The business of banking

From The Economist print edition

Banks have been at the heart of economic activity for eight centuries. In this,
the second of our schools briefs on the world of finance, we explain why
banks evolved, how they function, what they do, and the challenges they face

WHEN asked why he had robbed a bank, Willie Sutton, a 19th-century American
outlaw, replied: “Because that’s where the money is.” His reasoning is hard to fault:
since modern banking emerged in 12th-century Genoa, banks and money have gone
hand in hand.

Banks are still pre-eminent in the financial system, although other financial
intermediaries are growing in importance. First, they are vital to economic activity,
because they reallocate money, or credit, from savers, who have a temporary surplus
of it, to borrowers, who can make better use of it.

Second, banks are at the heart of the clearing system. By collaborating to clear
payments, they help individuals and firms fulfil transactions. Payments can take the
form of money orders, cheques or regular transfers, such as standing orders and
direct-debit mandates.

Banks take in money as deposits, on which they sometimes pay interest, and then
lend it to borrowers, who use it to finance investment or consumption. They also
borrow money in other ways, generally from other banks in what is called the
interbank market. They make profits on the difference, called the margin or the
spread, between interest paid and received. As this spread has been driven down by
better information and the increasing sophistication of capital markets, banks have
tried to boost their profits with fee businesses, such as selling mutual funds. Such
income now accounts for 40% of bank profits in America.

Deposits are banks’ liabilities. They come in two forms: current accounts (in America,
checking accounts), on which cheques can be drawn and on which funds are payable
immediately on demand; and deposit or savings accounts. Some deposit accounts
have notice periods before money can be withdrawn: these are known as time
deposits or notice accounts. The interest rate paid on such accounts is generally
higher than on demand deposits, from which money can be immediately withdrawn.

Banks’ assets also range between short-term credit, such as overdrafts or credit lines,
which can be called in by the bank at little notice, and longer-term loans, for example
to buy a house, or capital equipment, which may be repaid over tens of years. Most of
a bank’s liabilities have a shorter maturity than its assets.

There is, therefore, a mismatch between the two. This leads to problems if depositors
become so worried about the quality of a bank’s lending book that they demand their
savings back. Although some overdrafts or credit lines can easily be called in, longer-
term loans are much less liquid. This “maturity transformation” can cause a bank to
fail.

A more common danger is credit risk: the possibility that borrowers will be unable to
repay their loans. This risk tends to mount in periods of prosperity, when banks relax
their lending criteria, only to become apparent when recession strikes. In the late

1
1980s, for example, Japanese banks, seduced by the country’s apparent economic
invincibility, lent masses of money to high-risk firms, many of which later went bust.
Some banks followed them into bankruptcy; the rest are still hobbled.

A third threat to banks is interest-rate risk. This is the possibility that a bank will pay
more interest on deposits than it is able to charge for loans. It exists because interest
on loans is often set at a fixed rate, whereas rates on deposits are generally variable.
This disparity destroyed much of America’s savings-and-loan (thrifts) industry. When
interest rates rose sharply in 1979 the S&Ls found themselves paying depositors more
than they were earning on their loans. The government eventually had to bail out or
close much of the industry.

One way around this is to lend at variable or floating rates, so as to match floating-
rate deposits. However, borrowers often prefer fixed-rate debt, as it makes their own
interest payments predictable. More recently, banks and borrowers have been able to
“swap” fixed-rate assets for floating ones in the interest-rate swap market.

Minding the bank

Because banks provide credit and operate the payments system, their failure can have
a more damaging effect on the economy than the collapse of other businesses. Hence
governments pay particular attention to the regulation of banks. Individual banks have
reserve requirements; that is, they must hold a proportion of their deposits at the
central bank, where they are safe and immediately accessible. The central bank
typically pays little or, in America, no interest on these reserves. However it can
charge interest on its loans, which is one way in which the banking system pays for its
own regulation.

As a second cushion against a liquidity crisis, the central bank acts as lender of last
resort. That is, when it worries that solvent banks might struggle to raise money, it
will step in and provide finance itself. America’s Federal Reserve did this after the
1987 stockmarket crash. Ten years later the Bank of Japan did the same because it
thought that the difficulties the country’s banks had in raising money were only
temporary.

Another way in which regulators have tried to keep banks’ heads above water is to
force them to match a proportion of their risky assets (ie, loans) with capital, in the
form of equity or retained earnings. In 1988 bank regulators from the richest
countries agreed that the capital of internationally active banks should, with a few
variations, amount to at least 8% of the value of their risky assets. This agreement,
called the Basle Accord, is being revised, largely because the original makes only
crude distinctions between loans’ different levels of risk.

It is not just the failure of individual banks that gives regulators sleepless nights. The
collapse of one bank can spread trouble throughout the financial system as depositors
from other, healthy, banks suddenly fear for their money. Regulators step in because
they want to prevent a collapse of the entire system.

Governments try to minimise the risk of such failure in several ways. One is to impose
harsher regulation on banks than on other sorts of companies; often, the regulator is
the central bank. Another tack is to try to prevent runs on banks in the first place.
Following the collapse of a third of all American banks in 1930-33, the government set
up an insurance scheme under which it guaranteed to repay depositors, up to a
certain limit, in the event of bank failure.

2
Following America’s lead, other countries have also introduced deposit-guarantee
schemes. Even where they have not, depositors often assume that there is an implicit
guarantee, because the government will step in rather than risk a collapse of the
whole system. In this decade, the Japanese government went to the extreme of
guaranteeing all lenders (not just depositors) to the country’s biggest banks until the
end of the century.

Some argue that these guarantees make bank failures more likely, because they
encourage depositors to be indifferent to the riskiness of banks’ lending. Moreover, as
banks get bigger, they are also likely to conclude that they are “too big to fail”, which
is an incentive to take more risk. Both are a form of moral hazard.

To combat moral hazard, regulators try to be ambiguous about how big is too big, and
to restrict the amount of insurance they provide. In recent years, none of these
measures has prevented ill-advised lending by banks around the world. Failures
include the excessive loans of American banks to Latin America in the 1980s; and
banking crises in Japan, Scandinavia and East Asia.

In many countries, governments have responded to emergencies by nationalising the

worst banks, often pledging to inject capital, take on their dud loans, and reprivatise
them. This is fine in theory, but in practice it often distorts the market for the
remaining privately owned banks by keeping too many banks in business and by
allowing nationalised banks with the benefit of a government guarantee to borrow
more cheaply.

A mixed bank

Another difficulty increasingly faced by both

regulators and banks is the plethora of institutions
that conduct banking business. In Germany, for
example, less than a third of deposits by value are
held within the privately owned banking sector.
Retail banks—those that do business mainly with
individuals—often compete with mutual or state-
owned institutions. Often, such institutions were
founded to provide mortgage financing. Spain has its
cajas—savings banks owned by regional
governments; France, the Netherlands and Japan all
have agricultural co-operative banks that were
created to finance farmers.

Active mutual and state-owned banks can lower the

profitability of privately owned banks, since they
tend to care less about profits. In France and
Germany, banks’ returns are far below those in
Britain or America. As a result, banks’ assets—the
traditional measure of size—are often unrelated to the value of the banks on the
stockmarket (see chart 1).

Increased competition in lending has meant that over the past couple of decades
banks have expanded their lines of business. In Europe, in particular, a new type of
banking, called bancassurance (Allfinanz in Germany) has grown up. This is a fusion of
banking and other financial services and involves banks selling life assurance and
long-term savings products, such as pensions, as well as taking traditional bank
deposits.

3
Banks that combine all of these elements are known as universal banks. Germany’s
big three Frankfurt-based banks, Deutsche, Dresdner and Commerzbank, are all
universal banks, as is HSBC, the Anglo-Chinese giant. So, arguably, is America’s
Citigroup, which is the product of a merger between Citibank’s global-banking
business and the Travelers insurance group.

Banking is a lot messier than it was. In Britain, two big supermarket chains, Tesco and
Sainsburys, now take deposits. Many non-banking firms, such as General Motors, also
now provide credit, although regulators are less worried about institutions lending
money than about those collecting it. American credit-card operators, such as Capital
One and MBNA, have entered the market, using the techniques of database marketing
to identify the most lucractive customers. This has caused established lenders to trim
their rates. Last year the Prudential, Britain’s biggest life insurer, launched a
telephone and Internet bank, called Egg, which has lured a large amount of money
away from traditional banks.

Worse still for traditional banks, many companies in America raise money by selling
bonds rather than by borrowing from banks, a process called disintermediation. In
America, the share of business finance that comes from banks has shrunk from 59%
in 1970 to 46% today. With the single market and the euro, European firms are
increasingly following suit. This has benefited investment banks.

Investment banks, as distinct from ordinary “commercial” banks, help firms raise
money in the capital markets, and advise them
whether to finance themselves with debt or equity.
They underwrite such issues by agreeing, often with
other banks in a syndicate, to buy any unsold
securities, and are paid a commission for this
service. They provide a liquid market in securities,
and (now less than in the past) invest their own
capital, an activity known as proprietary trading. In
addition to advising clients on raising finance, they
also advise on mergers and acquisitions, usually their
most lucrative work.

In America the Glass-Steagall Act has prevented

commercial banks from acting as investment banks
and also from underwriting insurance. Japan has a
similar rule. Although the act is now likely to be
repealed in America, its effects have already been
much weakened. Commercial banks have been able
to underwrite some securities. Investment banks
have offered services that look exactly like current
accounts. Some now even offer credit cards.

As well as entering the investment-banking business, commercial banks have

responded to increased competition by trying to cut costs. The way is being led by
banks in America and Britain, where shareholders hold more sway, labour laws are
less restrictive and deregulation is more advanced. They have merged, replaced costly
branches with cheaper ones in supermarkets and such places, laid off expensive staff
and moved processing to cheap “back-office” administration centres. Those that fail to
get costs under control become the target of takeover bids. This is because it is easier
to cut costs by buying a rival bank and eliminating overlapping branches and
overheads. As chart 2 shows, this process has led to increased concentration in some
banking markets around the world (in Britain the figures are distorted by the

4
reclassification of many building societies as banks, thereby enlarging the banking
market).

Rivals in other countries, especially in continental Europe and Asia, have found it
difficult to cut labour costs. Until recently, they have tried to increase profits by
lending more or by expanding the number of branches (see chart 3). However, thanks
to the single market in financial services, European banks have tried to cut costs by
merging, too. Since monetary union at the beginning of the year, the restructuring has
gathered pace. In Japan, where consolidation has been painfully slow, banks are at
last starting to do the same.

Credit online

Despite all the effort being made by banks to consolidate and cut costs, the advance
of the Internet raises doubts over their long-term prospects. Without the expense of
running branches, online banks have lower costs and can thus offer cheaper products
or higher interest rates to depositors.

E-Loan, an American online bank, originally

offered cheap mortgages; now it also offers cheap
credit cards, car loans and small-business finance.
Americans, tired of the tedious paperwork of
writing cheques and reconciling bank accounts,
are ready to relinquish such duties to the likes of
paymybills.com. These services, which
concentrate on the most profitable customers for
each product, threaten to leave banks with the
unprofitable rump of depositors and borrowers.
Having seen share-trading move online more
rapidly than many had expected, traditional banks
are trying to come up with their own Internet
strategy.

In the next phase of net banking, the Internet will

be used by innovative financial firms to pool
information about customers’ entire financial
affairs, including their non-bank products, such as
mutual funds. This is an even greater danger to
traditional banks, because it threatens to be better than the service they can provide
offline. Future Willie Suttons may need a course in hacking: the money will be in the
ether.

The trouble with banks

From The Economist print edition

5
Nobody loves them, everybody needs them

BANKS have proved themselves to be the most hazardous economic institutions

known to man. Breakdowns in banking lie at the centre of most financial crises. And
banks are unusually effective at spreading financial distress, once it starts, from one
place to another. It it tempting to conclude that banks should simply be abolished.
Unfortunately, that is unlikely to be possible. Banks seem to be necessary.

To see why, consider the job that any financial system has to do. It has to bring
willing lenders and willing borrowers together. One way or another, this involves
processing information. Two kinds of problem arise. First, the lender needs to know
whether a would-be borrower is a good risk. To complicate matters, the keener the
borrower—and the higher the interest rate he is willing to pay on the loan—the more
likely he is to be a bad risk. This is called adverse selection: the most eager borrowers
will be the least desirable, making lenders less willing to lend. The possibility of
adverse selection inhibits productive lending and borrowing.

The other problem that financial systems encounter in processing information is moral
hazard. Once a borrower has his loan, he may try to cheat. In investing the money,
the most he can lose is the amount of the loan. But he may calculate that the greater
the risk he takes with the money, the higher his chances of doing very well. Because
his losses are capped, he is encouraged to take a bigger risk with his investment than
he otherwise would.

Moral hazard becomes acute if the borrower expects to lose the value of his
investment anyway. In that case he has nothing further to lose by taking a much
bigger risk in the hope of turning his fortunes around. If this “gamble on redemption”
works, he keeps all the proceeds after the loan is repaid, but the lender gets no extra
return. If the gamble fails, the borrower is no worse off than if he had acted prudently.
Moral hazard also takes more obvious forms: some borrowers will be tempted simply
to steal the money, or waste it, or otherwise do things that make it less likely that the
lender will be repaid. Lenders deal with moral hazard in the same way they deal with
adverse selection—by lending less than if they had all the information they needed.

This is where banks come in. They are specialists in dealing with adverse selection and
moral hazard, which is why their role in financial systems everywhere is so central.
They develop expertise in knowing what questions to ask borrowers seeking loans;
indeed, they will already know a good deal about them if the would-be borrowers are
existing customers. This allows them to screen out many of the bad risks. Access to
information also makes it possible to curb moral hazard. Banks can monitor what their

6
borrowers are up to; they can set restrictions on what the money is to be used for,
and enforce them by threatening to call in loans or withhold new ones.

Could all this not be done by financial markets, at arm's length? Up to a point, but
banks do have the edge. They are more likely to know a lot about the borrower to
begin with. Moreover, they keep all the benefits of effective appraisal and monitoring
to themselves, so they are willing to bear the risk.

Compare this with a loan that takes the form of a bond purchase. Suppose that one
investor has managed to gather the information needed to curb adverse selection and
moral hazard, and on that basis buys a bond from the would-be borrower. Other
lenders will be able to see this public transaction taking place, and will be able to buy
bonds too, profiting from the first investor's outlay on appraisal and monitoring at no
cost to themselves. Because of this open invitation to free-riding, a market-based
investor will not want to spend much on appraisal and monitoring: unlike the bank
lender, whose transactions are private, he cannot keep the benefits to himself.
Everybody will try to take advantage of everybody else's efforts.

As a result, there will be too little appraisal and monitoring, and the problems of
adverse selection and moral hazard will be much less well controlled. The cost of
lending is driven up; financial activity, and output and incomes in the wider economy,
will all be lower. This is why banks are necessary.

Make that “banks”

Banks may be necessary because of their core financial functions, but just how widely
the business of banking ought to be spread is more debatable. Some economies'
financial systems rely on bank finance far more than others. And the nature of
banking is changing fast almost everywhere. A core of traditional banking—deposit-
taking and straightforward commercial lending—may be indispensable, but how much
of it does a successful economy need? In many countries, traditional banking
represents a diminishing part of what modern banks do.

Economists have long theorised about the relative merits of bank-based finance and
market-based finance. For years it was taken for granted that financial systems
dominated by banks, such as Germany's and Japan's, were better at mobilising capital
and channelling it to the best uses than systems such as the United States' and
Britain's, which give financial markets a larger role. This was especially true, it was
believed, of economies at an early stage of development, where the information-
gathering advantages of banks were crucially important. Believers in market-based
systems emphasised the advantages on the other side, including improvements in the
governance of companies fostered by an active market in corporate control.

By the end of the 1990s these supposed advantages of bank-based systems seemed
rather less compelling. Japan (especially) and Germany were achieving a less-than-
stellar economic performance, in stark contrast to America's remarkable success
during the decade. The current economic consensus, underpinned by new research, is
not that one system is necessarily better than the other, but that either can work fine
so long as certain conditions are met.

The critical factor turns out to be the efficiency of the domestic legal system. If that is
working well, a financial system can deliver the necessary array of financial services
regardless of whether it is based mainly on banks or mainly on markets. Moreover, the
evidence confirms that financial development—measured by the breadth of financial
services, again regardless of whether they are delivered by banks or markets—plainly
promotes economic growth.

7
So the modern findings on financial structure, even though they have retreated from
the earlier idea that bank-based systems work best, offer no reason to jump to the
opposite conclusion. Get the legal fundamentals right, and banks are generally no
better or worse than markets in allocating capital and promoting growth.

In any case, financial innovation is eroding the distinction. Over the past 20 years,
deregulation, competition and technological progress have transformed banking
worldwide. In a sense, banks have become increasingly market-based in the way they
conduct their own business. The models are converging. Nowadays banks bundle
assets (loans) into securities and trade them; increasingly, they earn income from
fees as well as from interest. The once-sharp distinctions between commercial and
investment banking, portfolio management and insurance are blurring. And the use of
financial derivatives—in yet another new set of financial markets—is altering the way
banks manage risk.

These trends are welcome in some respects and worrying in others. Banks can achieve
a much finer degree of control over financial risk than before. In principle, this should
improve the terms of the trade-off between risk and return across the entire economy,
making it possible for an investor to achieve a given return at lower risk, or to earn a
higher return for assuming the same risk as before.
The critical factor
But there are two snags. One is the sheer complexity of turns out to be the
the positions that modern financial derivatives allow banks efficiency of the
to create. Often, according to practitioners themselves,
domestic legal
this outruns the ability of the institutions concerned to
system
manage their risks. The other problem is much deeper.
Sophisticated derivatives ought to help the economy find a
better balance of risk and return, with the risks more accurately allocated to those
who are willing and able to bear them. But what if banks are somehow predisposed to
take on more risk than they should? Innovations that allow banks to gamble bigger
sums would then appear in quite a different light.

This is exactly what modern financial instruments do allow. Leverage—increasing the

likely gain or loss from an investment of a given size—is the salient feature of many
derivatives. Buying an option to acquire shares in a company, for instance, is
equivalent to buying a larger number of actual shares using borrowed money. Selling
such an option can expose the seller to potentially unlimited losses. An increasingly
bewildering array of complex derivatives make it possible to create enormous
leverage. If banks for some reason tend to take on more risk than they should,
financial innovation has unquestionably made this easier to arrange—and harder for
supervisors, or even the bank's own managers, to monitor.

Unfortunately, banks do have a reason to take on more risk than they should. The
reason, paradoxically, is the safety net that governments put in place to prevent bank
failures. By trying to make banks safer, governments give banks the means and the
motive to behave recklessly.

Fond of a flutter

Banks are intrinsically fragile. They borrow from depositors with a promise to repay in
full and on demand, and then mostly invest those deposits in longer-term loans. If
depositors all suddenly decide to withdraw their money at once, as their contract with
the bank entitles them to, the bank cannot meet the demand for funds. It will fail.

Depositors might be induced to withdraw their money by fear that the bank might be
in trouble. Once this fear starts, it becomes self-fulfilling, because if there is any doubt

8
about the bank's safety, depositors have every reason to withdraw their cash: they
lose nothing by doing so. If one bank is perceived to be in danger, other banks are
likely to come under suspicion too. Bank runs, once they start, tend to spread. Note
that equity investors who fear a collapse in share prices face different incentives. As
concern mounts, equity prices fall immediately, which makes it less attractive to sell.
In a stockmarket, therefore, the price decline is somewhat self-limiting. Conversely,
once a bank scare begins, there is no fall in price to deter further withdrawals.
Deposits remain redeemable at par until the bank locks its doors.

At different points during the course of the 20th century, rich-country governments
decided that banks were too vulnerable to this danger. They were also aware that
bank failures could cause damage not just to depositors too slow to get their money
out, but much more widely across the economy. Banks are needed, after all, not just
for intermediation between lenders and borrowers but also to oil the wheels of
everyday commerce. If the banking system collapses, the infrastructure for making
and receiving payments collapses too, and the rest of the economy will follow close
behind.

The solution, governments decided, was to assure depositors that banks were sound,
by promising to step in themselves if need be. They promised to supply a safety net,
by arranging for deposits to be insured and in other ways. If depositors could be
persuaded that their savings were safe, there would be no danger of a bank run and
banks would not fail—or would fail only rarely, and would not take the rest of the
system down with them when they did. Confidence in the
banks would be self-fulfilling, in just the same way as in Once depositors stop
the absence of a safety net lack of confidence is self- caring about the
fulfilling. The cost of providing insurance would therefore soundness of their
be modest. banks, bad banking
quickly crowds out
Recall that banks exist because they are an answer to the good
problem of moral hazard: they can monitor borrowers to
make sure that the funds are not stolen or wasted. But
who monitors the monitors? Banks are borrowers too: they borrow from depositors.
What stops banks from wasting the money they borrow? Partly, the fact that
depositors will not trust their money to an institution that they suspect will be reckless
with it: they will place deposits only with banks that they judge to be safe.

Once governments arrange for deposits to be insured, however, there is no longer any
reason for depositors to worry about the safety of their bank. They will get their
money back anyway. So banks will be able to take bigger chances with the money
they lend. They will be able to lend to bad risks, charging more in interest and
therefore earning bigger profits. Higher lending rates will allow them to pay depositors
more too, enabling them to bid for a bigger share of the market. So once depositors
stop caring about the soundness of their banks, bad banking quickly crowds out good.

Enter the regulator

Governments have long understood this. Their solution is to monitor the banks
themselves. The quid pro quo for deposit insurance—itself absolutely necessary, they
say, to guard against runs—is careful supervision. Require the banks to keep a certain
minimum proportion of their assets in reserve, monitor their lending policies, place
restrictions on the businesses they can enter, and so forth. Having lifted the burden of
bank supervision from depositors, there is only one possible course: nationalise it.

This all seems logical enough, but the success of the policy has been mixed at best.
Banking and financial crises keep happening. And there is good specific evidence that

9
deposit insurance contributes to financial instability, especially in developing countries.
A recent study by Edward Kane of Boston College and Asli Demirguc-Kunt of the World
Bank shows that where effective bank regulation is lacking (as it is in many developing
countries), deposit insurance of the wrong kind does more harm than good. The wrong
kind means, in particular, that it is too generous in its coverage; too well-funded, with
reserves explicitly set aside for repayment of losses; and run by government officials
rather than by the private sector.

This helps to explain why banks have been so deeply implicated in the financial crises
of recent years, as the next section will explain. An exaggerated appetite for risk has
been part of the problem. And as banks have become more sophisticated, even the
best regulators have found it increasingly difficult to keep up.

So if depositors were responsible for supervision instead, would they do better? At first
sight, professional, highly-trained regulators seem a more likely bet than ordinary
depositors. The trouble is that regulators have allowed and even encouraged the
banks to become more “efficient”—lending ever more, against the backing of a
diminishing base of capital. Depositors acting on their own behalf would probably have
resisted that trend and insisted on a more conservative and less “efficient” style of
banking, which nonetheless had the considerable advantage of exposing their deposits
to less risk.

Be that as it may, moral hazard, which banks were invented to tame, has now become
one of the chief weaknesses in the international financial system. In reponse,
governments and regulators have been trying to push more of the burden of
supervision back to depositors and other parts of the private sector, without arousing
fears of bank failure or otherwise destabilising the existing system. This is an
extraordinarily difficult balancing act. The central role of banks in most, if not all, of
the recent big financial crises in developing countries underlines just how difficult—and
how important.

10