment opportunities. But if all your fellow investors are hoping for yet another great year in the stock market, then run, because there is no way fundamentals can justify another couple of years of very high returns.

Antonio Fatás

Sunday, May 4, 2014

Refocusing economics education.

Via Mark Thoma I read an interesting article about how the mainstream economics curriculum needs to be revamped (Wren-Lewis also has some nice thoughts on this issue).

I am sympathetic to some of the arguments made in those posts and the need for some serious rethinking of the way economics is taught but I would put the emphasis on slightly different arguments. First, I  am not sure the recent global crisis should be the main reason to change the economics curriculum. Yes, economists failed to predict many aspects of the crisis but my view is that it was not because of the lack of tools or understanding. We have enough models in economics that explain most of the phenomena that caused and propagated the global financial crisis. There are plenty of models where individuals are not rational, where financial markets are driven by bubbles, with multiple equilbria,... that one can use to understand the last decade. We do have all these tools but as economics teachers (and researchers) we need to choose which ones to focus on. And here is where we failed. And we did it before and during the crisis but we also did it earlier. Why aren't we focusing on the right models or methodology? Here is my list of mistakes we do in our teaching, which might also reflect on our research:

#1 Too much theory, not enough emphasis on explaining empirical phenomena. Models are easy to teach. Answering questions like "what caused the 2008 crisis?" or "what are the effects of an increase in the minimum wage"? is so much harder. So not only we tend to avoid them but also criticize those who provide answers by saying that there is too much uncertainty and no one really knows the answer. This is just a bad excuse because policy makers need to make decisions regardless of uncertainty. Economics students should be aware of the uncertainty surrounding these questions but they should also be taught how to answer them.

#2 Too many counterintuitive results. Economists like to teach things that are surprising. Teaching that consumption increases when taxes go down is not too exciting. But teaching that under some very implausible assumptions, consumers will save all the tax rebates to pay for future taxes makes you feel that you added some value in your class. Yes, clearly teaching something that our students have not thought about before can potentially be of more value, but this is only true if what we teach is relevant. If it just introduces confusion and makes our students be cynical about every economic policy proposal then we failed.

#3 The need for a unified theory. The idea that economics is a rigorous science pushes economists to look for consistency via a unified framework when we teach the subject. We want to have one model to explain everything. The use of small and some times inconsistent partial equilibrium models to explain real-world phenomena is seeing as a sign of weakness. A unified theory that is consistent (even if it does not explain much of what we seen in the real world) is always the way to go. Yes, a unified theory would be great, but we need to be realistic. Small ad-hoc models can be a lot more effective to learn about economic issues than the insistence on using the same unrealistic model to explain everything. And in most economics courses we spend all our time building this model and once we are done there is very little time to answer relevant questions. And when asked, we simply argue that "this model cannot capture that" (so back to mistake #1, too much theory, not enough emphasis on understanding empirics).

#4 We teach what our audience wants to hear. We conform too often to social beliefs about how the economy works and we simply support those beliefs with our teaching. Here is one example: when we teach about governments, cracking a few jokes about government inefficiency, bureaucrats, politics is very easy. Using models where government spending plays no productive role feels natural. But when we look at the private sector, we start with the opposite view, one of efficiency and absence of rents given the competitive environment in which firms operate (the famous analogy of no $100 notes sitting on the sidewalk). If you want to argue that it is the other way around be ready to fight a difficult battle. An it is not that we have plenty of empirical evidence to back up these statements. There is very limited research and in some cases with very uncertain results on the role of rents, inequality, market power in modern economies (although this might be changing). But rather than teaching about this uncertainty we start with models that take a very strong stance on these very fundamental questions. So we are being inconsistent. While in some cases we use uncertainty to criticize certain economic policies, in other cases we use the same argument to support a certain view of the world because it matches either the status quo or the beliefs that most in the audience have.

Antonio Fatás