Lectures in Macroeconomics
Chapter 12: Monetary Policy and Commercial Banking
Commercial Banking in the US
Theoretical Model of a Banking System
The Federal Reserve and US Monetary Policy
Application: Recent Changes in Reserve Requirements
Monetary Policy in Other Countries
Regulating Financial Institutions
Further Web Links and Readings
IntroductionTo this point we've been a little cavalier about money and financial markets, ignoring the distinctions between between currency and monetary aggregates (M in our theories) and the roles played by financial institutions in channeling saving to firms and governments, domestic and foreign. In the next two classes we'll rectify some of these oversights, and take a closer look at banking, financial intermediation more generally, and monetary policy in the US and around the world. With apologies to Goldman Sachs, the word "bank" will generally be used to mean commercial bank in this Chapter of the notes.
Institution Assets Percent Commercial banks 3442.5 29.1 Thrift institutions 1414.1 11.9 Insurance companies 2096.9 17.7 Pension funds 2729.7 23.0 Finance Companies 812.4 6.9 Mutual Funds 1352.0 11.4 Total 11847.6 100.0[The numbers are billions of dollars, from the Federal Reserve's Flow of Funds Accounts Financial Assets and Liabilities Year-End, Z.1.]
These figures should make it clear that banks are not, by a long shot, the only financial institutions in the US. In fact, their market share has fallen from about 39 percent in 1970 to 29 percent in the numbers above. Thrifts have done even worse, falling from 20 to 12 percent over the same period (and for obvious reasons). The big gainers have been pension funds and mutual funds.
While many financial trends are global, there are nonetheless substantial cross-country differences in financial institutions. The most obvious of these concern banks. The US banking system differs from many countries both in the range of services supplied by commercial banks and in the sheer number of them---roughly 12,000 at last count. This compares with 10 in Canada, of which 4 or 5 have almost all of the business. Or about a hundred in the UK, where 6-8 banks control about 80 percent of the market. We could make similar statements about France, Germany, and Japan: they have nowhere near as many banks as the US. [Again: I am referring here to commercial banks or their equivalent in other countries, not thrift institutions, investment banks, or anything else that sounds like a bank.] These banks differ, as well, in their range of activities. US banks have traditionally accepted deposits from customers (individuals and businesses) and used the proceeds to finance loans to businesses and individuals and investments in corporate and government securities. This is often allied with related businesses like credit cards, foreign currency transactions, and so on. But until recently the reach of commercial banks did not extend to investment banking activities, like underwriting. In many countries, however, banks provide a more complete range of financial services. In Germany, for example, "universal" banks provide investment banking and insurance services.
In the US, the banking and insurance industries have been separately regulated almost from the start, and commercial and investment banking were formally torn apart by the Glass-Steagall Act of 1933---hence the separate identities now of Morgan Stanley and JP Morgan. In recent years, the Act has been weakened, and many of the leading commercial banks now offer some range of investment banking services through so-called Section 20 subs. Bankers Trust has moved so far along this path that while it's still officially a commercial bank, it no longer has a retail business. Conversely, Merrill Lynch and other retail brokers now offer money market accounts that serve much the same purpose as checking accounts at commercial banks. In short, changes in regulations have had an enormous impact on the financial services industry in general, and commercial banking in particular, over the last twenty years or so.
Along with the reduction in the market share of commercial banking, we have seen, and will likely continue to see, a sharp reduction in the number of banks. The huge number of commercial banks in the US reflects a long-standing political impasse. On the one hand, Alexander Hamilton, resting in the Trinity Church cemetery near the head of Wall Street, argued that a strong banking system was essential to the economic viability of the country. On the other, Thomas Jefferson and his Virginia compatriots thought that large banks would place too much power in the hands of a few individuals. The result was a compromise that placed much of the oversight and control of banks in the hands of states. Thus we have, in some ways, not a single national set of banking legislation but fifty different sets. This has come up in talks with the European Community concerning reciprocal foreign banking regulations: the Europeans complain that access to US markets requires the approval of individual states as well as the federal government. As the regulations supporting the excessive segmentation of the banking industry are relaxed, we can expect to see a steady decline in the number of banks.
Currency 244.8 Checkable deposits 569.4 Travellers checks 8.4 M1 (total) 822.6[These numbers are billions of dollars; see Table A12 of the Federal Reserve Bulletin.]
[Note that we have lots of currency: about $1000 for every man, woman, and child in the US. Who has all this? And why?]
In this aggregate and others, "deposits" includes money at institutions other than commercial banks: savings and loans, credit unions, brokerage firms, and so on. This wasn't always the case, but since these institutions now have deposits that are like bank accounts in everything but name, they are now included in the data and are treated the same by regulators (reserve requirements, for example).
M1 is the aggregate that conforms most closely with our macroeconomic theory. We argued that "money" was useful for transactions, but paid no interest. [That led to the money demand relation, L(i,Y), that underlies the LM curve.] But as we often find, the world is more complicated than our theory. Particularly during the 1970s and 80s, the line dividing accounts used for transactions and those that were pure investments became extremely fuzzy. Banks, in competing for deposits, offered accounts that paid interest like time deposits yet had some of the check writing capabilities of checking accounts. In fact only a small part of the so-called checkable deposits above consists of pure demand deposits. They also offered, especially to large business customers, demand deposits that could be converted to time deposits or other interest-bearing asset at the end of the day to get higher interest.
As a result, economists who studied this found that M1 seemed to vary erratically over time as demand deposits were relabeled as time deposits and so on, and that the definition of M1 itself was becoming highly arbitrary. In the last decade, emphasis has switched instead to broader definitions of money that include both demand and time deposits and are thus less prone to variation as names and features of deposits change. The disadvantage of the broader definitions is that we are no longer talking about assets held purely for transactions purposes. In any case, the numbers for M2 and M3 are:
M2 3316.4 M3 4085.8M2 contains, in addition to M1, overnight RPs and Eurodollars, money market accounts, and small denomination time and saving accounts. M3 contains, in addition to M2, large denomination time deposits and Eurodollar accounts, and some other things. [See the notes to Table A13 of the Federal Reserve Bulletin for details.] You can see that the major difference is between M1 and M2: the latter is about 4 times as big as the former. The Federal Reserve currently emphasizes M2 in its policy statements. Our graphs of "money" in earlier lectures were also M2.
It's clear, then, that most of what we call "money" consists of bank deposits and is therefore not under the direct control of the Fed. To see how the Fed might influence M2 indirectly, we need to think a little about how the banking system interacts with monetary policy. That's what we do next.
The objective of this section is to provide a link between the money between the monetary aggregates used in our theory (think of this as M2) and the part of "money" that is under the direct control of the Federal Reserve (which we call the monetary base, MB). We try to spell out the link between Fed policy and monetary aggregates, and the role of the banking system in this process.
A bankless economy.To get ourselves warmed up, as it were, let's look at the balance sheets of the Fed and the Private Sector in a stylized economy that has no banking system, and the effect on these balance sheets of an open market operation. Then we'll go on to see how a banking system changes the analysis. Let us say, then, that the Private Sector (excluding banks) has, among other assets, 500 of treasury bills, 100 of currency, and some equity. Its balance sheet might then be something like
Private Sector Balance Sheet Assets Liabilities and Net Worth Currency 100 Net worth 8600 Treasury bills 500 Equity 8000[In real life, this would be much more complicated, but since this is theory we can go easy on ourselves.]
The Fed might have, say, an inventory of 100 in treasury bills and a liability of the same 100 in currency, since currency in the US is Federal Reserve Notes: in effect, interest free loans from the public to the Fed. (Read one sometime to see for yourself.) Thus the Fed's balance sheet is
Federal Reserve Balance Sheet Assets Liabilities Treas bills 100 Currency 100(The convention is that the Fed has no net worth: earnings accrue to the Treasury.)
In this economy, like the one I had in mind when we talked about the Keynesian model, the money supply is the supply of currency: 100. We can change this with an open market operation. If the Fed wants to increase the money supply by 10, it simply buys 10 worth of treasury bills from the public. [Work through this on the balance sheets for practice.] This changes the composition of the balance sheets of both the public sector and the Fed, but not their net worths. That's what was going on behind the scenes in our discussion of monetary policy in the Keynesian model: an increase in the money supply made the composition of the private sector balance sheets more liquid, in the sense that it included more money after the open market purchase than before.
A banking system.That was practice, now we develop the same idea for an economy with a banking system. We add bank deposits (and the corresponding loans) to the private sector's balance sheet and bring banks into the picture. A possible configuration is:
Private Sector Balance Sheet Assets Liabilities and Net Worth Currency 50 Bank Loans 150 Bank Deposits 200 Net worth 8600 Treasury bills 500 Equity 8000
Federal Reserve Balance Sheet Assets Liabilities Treas bills 100 Currency 50 Reserves 50
Commercial Banks' Balance Sheet Assets Liabilities Reserves 50 Deposits 200 Loans 150You'll note that net worth is zero for the Fed (it's "owned" by the Treasury) and Commercial Banks (they're owned by shareholders).
A useful example of a monetary aggregate in this economy is M = CU (Currency) + D (Bank Deposits). [This is simpler than we saw in the real world, since we only have one type of deposit. With more than one type of deposit we have more than one type of money and a more complicated theoretical setup.] The Fed, on the other hand, controls the amount of currency held by the private sector (as cash) and banks (as reserves). We call this quantity the monetary base, MB = CU + RE (Reserves).
The question is how an open market operation that changes the monetary base MB influences the monetary aggregate M---whether, that is, we can talk about the Fed influencing a monetary aggregate, when policy involves the narrower monetary base. We can derive the relation between the monetary base MB and the monetary aggregate M if we make some assumptions about behavior. Let us say, first, that private agents like to hold cash and bank deposits in some strict proportion:
CU/D = g ,where g is some number that we might expect to be roughly constant. The idea is that we make some transactions with cash, others with checks, and the proportions of the two doesn't change much. Let us also assume that banks hold a constant fraction of their deposits as reserves:
RE/D = r .This latter assumption is pretty good, since the Fed requires them to hold reserves proportional to their deposits (we'll see the details shortly). From a bank's point of view this acts as a tax on their deposits, since reserves earn no interest. Even if there were no minimum reserves, banks might be expected to hold some fraction of deposits in cash as part of their day to day business. From this, we can derive a relation between the monetary aggregate and the monetary base. We know:
MB = RE + CU (equilibrium condition) M = CU + D (definition of money)This leads (after some relatively simple algebra) to
M = [ (1+g) / (g+r) ] MB .The expression in brackets is referred to as the money multiplier, since we generally see that the stock of money is a multiple of the monetary base. In the US, for example, the multiple is about 3 for M1 and over 10 for M2 and M3.
We now have an answer to our question: if the ratios r and g are approximately constant, then by controlling the monetary base the Fed exerts indirect control over the broader monetary aggregates. In that sense, we can speak loosely about the Fed "controlling" M2 and other aggregates.
But are the ratios constant? We can get some idea by plotting the data.
In Figure 1 and Figure 2 we see
how monetary aggregates and related variables have behaved over time. In
Figure 1 I've graphed MB, M1, and M2 for the last thirty
years (each is scaled to equal 0.0 in the first quarter of 1959).
The trends are somewhat different, with M2 and M3 growing faster
than MB and M1. In Figure 2 we see the money multipliers
for the three aggregates.
Again there has been some variation over time (as there must be
since the aggregates have grown at different rates). You can see the same
thing in different form Figure 3, where the growth
rates of MB and M2 are drawn.
In short, the money multipliers are another case of a reasonable approximation, but in the short run we see some variation which is reflected in different growth rates across aggregates. Thus the money multiplier theory is only a rough guide and in the short run, at least, the Fed may have a difficult time affecting monetary aggregates.
Application: Money in the DepressionOne of the many unusual events of the 1930s is that the stock of money (think of this as M2) actually fell by 35 percent between March 1930 and March 1933. Some economists (notably Milton Friedman and Anna Schwartz) have argued that this decline was one the major factors in the Depression, and point to a 30 percent decline in the price level (deflation). Interestingly, while the stock of money fell, the monetary base rose by about 20 percent.
What happened? Clearly the money multiplier fell, but why? Two reasons stick out. (i) There was a great deal of uncertainty about the health of the banking system. One of the consequences was a sharp increase in the currency-deposit ratio as people pulled their money out of banks. (ii) Banks held large excess reserves, in anticipation of runs, and the Fed (in one of the bone-head moves of all time) increased reserve requirements to match. Thus both g and r rose and this led, as in our theory, to a sharp decline in the money multiplier.
Problems with the banking system in the 1930s led to changes in banking legislation that are still important today: Glass-Steagall, deposit insurance, and so on. Reports from the 1930s sound, in some ways, much like the late 1980s.
Application: Contrary Movements in Monetary AggregatesOver the last few years we've seen, as we did in the Depression, a divergence between the movements in the monetary base (MB) and monetary aggregates (like M2), with the base growing more rapidly than the aggregates. See Figure 3. Apparently the increase in the monetary base has been offset by declines in the money multipliers.
The story has some similarity to the Depression. For whatever reasons,
there has been, since 1986, a sharp rise in the ratio of currency to deposits
(this includes all the deposits counted in M2); see Figure
This implies, as we've seen, a fall in the M2 multiplier. Thus M2 over this period has grown less rapidly than the base. But why? Three possibilities cross my mind, maybe you can think of others: (i) Lack of confidence in the banking system led people to put less of their wealth in banks. Given deposit insurance this is probably a misplaced concern, but maybe it affected peoples' behavior. (ii) Growth in the underground economy (drugs?) led people to use more cash than before. (iii) Banks made less effort than before to attract deposits, since they had no desire to to make additional loans, when past loans were turning out so badly. Or a minor variation: alternatives to banks (mutual funds, brokers and dealers, etc.) attracted some of the funds that were previously invested in commercial banks, and thus led to a decline in the D part of the currency-deposit ratio. In other words, the decline in commercial banking's market share shows up here as a rise in the currency-deposit ratio.
Whatever the reason, it gives you some idea of the difficulties of "controlling" monetary aggregates. Some critics have argued that Greenspan has starved the banking system of funds; he replied, in essence, that the funds were there (base growth was reasonable) but that the banking system wasn't attracting deposits and (the other thing banks do) loaning them out. As the saying goes: "You can bring a horse to water, but you can't make him drink." Given the enormous changes we've seen in the financial system in the last fifteen years, it may simply be that broad aggregates like M2, which emphasize bank liabilities, are no longer good indicators of how well the financial system is meeting the needs of the economy.
We could tell a similar story about Japan: monetary aggregates have been growing more slowly than the monetary base, as people take money out of banks and invest it elsewhere, including the government's postal saving system. This shows up as a drop in the money multiplier. It's an open question, given the conflicting evidence, whether we view monetary policy in Japan as loose, tight, or in between.
The system also includes 12 regional Federal Reserve Banks, whose location tells you something about the US and its politics in 1913, when the Federal Reserve Act was passed. These regional banks do a lot of the bank supervision, check-clearing, and other day-to-day operations of the system.
With regard to macroeconomic policy, the Fed has three distinct instruments. By far the most important is open market operations, but they also control reserve requirements on deposits and the discount rate on borrowing from the Fed. Each of these policies is determined by a somewhat different combination of players. Open market operations are determined by the Federal Open Market Committee (FOMC), which meets eight times a year and can change policy between meetings in conference calls. The FOMC consists of the seven members of the Board of Governors and five presidents of regional banks. Of the five, the president of the NY Fed is always a member, and the other 11 members rotate (Chicago gets the odd share). Thus the president of the Minneapolis Fed is a voting member every third year. In off years, regional presidents are ex officio members.
Day-to-day open market operations are carried out with repurchase agreements, or repos, on US government securities with sanctioned US security dealers (which includes many of the big financial institutions of various types, including commercial banks). In a typical "system RP" (system distinguishing this from private sector repos) the Fed purchases government securities and sells them back again a few days later. The terminology means that the customer has "repurchased" the securities it temporarily sold to the Fed. The difference between the sale and repurchase prices indicates the interest rate on the transaction. The reverse transaction is called a reverse repo by the market, and a matched sale purchase (or MSP) by the Fed. Both of these instruments allow the Fed to affect the quantity of reserves and the monetary base.
The most popular indicator of the Fed's open market operations is the federal funds rate. In the course of satisfying their reserve requirements, banks often borrow and lend reserves. A bank with more reserves than it needs will loan them to a bank that needs more, and charge interest on the loan. This market for reserves is referred to as the federal funds market, and the rate the federal funds rate. With reserve requirements imposed weekly, the loans are generally of very short duration, often only a day.
The other two policy instruments are changed less frequently. As I mentioned, one of the tasks of the Fed is to provide short-term loans to banks. These loans are secured with government securities. In the old days this took the form of banks selling securities to the Fed at a discount, with the result that this arm of the Fed is known as the discount window. The interest rate on such borrowing is called the discount rate. The discount rates are suggested by the regional banks subject to the Board of Governors and in recent times have been below market rates. The loans are made at the discretion of the Fed, and banks that are seen as abusing their borrowing privileges may lose them. In practice, the discount rate is not viewed as an important aspect of monetary policy, but it is often used to signal the Fed's intentions. A rate cut, for example, may indicate that the Fed foresees lower rates, and possibly looser monetary policy.
The final policy instrument of policy is reserve requirements. As we've seen, banks (and other depository institutions) must hold reserves against deposits in the form of cash or deposits at federal reserve banks. These reserves do not pay interest. The Board sets these requirements within limits set by the Monetary Control Act of 1980. These requirement are changed much less often than the discount rate.
I should add one final policy instrument, foreign exchange intervention, which in the US is the joint responsibility of the Fed and the Treasury.
The Fed has responded to these concerns by changing some of the subtler aspects of monetary control in ways that should make funds more readily available to borrowers. One policy was to eliminate reserve requirements on some types of deposits. As of early December of 1990, required reserve ratios were
Transaction accounts 12% (3 on first 41 million) Time deposits Personal 0 Nonpersonal 3% (if shorter than 18 months) Eurodeposits 3%Other deposits were free of reserve requirements. As of December 27, 1990 the last two categories were reduced to 0. And on April 2, 1992 the rate on transaction accounts was reduced to 10 percent. The effect is to provide banks with more money to loan out and thus enable them to earn more (they avoid, in effect, the "reserve tax" on these deposits). The idea was to make banks healthier and get loans into the system (or, in our model, to increase the money multiplier). There has also been talk of easing the use of the discount window for bank borrowing from the Fed and of relaxing accounting rules on bad loans. All of these suggest that Fed policy consists of more than open market operations.
On the whole, the US has been moving closer to the British system, where reserve requirements are minuscule, thereby reducing the competitive disadvantage of non-interest bearing reserves for banks. Germany has been moving in a similar direction in an effort to put their banks on a more equal footing with those of the Benelux countries. A good guess is that reserve requirements will be much less important worldwide ten years from now than they were ten years ago.
Germany.The formal structure of the Bundesbank mirrors the US Federal Reserve System, although in practice the Bundesbank is widely thought to have greater autonomy. The Directorate consists of presidential appointments and state (Lander) nominees, who serve for eight years each. Price stability is its legally mandated primary objective. Open market operations generally work through repurchase agreements. The repo rate is generally bracketed by two rates on bank borrowing from the Bundesbank. A limited quota of credit is available at the discount rate, which puts a floor on the repo rate. The lombard rate on overnight loans to banks acts as a ceiling. Both are used, as is the discount rate in the US, to signal changes in interest rates. Reserve requirements have been reduced over the last few years. Foreign exchange intervention is used to fulfill obligations of the Exchange Rate Mechanism of the European Monetary System.
Japan.The policy making bodies of the Bank of Japan (BoJ) consist of both political appointees with five-year terms and career civil servants, and operate under the guidance of the Ministry of Finance (MOF). For this reason, most observers regard the BoJ as having little independence, but since their debacle with high inflation following the 1973-75 OPEC price increase, they have been very successful in maintaining price stability. Open market operations are executed with purchases of government securities in the long-term, repurchase agreements on a variety of securities in the short term; the Gensaki rate on bond repos is the most widely cited. Reserve requirements are changed infrequently.
United Kingdom.Authority over the Bank of England rests with the Chancellor of the Exchequer (the finance minister of the government), which gives the Bank more limited control over monetary policy than either the Fed or the Bundesbank. Policy operates through open market operations in Treasury securities ("gilts"). A "minimum lending rate" for Bank loans to discount houses (security dealers) is sometimes used to signal policy-induced changes in interest rates. Reserve requirements are minimal. Policy is circumscribed by its (wavering) commitment to the Exchange Rate Mechanism of the European Monetary System.
BackgroundThe US, as we have seen, has an enormous number of commercial banks, primarily as a result of past policies limiting operations across states. These restrictions have disappeared, for the most part, and many people expect there to be far fewer banks ten years from now than there are now (and we lost several thousand already in the 1980s). And during the last twenty years or so, commercial banks lost some of their traditional business to other financial institutions: large borrowers, for example, turned in increasing numbers to capital markets, where they paid lower interest rates (and commissions to investment banks). The outcome was a decline in the fraction of financial assets going through commercial banks, as opposed to other financial institutions, from nearly 40 percent in the early 1970s to under 30 percent now. Commercial banking, in other words, is under attack from inside and out.
These competitive pressures, and a less restrictive regulatory environment, led many banks to explore strategies with greater growth potential and greater risk. One of the outcomes, apparently, was the huge losses on loans to the third world, to lower-rated businesses, and on commercial real estate. From a taxpayer standpoint, at least, this trade of risk for return has been a mixed blessing.
To some extent these events were worldwide, and there has been a worldwide response to them. On the one hand, governments recognize that market innovations and competitive pressures are breaking down the strict differences, as in the US, between commercial banking, investment banking, insurance, and other financial services. Even in countries, like Germany, where universal banks have had access the full range of financial services, competition has changed the competitive environment. In the US we've seen steady dilution of the Glass-Steagall Act of 1933 that separated commercial banks from other financial activities and of similar limitations on interstate banking. On the other, there has been an international attempt, coordinated by the Bank for International Settlements, to develop uniform standards for bank regulation: the so-called BIS capital-adequacy requirements that went into effect in January of 1993. These standards are intended to guarantee the "safety and soundness" of the global banking system.
In short, the world banking system is much different now than it was ten years ago, and is likely to be more different still ten years from now.
Capital Structure in Banking and ElsewhereProbably the single factor that differentiates commercial banks from other financial institutions is the government guarantee of their deposits. This guarantee is provided in the US by the FDIC for deposits up to $100,000, in return for a "premium" paid by the bank. In other countries deposit insurance is not universal, but one guesses that in a large bank failure in (say) Germany, depositors would be bailed out by the government, thus providing de facto deposit insurance. The rationale for this policy is that sound banks are essential to the economy.
We can see what this does to banks' balance sheets by looking at some examples. As a start, the balance sheet of a firm (in highly simplified form) might look something like this:
Net Assets Liabilities 107,300,000 Bonds 37,300,000 Equity 72,000,000[This is adapted from Hilton's Managerial Accounting, p 774, but you can find similar examples in most accounting and corporate finance texts.]
You might hear, for this firm, that the debt-to-equity ratio is 0.52 [=37.3/72] or that the ratio of equity to net assets is 0.67. Obviously, this is two ways of reporting exactly the same information. In practice things are more complicated, in the sense that we have intermediate types of liabilities: preferred stock (senior to common stock but junior to debt), convertible debt, subordinated debt (debt junior to some other type of debt), and so on, but this example gives you the basic idea.
One user of such information is bondholders, who would like to evaluate their risk that the firm will default. Bonds, by law, have a senior claim on the assets of the firm: they get paid first, then equity owners get whatever's left, if anything. The more equity there is, the more protection bondholders have against an adverse movement in net assets. In this example, net assets can fall (in principle) by 72 million before bondholders lose a cent. [See, for example, the discussion in Bodie et al., Investments, 2nd ed., p 426.] This is one of the things that worried observers about the high debt load of some firms in the 1980s: highly leveraged firms (those with low ratios of equity to net assets) are more vulnerable to default and bankruptcy.
As an example, consider AT&T in 1990 (case prepared by the finance department). The ratio of equity to net assets was 0.63, which means the net asset position of AT&T can fall by 63 percent before bondholders lose any of their money. This is close to a ballpark figure for US nonfinancial corporate business: the ratio of equity at market value to net assets was about 56 percent at the end of 1991.
Another example, closer to the banking industry, is General Electric Capital Corporation. In 1990 it had a ratio of equity to net assets of 13 percent, which was a big enough cushion to provide its bonds with a Aaa rating. GECC is in the financial business, and with net assets of 70 billion is comparable to one of the largest banks in the country. Unlike a commercial bank, however, it had no deposits. Its liabilities included 17 billion of long term bonds and 40 billion of commercial paper.
By contrast, the average ratio of equity to net assets of the commercial bank industry is about 5 percent, which gives bank debt a relatively thin cushion against declines in asset values. One might think that this would make bank debt relatively unattractive (and it's true that very few banks these days have the Aaa rating of GECC). What enables them to have such a low equity ratio is the government guarantee: the largest category of debt holders is depositors, whose asset value is protected by the FDIC.
The conclusion: there's no question that commercial banks benefit from the government guarantee of their debt. In effect, the FDIC has stepped in to bear the risk that banks' liabilities exceed their assets. In return, government regulators generally require that banks not take undue risks.
The BIS Capital RequirementsBy an amazing stroke of luck, the major countries of the world agreed to a more or less common set of standards for regulating banks and related institutions. The core of the policy is a set of minimum standards on ratios of capital (analogous to what we termed equity above) to assets.
We saw that on average banks have a ratio of equity to net assets of about 5 percent. This summary of their capital position ignores two critical aspects of the banking business. The first is the riskiness of the assets: banks engaged in highly risky investments are more vulnerable to losses in asset value than those who are not. The BIS standards address this concern by giving different assets different risk-adjusted values. The second aspect is off-balance sheet transactions: banks engage in a wide range of activities that increase risk but do not appear directly on their balance sheets. These include standby letters of credit, in which they agree to supply funds on request, and derivative products, in which the initial net asset position is typically zero (the two sides of the transaction initially have equal value) but which can entail substantial risk as market conditions change. The BIS standards address this by adding a risk adjustment for off-balance sheet transactions.
The idea behind the BIS standards is that a minimum level of capital (equity and its close cousins) is required for every dollar of risk-adjusted assets. The adjustment for risk weights risky assets more than unrisky assets, so that banks engaged in risky activities must hold more capital. They must hold (for narrow "Tier I" capital, which includes common stock but not preferred stock) capital of at least 4 percent of their risk-adjusted assets. Banks with less capital than this are limited in the types of business they can undertake, banks with more can expand their business until they reach the limit. In this way the capital requirements keep banks' risks within reasonable limits. This also tends to reward successful banks, since high earnings can be used to increase shareholders' equity.
Risk adjusted assets. Suppose a commercial bank has a net asset position of 150 (think of this as billions, to make it a big bank). If all of the 150 billion were invested in treasury bills, then there's little risk to the asset value of the bank and the bank needs little capital to protect it from declines in asset value. But if the 150 were invested in (say) commercial mortgages and then the risk is considerably greater. The BIS capital requirements give assets weights, with larger weights on riskier assets, so that a bank with a risky portfolio gets a high risk-adjusted net asset position.
Off-balance sheet transactions. We do the same with assets that do not appear explicitly on the balance sheet. As an example, consider a standby letter of credit from a bank to a firm who uses it as a guarantee for an issue of commercial paper. They draw on the "SLC" if they are short of funds otherwise. Until the funds are drawn, this transaction does not appear on the balance sheet. Nevertheless, it adds to the risk of the bank. Similarly, a bankers acceptance is a guarantee by the bank that only shows up on the balance sheet if it's used. The BIS standards add these items to risk-adjusted capital.
Example. Consider a bank with (Tier I) capital of 6 and net assets of 150---to be specific, say 90 of commercial mortgages and 60 of unsecured business loans. Under the rules, commercial mortgages get a weight of 0.5 and unsecured loans a weight of 1.0. The risk-adjusted assets on the balance sheet are then 105 [=(0.5)90 + (1.0)60]. Suppose the bank has, in addition, 30 of bankers acceptances, which get a weight of 0.2. The total risk-adjusted assets is then 111 [=105 + (0.2)30]. This bank has a risk-adjusted capital ratio of 5.4 percent [=6/111], so it is within the BIS limit and can expand its business, either by expanding its assets or increasing their risk. Alternatively, if the bank were below the limit, it could raise its capital ratio by shifting into less risky assets. By shifting 30 from business loans to mortgages, for example, they raise their risk-adjusted capital ratio to 6.25 percent [=6/96].
[If this seems terse, it is. I'm trying to get the idea across without getting bogged down in the details.]
In short, the BIS capital requirements limit, in a crude way, the amount of risk a bank can assume given its capital. In this way it provides some protection for the guarantors of deposits, the taxpayer. Banks complain that the risk weights are often poor indicators of risk, which is true, but I don't think there's much question that some risk adjustment is called for. We can expect to see some refinements of this system as its obvious flaws are revealed and exploited.
Other Regulatory IssuesRisk-based deposit insurance. The FDIC now charges different rates for deposit insurance to banks with different risk-adjusted capital ratios. This tends to make risky banks bear more of the burden of their risk to depositors. The rates, though, do not reflect the true, actuarial risk: low-risk banks (and taxpayers) still effectively subsidize high-risk banks.
Market value asset valuation. The BIS requirements generally apply to book values for assets and equity. In the current environment, this is probably less restrictive than market value accounting. Banks with bad loans generally haven't realized all the losses on their books, and thus overstate their assets and equity/capital. The FASB, though, has been moving in the direction of market-based valuation, at least for assets whose value is easily determined. For nonmarketable assets it's not clear how to implement this idea, but as the range of markets expands this may become less of an issue.
The course home page on the Fed and its decisons
on the Federal Funds Rates at FOMC meetings is a useful source of material
on the conduct of monetary policy.