The Non-Inflationary Great Leveraging

0:00So I would like to thank the organizer of the conference, in particular Professor Salerno, for giving me the opportunity to present my work today, even if I am a central banker, so it's relatively rare for having central bankers here, I think, but so the paper I present is entitled The Non-Inflationary Grade Leveraging, and it is a joint work with Maya Ganarin, and the usual disclaimer applies, the view expressed in this paper do not necessarily reflect the views of the Swiss National Bank. So the macroeconomic development in many OECD countries over the last 30 years provides an apparent conflicting assessment of the monetary policy stance.

0:49We can illustrate this apparent conflicting assessment with US data, but in that respect, the macroeconomic development in the US is very representative of any OECD countries. So on the one hand, you see on the chart of the left that inflation has been falling relatively low over the last 30 years. And this suggests that monetary policy has been restrictive over the last decades. And there is indeed a large literature supportive of the view that typically that monetary policy, for instance, violated the so-called Taylor Principle before the 1980s, but followed the Taylor Principle thereafter. Then on the other hand, if you look at the chart on the right, these show the US debt as percent of GDP.

1:39And these illustrations show that the financial system has been extremely expansionary over the last 30 years. And it is not obvious to understand why this huge increase in leverage has been accompanied by relatively low level of inflation. Because one may indeed expect the granting of credit to enhance aggregate demands, and therefore to exert inflationary pressure on the consumer price inflation. And this raises a question, as put by Clarida, as to why all this leverage and the rise in aggregate demand it supported did not result in inflation or, at minimum, a Rise in Inflation Expectation.

2:25So the aim of this paper is to provide the solutions to this dilemma. And basically, we just, the hypothesis of this paper just sets on quantities or positions on banks' balance sheets to explain why inflation has been so low. And to put it as simple as it is, inflation has been low because the quantity of money, as money should be properly measured has been low over the last decades. So in the conduct of monetary policy, the central bank orchestrates the development of the banking system. Now, the specificity of the fractional reserve banking system consists in the fact that it allows banks to grant credit simply by creating money.

3:12So when non-banks or capital markets grant credit, They simply lend money that someone has decided to save. So non-banks or capital markets are just true intermediaries between savers and borrowers. By contrast, when banks want credit, they create the money that they lend. So money and credit are two sides of the same coin, because banks supply both loans and deposits. So the question is, whether it is the asset side, that or the liability side, money of banks balance sheet that matters for the determination of aggregate demands and inflation. And this is an old question because the evolution of the banking system over the last centuries created difficulties for the exponent of the quantity theory to adjust their theory to the new mechanism of exchange and to specify the money stock variable which is so essential to this theory.

4:12For example, David Leidler writes that in the middle of the 19th century, John Stradmill could not make up his mind whether it was bank credit per se or its counterpart liabilities which affected prices. And still today there is a lot of confusion in the economic professions to know whether it is money or credit which is relevant for the determination of aggregate demand. But this confusion between money and credit can be easily explained because in the fractional reserve system you may expect money and credit to expand in the same proportion. However, since the 1980s, monetary aggregates have expanded much slower than their credit counterparts. You have here on this chart on the left the data for Switzerland, on the right for the US.

5:02The lower panel shows CPI inflation. And on the upper panel, you see in red the bank law as loans to GDP, so the leverage ratio of the economy. Then in the blue, it's the monetary aggregate M2 to GDP, and in gray, M3 to GDP. And what is very interesting to see a part of the huge increase in the leverage is the fact that until the 1980s, money and credit were highly collinear. But since the 1980s, 1918, 1985 in the US, monetary aggregates have expanded much slower than their credit counterparts. And I will try to convince you that these explain why inflation has been so low over the last decades.

5:50So in this paper, we discuss the relevance of the money view and the credit view as regards the way in which the expansion of the banking system is transmitted to aggregate demand and to inflation. And in particular, we emphasize the theoretical weaknesses of the credit view as we got the development of aggregate demand. And we derive the theoretical consequences of the decoupling of money and credit. And then from an empirical point of view, we exploit the decoupling of money and credit to provide empirical evidence in favor of the money view. So there are two views, two competing views, is one based on money, the other on credit, as regards the way in which the expansion of the banking system is transmitted to aggregate demand and to inflation.

6:40So according to the money view, it is the creation of money by the banking system that is responsible for stimulating aggregate demand. And the expansion of the banking system stimulates aggregate demand because the incremental spending that the banking system finances are not compensated by an equivalent reduction in spending because banks create the money that they lend. And then, when the borrowers spend the deposit that have been created by the banking system, these sets in motion a chain of transactions which stimulates aggregate demand as long as deposit circulates, that means as long as money circulates in the economy, and as long as the price level has not entirely adjusted to the new quantity of money.

7:29Now, the second view is the credit view, according to which it is the granting of credit by banks that stimulates aggregate demand because many economic agents cannot substitute non-monetary or non-banks credit or capital market credit for bank credits in an environment characterized by imperfect information. So, because banks have a monitoring advantage, bank loans are special in the sense that there is no close substitute for bank loans on capital markets. So, for example, if there is a reduction in bank credit, this may lead to a decrease in aggregate demand because many economic agents, like as small enterprises or households, Now, which view best describe the transmission mechanism of the expansion of the banking system on aggregate demand and inflation?

8:21So, at first glance, both view seems plausible. However, it is important to stress two related weaknesses of the credit view as regard to the development or the determination of aggregate demand. At first glance, both views seem plausible. However, it is important to stress two related weaknesses of the credit view as regard to the development or the determination of aggregate demand. First, the credit view completely overlooks the very specificity of bank credit, namely the fact that banks create the mean of exchange when they grant credits. To illustrate the fact that exponent of credit view completely overlooks that, we can quote Blinder and Stiglitz They write, if information were perfect or cheaply acquired, then a reduction in bank credit would be offset by an increase in non-bank credit.

9:08Central bank policy would change the locus of borrowing but would change neither the total volume of credit nor who gets it. Though this statement is clearly wrong because it is equivalent to claiming that the quantity of credit granted in the fractional reserve system would be the same as the quantity of credit granted in the 100% reserve requirement system Because non-banks cannot create money exactly as banks would not be allowed to create money in the 100% reserve requirement system. So this observation does not deny the monitoring advantage of banks, but it just shows that this credit view does not or completely ignores the very specificity of bank credit. Then the second related weakness comes from the fact that one unit of credit reflect the is the occurrence of one transaction, which occur once credit is granted.

9:58By contrast, one unit of money may reflect many transactions because money circulates. So the fact that money circulates and may finance many transactions suggests that the quantity of money must be a much better indicator or much more effective at influencing the aggregate demand and inflation than the quantity of credit. To put it in another sense, it is the increase in credit which reflects the occurrence of transactions but it is the stock of money times its velocity of circulations which reflects the quantity of potential transaction in the economy. So now how can we understand the decoupling of monetary and credit aggregates? So when banks grant credits they create at the same time deposits. Now, the decoupling reflects the substitution of non-monetary banks' papers, non-monetary liabilities for bank deposits.

10:58To borrow an expression from central bankings, when commercial banks issue non-monetary papers, they sterilise the quantity of deposits in circulation of the economy. And non-monetary papers are papers on the liability side of banks' balance sheets, but which are not used as common medium of exchange. That are money market papers, certificate of deposits, structure notes, and bonds, and so on. So the decoupling reflects the substitution of non-monetary papers for money. And these reflect also the willingness of deposit holder not to spend further their deposits, but to save the deposits that have been created by the banking system. We can illustrate this simple mechanism with standardized balance sheet.

11:46You can in a deposit-based banking system, you may expect deposit and credit to evolve in the same proportion. For example, the banking system pledged a paper issue by the household at the central bank against a cash payment and then the banking system create money and loans and deposits. So in the first case, loans are just funded by money, which can circulate. Then if banks start to issue non-monetary papers, they just sell these non-monetary papers against deposits, and this leads to a situation where loans are funded not by money, but by non-monetary papers like long-term debt. What are the implications for inflation of these mechanisms? So the substitution of non-monetary papers for deposits reflect the interruption of the chain of transactions initiated with the spending of created deposits.

12:39So when borrowers spend the deposit created by the banking system, these set in motion of chain of transactions. Now when deposit holders substitute non-monetary papers for their deposits, these reflect their willingness not to spend their deposit further and these reflect an interruption of the chain of transaction. So, according to the credit view, these transformations should have no impact on the development of aggregate demand and inflation because it is the granting of credit by banks which is supposed to stimulate aggregate demand. By contrast, according to the money view, the decoupling should exert a dampening effect on inflation because these reflect an interruption of the chain of transaction. The increasing use of non-monetary papers also facilitates the leverage of the economy because by substituting non-monetary papers for deposits, it relaxes the minimum reserve requirement of banks.

13:40So if there is less deposit in the economy, banks can expand their balance sheet. And also, if you introduce in this game some non-banks, financial intermediaries, which would issue non-monetary papers and then by packages of loans of the banks, this would unload the balance sheet of banks and relaxes the capital requirement of banks. So these may also not only have an impact on the development of inflation, but also it facilitates the leveraging of the economy. Now I turn to the quickly empirical part of the paper. So empirical analysis were relatively difficult to perform in the previous decades because money and credit evolved hand in hand in the past. So, for example, blind and stiglitz said, while we have two competing theories, one based on credit, the other on money, that are conceptually distinct, the data will have difficulty distinguishing them between them because credit and money normally are highly collinear.

14:36Now, we ask collinearity between money and credit where the fact when blind or stiglitz or Benjamin Friedman wrote their paper in the 80s, the decoupling we do have now since 1980s allows to test these hypotheses clearly. So the long-term relationship between money and inflation was first highlighted by Milton Friedman in the 60s. Then more recently, authors pretended that this relationship completely disappeared. However, I don't want to go too much into details, but it is possible if you adjust the quantity of money in the right way, if you in fact adjust the quantity of money by the growth rate of potential GDP and by the changes in the opportunity cost of money-holding, it is possible to find a long-term, still today, a long-term relationship between money and credit.

15:29So I don't have time to go too much into details, but if you just apply this simple adjustment to the money quantities, you find a very strong relationship between the CPI level and the quantity of prices. So here in this chart again you have data for Switzerland on the left, data for the U.S. on the right. The thin blue line is first the log of U.S. loans, then the dotted red line is a log of U.S. M2 aggregates, and then once you adjust you have the decline blue for credit and red for M2 adjusters. And there is a long-term relationship between the CPI level and the quantity of M2 adjusted according to the formulas I just show you here.

16:19So this clearly shows that it is not the quantity of credit that drives the aggregate demand and inflation in the economy, but it is the quantity of money. And we can also perform some different tests to show the long-term relationship. Now, what are the implications for monetary policy of these facts? So in this paper, we show that the expansion of the fractional reserve banking system does not necessarily lead to CPI inflation. So as I highlighted in my presentation, the decoupling of monetary and credit aggregate since the 1980s reflects an interruption of the chain of transaction initiated with the spending created by deposits.

17:07So now, since central banks orchestrate the development of the banking system, which objective should then central banks target? So as long as money and credit evolve hand in hand, monetary or inflation targeting implicitly prevents an excessive expansion of credit. Because if you target inflation, you also target monetary aggregate. And if credit expands in the same path as credit, you also prevent an expansion of credit. Nevertheless, the expansion or the decoupling of monetary and credit aggregates challenges the paradigm of price level stability because the focus on inflation may make policymakers overlook an excessive expansion of credit.

17:57And the intrinsic instability of credit source boom, for which the current crisis gives a prime illustration, was first theoretically exposed by Ludwig von Mises and then there was also, there after an extensive literature showing that credit, expansion of credit usually always precede big economic troubles. So there are strong presumptions that the paradigm of price-level stability is mainly responsible for the blow-up of financial and economic instability over the last 30 years. And to conclude, I would like to draw a parallel between the current situation and the critique of the currency school made by Ludwig von Mises.

18:46So whereas the currency school overlooked that the expansion of credit can result not Not only from the issue of bank notes, but also from the issue of bank deposits, the monetary school or the exponent of inflation targeting overlook that the expansion of credit can result not only from the issue of bank deposits, but also from the issue of non-monetary, non-inflationary papers. So thank you very much.