For six plates this volume has watched a single kind of middleman, the dealer who stands in a market quoting two prices and keeping the gap between; this plate lifts its eyes to a different and far older one. Between the saver, who has money and no present use for it, and the borrower, who has a use for money and none of his own, there stands the bank — and its spread, the gap between the interest it charges the borrower and the lesser interest it pays the saver, is among the largest and oldest of all the prices of standing between. The question the bank raises is the one the dealer raised, but deeper, for it is not obvious why the bank should exist at all. The saver and the borrower want the same thing, the one to lend and the other to borrow; why do they not simply find each other and deal direct, sharing between them the spread the bank would otherwise take? The answer, worked out most fully by Douglas Diamond, is that lending is not the simple handing-over of money it appears to be, but a relationship that must be watched — and that the watching is precisely what the saver cannot do, and the bank can. The bank's spread is the wage of that watching, done for the saver who cannot do it himself; and the plate's task is to weigh that wage, and to find in the bank both the most useful of the volume's middlemen and, in one respect, the most dangerous.
Consider what the saver would have to do to lend his money safely on his own. He would have to assess the borrower before lending — judge whether the venture is sound, whether the man is honest, whether the loan will be repaid. He would have to watch the borrower after lending — see that the money is used as promised and not gambled away, that the business is run with the repayment in mind, that no quiet ruin is gathering. And he would have to be prepared to act if things went wrong — to intervene, to call the loan, to seize what was pledged. All of this is costly, in effort and expertise and time, and most savers possess neither the skill nor the inclination for it. Worse, the cost falls absurdly when a single borrower is funded by many small savers, as large borrowers must be. If each of the hundred savers who together fund a firm tries to watch it, the watching is done a hundred times over, at a hundredfold waste. And if, seeing that waste, each saver leaves the watching to the other ninety-nine, then by the ordinary logic of the shared burden the firm is watched by no one at all, each saver free-riding on a vigilance that, because all rely on it, none supplies. Direct lending by dispersed savers thus founders on the watching it requires: performed, it is ruinously duplicated; economized, it vanishes. The borrower goes either over-watched or unwatched, and in neither case is the saver well served.
The bank is the institution that cuts this knot, and Diamond showed exactly how. Instead of each saver lending to the borrower and trying to watch him, the savers lend to the bank, and the bank lends to the borrower and watches him — once, on behalf of them all. The hundredfold duplication is collapsed into a single watching; the free-riding is abolished, because the one who watches is also the one on the hook to repay the savers, and so has every reason to watch well. The savers have delegated their monitoring to the bank, and the bank has become their delegated monitor, doing for all of them together the watching that none of them could do alone or would do in common. This is the bank's first and central service, and its spread is the wage of it: the borrower pays a higher rate than he would owe a perfectly trusting lender, the saver receives a lower rate than the borrower pays, and the difference is what the bank is paid for standing between and watching — for converting a relationship that the saver could not safely enter into one that, through the bank's vigilance, he can. What the bank sells is not money, which the saver already has, but the monitoring of the borrower, which the saver lacks the means to perform.
What the bank sells is not money — the saver has that — but the watching of the borrower the saver cannot do for himself; and its spread is the wage of a vigilance performed once, for all.
But the delegation seems, at first, only to move the difficulty rather than to solve it. If the saver could not trust the borrower without watching him, how can he trust the bank without watching it? Has he not merely exchanged a borrower he must monitor for a bank he must monitor, and gained nothing? Diamond's answer to this is the subtle heart of the theory, and it turns on diversification. The bank does not lend to one borrower but to a great many, and the misfortunes of borrowers, taken in number, largely cancel: some fail, but most repay, and the proportion that fails is, across a large and varied book of loans, highly predictable. This means that the bank, though each of its loans is risky, can itself be very nearly safe — its total receipts steady even as individual borrowers default — and can therefore promise its depositors a return that is almost certain to be paid. And a promise almost certain to be kept needs little watching. The saver need not monitor the bank closely, because the bank's diversification has made its promise to him nearly riskless; the cost of trusting the bank is small precisely because the bank, by spreading its lending across many borrowers, has made itself trustworthy. Diversification is thus what makes the delegation work: it is the reason the monitor need not itself be much monitored, and the reason the chain of trust, which seemed only to have been lengthened, is in fact made sound.
There is, beneath the monitoring, a second service the bank performs, and it is in some ways the more remarkable, for the bank does not merely pass the saver's money to the borrower but transforms it in the passage. The saver wants a claim that is short and safe — money he can have back whenever he likes, with no risk of loss. The borrower needs funds that are long and patient — committed for years, and able to bear the risk that his venture takes time or goes awry. These two wants are opposite, and could never be matched directly: no saver would lend long and risky what he wishes to hold short and safe. The bank reconciles them by being both things at once. On the side it shows the saver, it is short and safe: deposits payable on demand, secure against loss. On the side it shows the borrower, it is long and risky: loans committed for years, exposed to the borrower's fortunes. The same money is short and safe as it enters and long and risky as it leaves, and the bank's art is to stand astride that transformation — to give the saver the liquidity and safety he wants and the borrower the patience and risk-bearing he needs, out of the one pool of funds. This maturity and risk transformation, layered atop the delegated monitoring, is the deeper reason the bank's service cannot easily be done without; and it is also, as the plate will close by showing, the deeper source of the bank's peculiar fragility.
The same money is short and safe as it enters the bank and long and risky as it leaves it; the bank's art is to be both at once, giving the saver liquidity and the borrower patience from a single pool.
For all its usefulness, the bank stands where every middleman in this volume has stood, between two parties who must pass through it, and it is heir to the same two temptations of the position. The first is power. The bank sits at a chokepoint between the saver and the borrower, and where banks are few, or a borrower has nowhere else to turn — a small business in a town with one lender, a household with no other source of credit — the spread can widen past the cost of the monitoring and the transformation into the toll of mere position. The lending rate climbs higher, and the deposit rate sinks lower, than any service the bank performs would warrant, the excess reflecting not the cost of watching the borrower or transforming the term but simply the borrower's and the saver's want of an alternative. This is the bank as toll-keeper rather than servant, charging for being unavoidable, and it is the same figure the earlier plates met in the dealer who quoted a wide spread because no rival quoted a narrower; here only the stakes are larger, for credit is more nearly a necessity than immediacy, and the captive borrower has even less choice than the impatient trader. Competition among banks presses this toll down, as competition always does; concentration and captivity let it back in.
The second temptation is not a temptation at all but a vulnerability, and it is the gravest in the volume, for the very transformation that makes the bank valuable is what makes it fragile. The bank has promised the savers their money on demand, but it has lent that money out long; at any moment, only a fraction of it sits ready to be paid. So long as the savers come for their money a few at a time, this is no trouble, for the bank holds enough for the ordinary trickle. But let the savers come all at once — let a rumour spread that the bank is unsound, whether true or false — and the bank cannot pay, for its money is out in loans it cannot call in; and the saver who believes the others will demand their money has every reason to demand his first, so that the mere fear of a collapse can cause one, and a bank perfectly sound can be destroyed by a panic alone. This is the bank run, and it is the dark twin of the bank's great service: the liquidity the bank manufactures for its savers is liquidity it does not truly have, conjured from the assumption that not all will want it at once, and that assumption can fail. Worse, because all banks transform alike and lend into the same economy, a shock that frightens the savers of one frightens the savers of all, and the diversification that makes each bank safe against the failure of its own borrowers avails nothing against a storm that strikes every bank together. The bank's spread, then, is the wage of a service genuine and profound — the monitoring delegated, the transformation performed — but it is underwritten, always, by a fragility that the calm conceals and the panic reveals, and shadowed, like every cut in this volume, by the toll its position can command.
So the bank is at once the most useful of this volume's middlemen and, in its fragility, the most perilous — the delegated monitor who watches the borrower the saver cannot watch, the transformer who makes the saver's short safe money into the borrower's long risky loan, and the keeper of a promise of liquidity that a panic can expose as more than it can keep. Its spread is earned in the watching and the transforming, swollen where its position lets it levy, and underwritten by a danger that sets it apart from every other price in the table. The matter of why such an institution must exist at all — why the watching and the transforming are done within a firm rather than across a market — belongs to the volume on the firm, where the question of which tasks the market hands to institutions, and why, is taken up in full; the bank is only its clearest and most consequential instance. The next plate returns from the bank to the trading floor, and to a middleman of a wholly different temper: not one who stands between a buyer and a seller, nor between a saver and a borrower, but one who stands between two prices for the same thing in two places, and, by buying where it is cheap and selling where it is dear, presses the two into one. As the table says of every price, first to last: the argument remains open.