In preparing an essay defending the now-unpopular view that (ordinary) banks are credit intermediaries, I'm re-reading the literature saying otherwise. What rubbish! A jumble of misrepresentations, non-sequiturs, and otherwise shoddy arguments.
Consider this article by Michael Kumhof and Zoltan Jakab, who are among the more high-profile critics of the intermediation view, with impeccable Bank of England and IMF credentials. justmoney.org/m-kumhof-and-z…
In it, they say that Irving Fisher and other 1930s era proponents of 100-percent reserve banking favored it because, having realized that fractional reserve banks were "creating" money instead of acting as pure intermediaries, they wanted to make them true intermediaries.
W/ all due respect, that isn't so. Fractional reserve banking is as old as banking itself (in fact, 100% reserve banking isn't "banking" at all), and _everyone_ understood by 1930, and long before, that banks "create" money by lending. Adam Smith, certainly did.
But--and here is one of the principal errors of the "anti-intermediation" view: that banks "create" exchange media, including deposits and, in Smith's day, "banknotes," in amounts exceeding their reserve or "base money" holdings, _doesn't_ mean that they "create" credit.
So one can either object to fractional reserve banking, or favor it, without rejecting the "intermediaries" view. In fact, as a moments reflection should make clear, banks cannot possibly be "intermediaries," pure or otherwise, _unless_ they hold fractional reserves!
In contrast, a bank that merely warehouses coin is...a warehouse! To be an intermediary, a bank has to lend; and as soon as it lends its assets can't but exceed it's holdings of base money reserves.
So when Kumhof and Jakab describe not just older but "modern neoclassical intermediation theories" as resting on the assumption that "banks are...warehouses..in a very literal sense," they're taking utter nonsense!
It's true that some, including followers of Murray Rothbard, imagine that forcing banks to be mere money warehouses would make them "purer" intermediaries; but they are also spouting nonsense, as I've argued at length elsewhere.
If fractional reserve banking, and the associated "creation" of exchange media consisting of bank IOUs, doesn't itself contradict the intermediation theory, what, then, _does_ define the limits of such intermediation? Or is _any_ amount of money creation consistent with it?
To answer, let me first point to another false statement in Kumhof and Jakab's essay. They hold that banks remain intermediaries only to the extent that growth in their balance sheets starts w/ "banks collecting new deposits of previously saved physical resources from savers...
and ends with the lending of those same resources." Note, first of all, that even if this were all banks did, it could suffice to give rise to arbitrarily tiny reserve ratios and arbitrarily large balance sheets!
In fact, what the authors suppose to be a "requirement" of "modern neoclassical intermediation theories" doesn't rule out _infinite_ monetary expansion, which would occur if all banks chose to lend every bit of the "saved physical resources" deposited with them. In that case,
that physical stuff would be like so many hot potatoes, passed on and on, or lent and relent _ad infinitum_, because no bank wants or needs them.
In fact, proponents of the intermediation view have a quite different understanding of what it means for any bank or system of banks to engage in pure intermediation, and hence what it means for banks to extend credit beyond a point consistent w/ that ideal.
Here again, Adam, Smith already had it right. No, I'm not about to defend the real-bills doctrine! I've always rejected it. But Smith also referred to a different notion--that of money's "channels of circulation."
The pure intermediation case is one in which banks confine their money creation so as to fill but not "overflow" those channels. Now let me translate the metaphor into more modern terms.
In those terms, taking the level of prices as given, so long as any additional money banks create by lending is matched by an equivalent growth in the demand for real (bank) money balances, the banks are acting as savings-investment intermediaries.
The "savers" here are those willing to add more bank deposits (or notes) to their average real) money balances. Smith's "channels of circulation" are thus the many "pockets" of money holdings across which expenditures flow. Deeper pockets = increased demand for real balances.
So, imagine a fractional reserve banking system with X (fixed) reserves and initial equilibrium values of M (> X) and P. As the demand for real balances grows, banks can expand their balance sheets at the same rate, and to any extent, without ceasing to be intermediaries.
(Of course, all this assumes that people are happy to hold bank-created money. If they cease to be happy to do so, that also shrinks the Smithian channels of circulation--perhaps dramatically. Smith got that, too.)
Now, several things. First note that, there is _nothing_ about this alternative view of the conditions for banks to be intermediaries that says that "physical resources" have to come to them _before_ they can make new loans.
So long as the stuff they lend ends up in one of those pockets that make up the "channels," the lending is still intermediation. So defending the pure intermediary does _not_ mean believing that bankers have to wait for fresh deposits to come their way to them to make loans.
It _does_ mean, however, that someone, somewhere, has to be willing to hold more of a bank's liabilities in order for it to increase the size of its balance sheet. (To be continued.)
Thus, banks can also be "pure intermediaries" while relying on wholesale funding. Or they can lend in anticipation of growth in the demand for their money. In any case, it isn't essential to the theory that they have prior deposits or unwanted reserves.
It's wrong, on the other hand, to suppose, as Kumhof and Jakab do, that a bank isn't acting like an intermediary is it lends after "collect[ing]] checks or other similar financial instruments," as opposed to to deposits of "physical" money, e.g., gold or central bank notes.
They say that wouldn't be intermediation because banks in that case the collected instruments don't really represent savings. But while they may not represent net aggregate savings, such funds are certainly net savings for the banks that acquire them. Sometimes...
Bank A does more intermediating at Bank B's expense, as happens, say, when someone switches banks, or a borrowers from one bank deposits funds in another.
So far I've said nothing about the forces that can confine banks to being intermediaries, instead of "creating" not just money but credit--that is, lending beyond levels consisting with prior real demand for their liabilities. But these forces exist, and they're very important!
To appreciate them, one must first understand that these forces confront _competing_ banks only. They do not confront fiat-money issuing central banks. Nothing keeps the latter from creating oodles of credit, and thus doing other nasty things, like fueling inflation.
It is, of course, central banks' unique credit-creating capacity that allows them to perform most of the special roles they perform--roles no ordinary, competitive bank is capable of performing.
Indeed, it's not at all improper to say that such banks create money "out of thin air." But to suppose that ordinary commercial banks do the same is to betray a very serious failure to understand the constraints ordinary banks face. (To be continued.)
What sets central banks apart is the fact that they alone supply their economy's "final" means of payment and settlement. Other banks' demandable IOUs can serve instead of central bank money for most payments. But they are mere promises, not means of final settlement.
In particular, they aren't means of interbank settlement. When an ordinary commercial bank receives a credit from some rival bank, it doesn't hold that banks IOUs: instead, it "collects" from the rival bank, in a process known as "clearing and settlement."
Nowadays most of this clearing and settlement happens on central banks' books: the CB credits the settlement (or reserve accounts) of banks that receive items from rival banks, and debits the accounts of those that owe funds to others.
These routine (at least daily) settlements are why banks keep accounts with central banks, and why they generally hold some CB money in them. It is the need CB money for net interbank settlement that limits ordinary banks' ability to create credit.
Settlement balances aren't "thin air."They are real resources ordinary banks must come buy one way or another. And though at any time banks may hold a slim reserve of such balances only, it doesn't mean that the don't actually employ large quantities.
Now comes the important part. Suppose a bank's balance sheet is such that it qualifies as a pure intermediary: the demand for real balances of its money equals the quantity outstanding. Loans are being repaid, but new ones take their place,
Deposit withdrawals are offset by equal new deposits. The bank owes no more to other banks at the end of each clearing session than they owe to it. It could of course borrow money, but it has no loan prospects such as will allow it to re-lend the borrowed funds at a profit.
In short, our bank is in a state of stationary equilibrium. A "pure intermediary" stationary equilibrium.
Alas, our banker is unhappy. He has only a few dollars on hand in his CB account--plenty under the circumstances. but he wants to be a bigger banks. He has read some Bank of England papers, and he decides that he can create money out of thin air after all.
So, he makes another $1 million loan _above and beyond_ whatever amounts he had on his books, by offering an attractive rate. He finds a borrower and credits his account accordingly. So far so good!
Al;as for our banker, people seldom borrow money to keep it in a bank. They borrow it to _spend_ it. (This, BTW, is a crucial difference between the deposits banks create when they make loans, and their "core" deposits, which are _brought_ to them by other customers....
Bankers understand the difference. Would that all who theorize about banking understood it!). Now consider: in the U.S. we still have over 10,000 banks and credit unions. So, when our banks' loan gets spent, chances are the funds end up being placed in other banks.
(Recall that, _ex hypothesi_, the real demand for our bankers' liabilities hasn't increased. That is, they are now overflowing the Smithian "channels of circulation.") The banks that are the beneficiaries of these deposits duly arrange to clear and settle them.
Whence, low and behold (and assuming the other bankers were also enjoying a state of stationary equilibrium, which they wisely chose not to disturb), our banker finds himself having to scrounge-up $1million, by borrowing from the Fed or from other lenders.
But (again, _ex hypothesi_), the banker will lose money by funding the loan that way. We are, in short, a long way from the "thin air" hypothesis. (We can see, however, that the more generous the CB's lending terms are, the more reality will appear to fit that hypothesis.)
Also, if our bankers does manage to borrow from some other bank at a profitable rate, he will loosen the constraint on his own lending only by further constraining the lending bank, which must itself secure resources sufficient to cover its own clearing loss.
So ordinary banks are in fact constrained in their ability to extend credit, and because the constraints they face depend on the demand for their monetary IOS, they help to confine their lending w/in "pure intermediary" limits.
This doesn't mean that those limits are always respected: for all sorts of reasons actual, competitive banking systems don't always conform to the "pure intermediary" ideal. But saying so is a long way from conceding that there's no merit to the intermediary view of banks.
On the contrary: IMHO, despite being in vogue, its the opposite view--the view that holds that banks can create credit without regard to the public's willingness to save by holding their IOUs--that is furthest from the truth, making it a very poor basis for banking policy. (End.)
(Note: it also doesn't mean that bank lending can never help restore a depressed economy to full employment. If an excess demand for real balances is behind the excess supply of goods and labor, the supply of bank money is, ipso facto, < what perfect intermediation entails.)
(1930s-era 100-percenters were mainly concerned about bank runs and deflation, not banks being able to "create" too much credit.)
For "taking" please read "talking."
• • •
Missing some Tweet in this thread? You can try to
force a refresh
