This month's featured article
Two Market Model Sheds Light on Banking Crises
KOBAYASHI KeiichiroSenior Fellow, RIETI
Visiting Professor, Chuo University
Research Director, Canon Institute for Global Studies
A Monetary Model of Banking Crises
The global recession triggered by the financial crisis that broke out in the United States in autumn 2008 reminded us of the degree of the impact that the destabilization of the financial system may have on the real economy. A series of emergency measures implemented by governments across the world have managed to bring a lull in the global economic turmoil. Meanwhile, criteria for assessing those crisis response measures have yet to be established.
Against this backdrop, RIETI Senior Fellow Keiichiro Kobayashi has built a framework (theoretical model) for assessing the efficacy of a series of measures taken in response to the crisis such as fiscal spending, monetary easing, and banking reform. One of the policy implications derived from this model is that banking reform designed to restore bank solvency - i.e. the acceleration of bad asset disposals and capital injections - is the most effective as a response to a crisis. Of course, fiscal spending is effective as a short-term policy. However, it would not be a definitive cure because a government cannot afford to continue such policy forever. The same can be said about monetary easing.
* * *
- What is the background and purpose of the new theoretical model you have proposed in your latest study?
For nearly 10 years, I have conducted research on the collapse of the economic bubble and the long-term economic slump that ensued, taking various approaches and focusing primarily on the real economy. In the mean time, the latest global financial crisis and resulting economic recessions across the world painfully reminded us of the sheer size of the impact of a financial system disruption on the real economy. Asset prices had gone up to levels that were irrationally high compared to the actual state of the economy and, in tandem with the rise of asset prices, the amount of outstanding debts ballooned. Then suddenly came a sharp fall in asset prices and massive amounts of loans became irrecoverable. This sequence of events in the latest crisis is identical to events surrounding the collapse of Japan's bubble economy. Creditor-debtor relationships are more difficult to discern in the present-day crisis because financial instruments have become far more complex and sophisticated over the years. So, that can be cited as a difference between now and then. But essentially, the latest crisis has reconfirmed that a disruptive event on the money side of the economy, as exemplified by bank runs, is what lies behind disruptions in the real economy.
Some people might point out that bank runs - those in their classic form where anxious depositors would line up outside a troubled bank - were hardly observed in the latest crisis, except in the case of Northern Rock Plc., a British bank. But this is because depositors are assured of the recovery of principal up to a designated amount even in the event of a bank failure thanks to the presence of a safety net, namely, enhanced deposit insurance. However, problems did arise in interbank transactions, for which no such safety net exists. With less creditworthy banks increasingly denied access to short term funds, some banks became cash-strapped.
In order to properly assess the existing government policies and propose those for future implementation, it is indispensable to describe those developments in the economy in the form of a theoretical model and analyze the spillover effects of various policies. But as far as I know, models used in preceding studies on the banking system have not explicitly distinguished between money and the real economy.
Such was the case, for instance, with "Bank Runs, Deposit Insurance, and Liquidity" (1983), a work co-authored by Douglas W. Diamond and Philip H. Dybvig, which had a significant influence on studies on the banking system in recent years. In extreme terms, it may as well be said that Diamond and Dybvig treated money no differently than consumer goods. With such a model, it is not possible to analyze what kinds of impacts would be felt on the real economy as a result of the slow or clogged circulation of money, an intermediary agent for the exchange of goods. So, I decided to develop a new model that allows for making a distinction between money and consumer goods.
Analysis using a model with two markets, one for day and one for night
- What are the specific features of your model?
What I am going to say may be too technical, but one reason why it has been difficult to construct a model that treats goods and money differently is that making such a distinction would make the mathematical analysis used to determine market equilibrium too complex. An idea for overcoming this mathematical barrier was provided by "A Unified Framework for Monetary Theory and Policy Analysis" (2005) by Ricardo Lagos and Randall Wright. By introducing two markets, they have made the computation of market equilibrium much simpler.
In my paper, I applied their concept to the analysis of the banking system. Specifically, I introduced the two markets - the day market and the night market - to describe the relationship between the banking system and the real economy. As I mentioned earlier, disruptions in the interbank money market were the primary factor behind the global financial crisis. But for the sake of simplicity, I have substituted the interbank relationship with the bank-depositor relationship and constructed a model describing a typical bank-run episode.
- What role does each market play?
Without getting into too much detail, the role of each market is as follows (figure 1).
First, for the day market, it is assumed that goods can only be purchased with cash. More specifically, a sale transaction of goods takes place through a random encounter between a buyer and a seller. All transactions are settled via cash payment because it is premised that the buyer and seller are strangers who have had no interaction prior to the transaction. A person wishing to buy goods withdraws money from his bank account and gives the money to the seller in exchange for the goods. The seller then deposits the money into his bank account as long as he has confidence in the banking system.
Repetition of this process in the day market would generate sales transactions of goods in an aggregate value worth many times the amount of cash held in banks. An increase in sales transactions of goods leads to an increase in production. It also leads to an improvement in the solvency position of banks because, strictly speaking, the rate of return on deposits is higher than the rate of return on cash.
Meanwhile, the night market is for the settlement of the day's transactions, including short-term interbank lending and borrowing as well as the settlement of loans to households. Through such settlements, banks prepare cash for withdrawals in the next day's day market. A delay in loan repayments and/or a failure to ensure the smooth interbank transfer of funds in the night market would cause liquidity constraints in the day market and, in due time, the episode would come to the attention of depositors. In this manner, any trouble that has occurred in the night market causes disruptions in sales transactions of goods in the day market, an activity based on the premise of sustaining confidence in the banking system.
Credit uncertainty disrupts the flow of money, causing adverse effects on the real economy
- What form does a financial crisis take in your model?
One of the simplest examples is a case in which participants in the day market become increasingly uncertain about the future and suddenly fall into a panic. Under normal conditions, they would not try to keep more cash on hand than necessary for the payment of goods purchased, knowing that they can withdraw their bank deposits at any time when the need arises. Likewise, sellers would not keep excess cash on hand, depositing the money received in exchange of goods in their bank accounts.
However, when participants in the day market become uncertain about whether or not they will be able to freely access their money when necessary, panic sets in and they try to withdraw all of their money from their bank accounts. But banks do not have the cash reserves on hand to cope with such massive withdrawals because all but a fractional portion of deposits are loaned out or invested in other forms. Furthermore, in such a situation, sellers of goods are also anxious and would not put the money received into banks (they would instead hoard the cash in what is referred to as "tansu yokin", which is the equivalent of cash under the mattress). As a result, banks would quickly run out of cash and some market participants become unable to withdraw cash necessary for the purchase of goods. When concerns become a reality with the occurrence of an incident in which only some people can withdraw deposits, depositors are then thrown into a greater panic. Furthermore, the scarcity of money in circulation stagnates sales transactions of goods and slackens production activities, altogether adversely affecting the economy.
Of course, in such an event, the central bank would increase the money supply. However, inevitably there is a time lag before the central bank takes action, during which transactions stagnate. And if deposit withdrawals do not stop, additional money supplied by the central bank would soon dry up. This process would continue till market participants withdraw all of their deposits or public confidence in banks is restored.
Put into terms more relevant to what has happened in recent years, it can be rephrased as follows: when interbank loans extended in the night market turn sour or so feared, participants in the day market begin withdrawing their bank deposits. The primary cause of interbank loans going bad is a decline in asset prices. At the same time, however, distressed goods transactions have a negative impact on asset prices. This, in turn, would induce more deposit withdrawals by making people expect more loans to turn sour as a definite or likely course of events.
The result is a greater degree of scarcity of money and further stagnation in goods transactions.
Both monetary and fiscal policies are nothing more than stopgap measures
- In your model, how do you define government response measures to the financial crisis?
The direct factor behind the stagnated goods transactions in the day market is the scarcity of money or the means of payment. Emergency measures can be implemented in two ways.
One way is by means of monetary policy. A central bank could provide money to banks to assure depositors that they can withdraw their deposits at any time. However, in order to be able to properly support goods transactions in the day market, the central bank must provide money to all banks, i.e. not only sound banks but also troubled ones. Furthermore, so far as the problem of bad loans, which is the primary cause of bank runs, remains unsolved, the central bank needs to continue to provide money to banks. And this could go on forever.
The use of monetary policy as an emergency measure is necessary. However, as evident from Japan's experience with its financial crisis in the late 1990s, the removal of bad loans from banks' balance sheets and capital injections into banks - both designed to restore confidence in the banking system - are imperative to overcoming a crisis.
The other way is by means of fiscal policy. That is, a government, acting in place of market participants, could purchase goods with fiscal funds to give money to sellers. But so far as concerns over the banking system remain unsolved, money would not return to banks. Thus, the government needs to retain the goods purchased and keep on spending additional funds to purchase more. As is the case with monetary policy, this could go on forever.
From these, we can clearly see that banking sector reform designed to restore bank solvency, i.e. bad loan disposals and capital injections, is the key to the recovery of trade in goods and an increase in asset prices. The cost of these policies may look enormous because of the size of public funds injected into banks for recapitalization. However, the ultimate cost would be limited because banks are to return the public funds once the banking system stabilizes.
A pitfall of U.S. optimism
- How do you assess government responses to the financial crisis around the world?
In the U.S., which has been the epicenter of the global crisis, the Federal Reserve Board acted quickly and decisively in easing the monetary policy. Given the promptness of action and the degree of easing, it seems that lessons have been learned from Japan's experience with its past crisis. I believe that such massive monetary easing has been surely necessary as an emergency measure.
According to local media reports, it appears that optimism is already spreading in the U.S. reflecting the bottoming out of house prices and some other bright signs. If such optimism helps restore confidence in the banking system, sales of goods may increase which in turn leads to an economic recovery and - through rising asset prices - to greater confidence in banks, thus creating a virtuous cycle, as shown in my model.
But my personal view is not as optimistic as that. It certainly looks like the worst is over for house prices. However, commercial real estate prices are still in the doldrums. And if this puts downward pressure on overall asset prices, people's anxiety over the stability of the banking system, which has somewhat subsided for now, may be reignited. It is also a concern that the high level of unemployment, above 10% at the moment, may dampen consumption, driving the real economy down and thus generating further bad loans, which would be nothing but the duplication of what Japan underwent in the 1990s.
It seems to me that the U.S. has yet to take aggressive steps to accelerate the disposal of bad assets such as those undertaken in Japan from the late 1990s through early 2000s. I believe it is necessary for the U.S. to make steady progress on the disposal of bad assets.
Meanwhile, looking to the future course of the global economy, we can no longer rely on the U.S. economy, or consumption demand in the U.S., as the sole driver of the world economy. We need to establish a balanced global economy through the expansion with all countries - including emerging economies - working to expand domestic demand. At the moment, however, even China, a country that is counted on as the greatest driving force for the recovery of the world economy, seems to be unable to cast off the idea that exports are the No. 1 growth engine. I am afraid that the economy, in which each country counts on demand-boosting measures implemented by other countries for its recovery, must be judged to be fundamentally weak. We must discuss the financial system, international trade, and measures to stabilize and revitalize the real economy not only within the league of developed economies but also under the framework of Group of 20 (G20) countries. Otherwise, we may see the spread of beggar-thy-neighbor policies as was the case during the Great Depression in the 1930s.
- What are the plans for your research activities?
The model I have proposed this time describes a financial crisis in terms of the relationship between banks and depositors. However, as discussed earlier, what we see in reality is a phenomenon in which a financial crisis undermines confidence between banks, leading to disruptions in interbank money flows. So, I am studying ways to express such interbank relationships in a simple model.
Also, I am trying to develop a framework for a model that can analyze a financial crisis in combination with a business cycle. What are the differences between an ordinary business cycle and a financial crisis? Because of such differences, should we apply different criteria in assessing government policies? I would like to make a model that can analyze these aspects.
Fellow titles and links in the text are as of the date of publication.
For questions or comments regarding RIETI Report, please contact the editor.*If the "Send by mailer" button does not work, please copy the address into your email "send to" field and connect the prefix and the suffix of the address with an "@", sending it normally.
RIETI Report is published bi-weekly.