Wednesday, March 3, 2010

Leverage and Financial Intermediation

A bank is a financial intermediary. A financial intermediary is a "middleman," it borrows funds from some and then lends those funds to others. The revenue of the financial intermediary is the difference between the interest rates it charges and it pays. Like other businesses, a financial intermediary has operating costs--facilities, labor, and the like.

The concept of leverage from either personal or business finance applies only loosely to financial intermediation. Compare to another type of middleman, a retail business. Retailing involves the purchase of goods at wholesale and then the resale of those goods at retail. The retailer's revenue is the difference between the prices it charges and the prices it pays. A retailer has operating costs--facilities, labor, and the like. And the retailer must finance its inventory and operations, using either debt or equity.

The financial intermediary, on the other hand, has no need to finance its inventory because its inventory is funds borrowed by the financial intermediary. But given the nature of this business, even if all of the operating costs were financed by equity, the financial intermediary would still be "highly leveraged."

In banking, the more common term to refer to relationship between debt and equity is the capital to asset ratio. This ratio is the net worth of the intermediary divided by its total assets. It shows the proportion of the total assets of the intermediary that are financed by equity, that is, by the owners. Since bank "capital" is the difference between assets and debt, the capital ratio is related to leverage. If that capital to asset ratio is cr, and the leverage ratio (debt/equity) is lr, then the relationship between the leverage ratio and the capital to asset ratio is:

lr = (1/cr) - 1

Before deposit insurance was instituted in the thirties, banks would typically put a very abbreviated balance sheet on their door--Assets, Deposits, and Capital. A typical capital ratio would be 15 percent. So, a bank with $100 million in loans and other assets would fund those with $85 million in deposits. The net worth, or funds provided by the stockholders, would be $15 million. The capital to asset ratio is 15 percent. The typical leverage of banks during the twenties would be (1/.15) - 1 = 5.6.

Why would banks put these balance sheets on their doors? It was to reassure their depositors. If the bank took losses from bad loans, then it would remain solvent and able to pay off the depositors eventually, as long as the losses were less than 15 percent of total assets.

Treating this situation as being similar to an individual investor "leveraging" his bet on the stock market by taking his capital and then borrowing funds from the broker is very confused. Like any other businesses, the issue is how will these value-enhancing activities be funded--with debt or equity. The owners are residual claimants and are the first to bear losses. If the losses are too severe, then part of the losses must be suffered by those providing the debt finance. For the bank in the twenties (or before), that would be depositors. The only thing unusual about banking with regard to leverage was that the nature of the business made it high.

After deposit insurance was introduced, capital ratios soon became an element of regulation. The basic capital requirement was 6 percent, which implies a leverage ratio of 15.6. Why did depositors accept such a low capital ratio? It was because they were insured by FDIC. Depositing money into a more highly leveraged firm because the funds are insured is an example of moral hazard--taking risk because the insurer covers the loss.

The nature of the problem was made obvious when all of the banks scraped the simple balance sheets off their doors and replaced them with what remains there today, a sign that says "FDIC Insured."

Because savings and loans institutions concentrated all of their lending in home mortgages, which were "so safe," regulators allowed them to keep an even lower capital ratio of 4%. That was a leverage ratio of 24. When high expected inflation in the late seventies caused short term interest rates to rise into double digits, the low credit risk of the home mortgages did little to solve the massive interest rate risk created by borrowing through savings accounts and then concentrating lending in 30 year home mortgages.

After the savings and loan fiasco, regulators increased capital ratios. In the U.S., the minimum rose to 8 percent. That would be a leverage ratio of 11.5. However, banks that were "well-capitalized" were allowed more freedom to expand into "other" activities, such as underwriting securities, and that required a capital ratio of 10 percent, in other words, a leverage ratio of 9.

It seems sensible that financial intermediaries making more risky loans should be required to have more capital. Since the point of requiring capital ratios is to protect depositors, or more precisely, deposit insurers, the greater the risk of loss from the loans, the more likely the deposit insurer will have to pay off insured depositors. Of course, the extra low capital requirements for the saving and loan industry didn't work out very well. But that just would suggest that mortgages involve more risk than had been understood, so in the future regulators would require more capital for home mortgages.

Risk based capital requirements then require different levels of capital depending on the risk of the assets. For example, vault cash or balances at the Fed require no capital. Government bonds from a developed country, like Greece, require no capital. Commercial loans to small businesses, on the other hand, are very risky, and require the full 8 percent minimum capital requirement. Incredibly, after the saving and loan mess, home mortgages require 4 percent capital. And mortgaged backed securities rated AA or AAA by S&P, Moody's, or Fitch require only 2 percent capital!

How does bank capital and so, bank leverage, relate to nominal expenditure in the economy? In the absence of monetary disequilibrium, there is no relationship between bank leverage and aggregate nominal expenditures. Of course, bank do some borrowing by the issue of monetary liabilities--transactions accounts these days. And banks hold vault case and reserve deposits at the central bank. Banking and monetary disequilibrium are intimately related. However, does leverage, even "excessive" leverage, generate changes in nominal expenditure independent of monetary disequilibrium? No, it does not.

It is easy to make a false analogy from the individual stock speculator leveraging his capital to a bank expanding its balance sheet. Suppose the banks are holding 15 percent capital as they did before deposit insurance, and so, are leveraged a bit less than 6 to one. Then, they decide to leverage up, to the level of well capitalized banks today, holding only 10% capital and so leveraging 9 to one. Perhaps they can go to the pre-S&L crisis capital ratio of 6 percent and so leverage their capital more than 15 times? Could they go as far as the S&L's before the crises? A 4 percent capital ratio? Leveraging their capital 24 times? If they limited their asset portfolio to AA mortgaged back securities, why not follow the apparent advice of the regulators, keep 2% capital and leverage up 49 times?

Surely, all of that bank lending would result in a credit fueled boom? As the banks increase leverage, they are lending more and more. They might lend to households purchasing new cars or homes. Perhaps they lend to business, financing new buildings or equipment. Perhaps they lend to local government by purchasing bonds to help finance schools or road improvements.

Since nominal expenditure is made up of consumption expenditure by households, investment expenditure by firms, and government expenditure, as banks leverage up, the expansion of credit fuels an expansion of nominal expenditure on many fronts, and the economy booms. Firms can sell more, so they expand production. Profits are strong. Jobs are created.

But there is a limit to this constant expansion of credit through ever growing leverage. If leverage ratios get small enough (a tiny fraction of one percent) then any bad loan will force banks into insolvency.

So, why the bust? When the ever increasing loans slow or stop, spending slows. Firms and households that could easily make loan payments when the economy was booming, have difficulty. When debtors can't pay, the banks have bad loans. This reduces the banks' capital, because losses reduce net worth. Already being over-leveraged, the banks must further reduce lending.

It seems so plausible, but it just doesn't add up. During the supposed credit fueled boom, the banks may expand their lending, but they must also expand their borrowing. In the absence of monetary disequilibrium, those lending to the banks have less money to spend on other things.

For example, a household refrains from going out to eat and thriftily puts those funds in the bank. The bank takes those funds and lends them out to another household that purchases a car. There is more consumption expenditure in the car industry, but less consumption expenditure in the restaurant industry. There is no change in aggregate consumption expenditure.

Consider another example. A profitable firm was going to retain earnings and purchase new equipment to expand output and obtain future profits. However, the firm notes that the expected profit is less than the interest that could be earned by putting money in the bank. The firm lends the money to the bank. The bank then lends the money to another firm that uses the funds to purchase equipment and expand its profit. That firm anticipates that the added profit will allow it to pay back the bank loan and make money for the owners. While investment expenditure on some types of equipment might rise, it falls for other types of equipment. Investment expenditure in aggregate isn't changing.

Of course, the household that refrains from going out to eat might be funding some firm's equipment purchase. Or the firm that retains earnings and puts them in the bank may be funding some household's car purchase. But unless monetary disequilibrium is generated, there is no impact on total nominal expenditure. Credit, even if financial intermediaries are involved and are increasing their leverage, simply shifts funds between and among households and firms.

If this increase in bank leverage is taken as "exogenous," that is, banks and depositors, (or regulators) decide that lower capital ratios are acceptable, then there remains market based limitations to bank expansion. Again, in the absence of monetary disequilibrium, a bank seeking to expand its balance sheet must attract more depositors and more borrowers. This involves paying higher interest on deposits and charging lower interest on loans. The bank's interest margin shrinks. While an expansion in the bank's balance sheet increases revenue, this is partly offset by the decrease in the margin between the interest rates charged and paid, (which is the standard logic of marginal revenue with monopolistic competition,) and then there are additional operating costs from servicing the additional deposits and loans. Even if capital ratios were irrelevant, banks would have no incentive to expand their balance sheets beyond the level that maximizes profit.

To the degree banking, or some elements of it, is perfectly competitive, a "price taking" bank can obtain deposits, make loans, and expand its balance sheet an "infinite" amount. But if the banks in general are reducing capital ratios and raising leverage, the added demand for deposits and supply of loans will reduce the margin between interest rates charged and paid until it equals the marginal cost of operating with larger balance sheets.

Consider the opposite scenario. The banks suffer losses--they have made some bad loans or hold some bonds that default. The banks' capital is reduced. Suppose that the banks need to meet capital regulations or else depositors, used to observing capital ratios on the bank door, will begin to remove funds unless the banks increase their capital ratios. What if the banks must reduce leverage? The simplest approach is for a bank to refrain from making new loans as it collects on existing loans and use the funds to pay off depositors.

Doesn't this contraction in bank credit reduce nominal expenditure? In the absence of monetary disequilibrium, an imbalance between the quantity of money and the demand to hold money, there is no impact on nominal expenditure. While those who would have borrowed from the bank and spend the funds have less money, the depositors whose funds are repaid by the bank have more money to spend.

The possibilities are endless. Perhaps the firm that had retained earnings by putting money in the bank will now invest the funds internally. So, rather than someone borrowing money to buy a car, some business purchases new equipment.

Or, perhaps the bank pays back a firm, and that firm purchases corporate bonds. The business that would have borrowed money from the bank instead issues new corporate bonds. Perhaps depositors, receiving repayment of deposits, purchase stock in banks, allowing banks to expand their capital without decreasing loans.

Since banks have a lower demand for deposits and are reducing their supply of loans, they can lower the interest rate paid on deposits and increase the interest rates charged for loans. This increases the margin between the interest rates paid and charged. While the shrinking balance sheet lowers revenue, the expansion in the interest margin raises revenue. Along with reduced operating cost because fewer deposits and loans must be serviced, operating profits could expand.

If, however, the problem really was past losses reducing bank capital, and some outside constraint is forcing banks to shrink their balance sheets, and the banks were maximizing profit before, then the result will be reduced profit for the existing banks. The solution would be for banks to obtain more equity investment, selling new shares on the market.

Consider the situation after the banks have reduced their balance sheets to return capital ratios (leverage) to what they count as an appropriate level. The banks balance sheets are too low, and expansion of both lending and borrowing would raise profit. If depositors (or regulators) require capital for this profitable expansion, then the banks can and should increase capital.

Rather than just imagining that banks decide to increase or reduce capital ratios, suppose there is a change in supply of deposits or the demand for loans. For example, suppose a new production technology is introduced--single family homes will magically begin generating output in the near future. The demand for loans to purchase the magical houses rises. The banks raise the interest rates they charge on loans, and make more profit. Competition among banks results in them also raising the interest rates paid on deposits to attract more deposits and make more loans. If the banks fail to retain part of their earnings or issue new shares, then capital ratios fall and leverage increases.

Assuming no monetary disequilibrium, there is no impact on aggregate nominal expenditures. Some of those who would have borrowed from banks at lower interest rates are crowded out. Perhaps fewer cars are purchased with borrowed funds. Perhaps fewer machines, buildings or equipment in sectors unrelated to housing are purchased. And, of course, those depositing additional funds in the banking system must be considered. Perhaps they purchase fewer stocks and bonds. Perhaps they go out to eat less often. The end result is that some other element of investment or consumption expenditure is reduced.

Because this is a profitable expansion of the banking industry is due to increased demand for its product, there should be little problem with banks retaining earnings or selling new shares, and maintaining capital ratios. However, if they don't, the notion that this expansion is being caused by increased leverage, or even "excessive" leverage is absurd. It is being caused by the increase demand for bank loans. And the increase in the demand for bank loans was caused by the new technology. Of course, when people discover that there is no magic in the world, some of the home buyers will be disappointed. And banks that made loans secured by homes at prices that reflected the belief in magic will take losses. And the amount of capital that the banks had will turn out to be very important for the depositors, or perhaps, the deposit insurer.

Consider another possible scenario. Rapid productivity growth in China results in rapid growth in incomes. The people expand consumption more slowly, perhaps because they can't believe the good times will last. Some of their saving finds its way into the U.S. banking system. The supply of funds available to the banking system rises. The banks can lower the interest rates they pay on deposits. This increases the banks profits. Competitive banks will expand their balance sheets by lowering the interest rates charged on loans. If the banks don't retain earnings or sell new shares, capital ratios will fall and leverage will rise.

In the absence of monetary disequilibrium, nominal expenditure is not effected. The Chinese export goods to obtain funds to deposit into the banks. Import competing industries sell less because domestic purchasers, say consumers of toys and underwear, purchase Chinese goods. The lower interest rate on deposits may cause some to remove funds from banks. They could use those funds to purchase financial assets like stocks or bonds. Or they could go out to eat more frequently. The banks, expand loans, and the borrowers purchase homes, cars, or machinery and equipment.

With this additional supply of funds, the banking system is more profitable. The banks should have no problem attracting increased equity investment for their expansion. If they fail to do so, the decrease in capital ratios and increase in leverage does imply an increase in the risk that losses from bad loans or investments will result in insolvency and bankruptcy. Still, it wasn't increased bank leverage that caused the expansion in the size of banks' balance sheets.

Does that mean that leverage is irrelevant? Of course not! The relatively safe banking system of the twenties could suffer losses equal to 15 percent of total assets before depositors would take a loss. Government regulators allowing banks to keep 2 percent capital for a portfolio of mortgage backed securities were gambling with the taxpayers money. As long as housing prices kept on rising, mortgage backed securities were perfectly safe. The taxpayers wouldn't have to make good on deposit insurance promises because the banks wouldn't fail. If, on the other hand, housing prices fell 30 percent, then the stockholders of banks investing in those securities lose everything, and deposit insurers will be on the hook to cover the depositors' share of the losses.

Capital ratios and leverage are closely related to bankruptcy. How does default and bankruptcy for financial intermediaries impact the economy? And, more importantly, do capital ratios and bank leverage impact monetary disequilibrium? More later.


  1. Keep 'em coming; these are very stimulating posts. A few comments/questions:

    1. "A financial intermediary is a 'middleman'; it borrows funds from some and then lends those funds to others." A financial intermediary/middleman might simply be a loan broker, arranging for savers to make loans to borrowers. That's almost the position of a mutual fund or hedge fund, which is more of a "pass-through" entity than an independent borrower-and-lender. Its intermediary function is similar to that of a bank, but the implications for "leverage" are quite different (which makes me wonder how important the concept of leverage is).

    2. "Incredibly, after the saving and loan mess, home mortgages require 4 percent capital." The justification for this would be that the Fed has finally [!] learned how to avoid big changes in expectations of inflation, so we can be confident that the experience of the S&L debacle will not be repeated.

    3. "Commercial loans to small businesses, on the other hand, are very risky, and require the full 8 percent minimum capital requirement." If this is the regulation, it seems awfully crude. Small businesses vary enormously in how risky it would be to lend to them.

    4. If the (pre-deposit-insurance) bank thinks it needs vault cash of 10% of deposits, when a new deposit of $100 is made it will lend out only $90, putting $10 in the vault. If it then (exogenously, whatever that means) changes its view and decides 5% of deposits is sufficient vault cash, it will start lending out $95 for every $100 of new deposits; this will (probably?) result in a higher velocity of money--in greater nominal expenditure--*unless the monetary system adjusts* to it.

    But if the monetary system is going to adjust perfectly to everything that happens, *nothing* will affect the course of nominal expenditure--assuming "perfect adjustment" means *steady growth in nominal expenditure*. Of course, people who say greater bank leverage will, *ceteris paribus*, increase nominal expenditure are assuming *imperfect adjustment*--assuming that the quantity of money *remains the same* after the change in bank leverage. (Or, if not the quantity of money, something like that; when people say "ceteris paribus" it is usually not quite clear *just which* other things are to remain the same.)

    Suppose a bank loses capital and has to contract its lending. Does this reduce nominal spending? You point out that the bank may have a choice between curtailing its loans and failing to pay back its depositors, which would have no differential impact on nominal expenditure. But it is irrelevant to compare what the bank is doing with what else it might do; the proper comparison is between what the bank is doing and what it used to do. It used to make loans that it is no longer making; that, *absent an adjustment by the monetary system*, will reduce nominal expenditure. (Again, with an *appropriate monetary adjustment* there will be no effect on nominal expenditure--this is trivially true.)

  2. Bill, I had been meaning to come back and look at your posts. I have a lot of learning to do, and to catch up on these matters.

    With so much happening around the world, I really do need to understand these concepts. Your site, I am sure will help me to gain a better understanding of these matters.

  3. Many of the payroll service companies use the latest payroll software that makes the process of payroll information and payment very easy.

    Payrolling information

  4. I discovered your web page ideal for my needs. It contains useful content. It features very beneficial subjects for many individuals to obtain a lot from it.
    Business Loans

  5. This comment has been removed by the author.

  6. Leverage plays key role, but of course we need to work carefully. I do trading with OctaFX broker and with them, I am able to use 1:500 leverage, it is so great because, I don’t have to fear of over paying, as they got protection for negative balance, so I don’t have to pay anything outside my own investment, it is the thing that makes me really pleased and allows me to work with freedom due to knowing there is no risk at all.