Krugman vs Minsky: Who Should You Bank On When It Comes to Banking?

L. Randall Wray | April 2, 2012

Last week I explained why Minsky matters, outlining his main contributions. This was, in part, a response to a blog post by Paul Krugman that appeared to dismiss the importance of trying to find out “what Minsky really meant.” But, more importantly, it was a response to his defense of a simple model of debt deflation dynamics that left banks out of the picture. In Krugman’s view, banks are not very important since all they do is to intermediate between savers and investors, taking in deposits and packaging them into loans.

In my post last week I promised to go into more detail on Minsky’s approach to banking. And right on cue, Krugman expanded on his views in this post.

Now, I know that Krugman’s own specialty is not money and banking, so one would not expect him to have a deep understanding of all the technical details.  However, he is an important columnist and textbook writer, so if he is going to expound upon “what banks do,” he should at least have the basics more-or-less correct. But he doesn’t. Indeed, his views are outdated by at least a century, or more. Can one imagine a science writer at the NYTimes presenting Newtonian physics as state-of-the-art?

If there is any banking “mysticism,” it is what Krugman is presenting—not what Minsky’s followers are arguing. Yes, we need Minsky—whose views even from the 1950s are far more relevant to today’s real world banks than are Krugman’s.

I mean no disrespect here. Like the rest of Krugman’s followers, I think he’s one of the best columnists at the NYTimes–and he covers a great range of topics with flair and good insight. But he cannot be trusted when it comes to money—he just doesn’t get it. What he is presenting is a strange combination of early twentieth century theory plus a throwback to a particular nineteenth century view that was based on an even older “goldsmith” story. Let me explain.

From the 1920s a peculiarly American misunderstanding developed according to which the quantity of bank reserves issued by the Fed could somehow control bank lending and deposit creation. This was called the “exogenous money” approach (the money supply is “exogenously” controlled by the central bank through restriction of the quantity of reserves supplied). It became the starting point for Milton Friedman’s monetarism—which finally ended in the disastrous Great Monetarist Experiment of the early 1980s in the US and the UK in which the central banks tried formal targeting of growth of the money supply.

It didn’t work, and money targets were completely abandoned by all developed nation central banks by 1990.

There was always another tradition, dating back to the Banking School of the early 19th century through Marx and then Keynes, and on to Schumpeter, Gurley&Shaw, Minsky, N. Kaldor, B. Moore and finally to yours truly at the end of the 1980s. It is called the “endogenous money” approach that insists central banks cannot control private money creation by banks through control over reserves.

Very briefly, the idea runs like this. Modern central banks are responsible for maintaining a smoothly operating payments system, which among other things requires that bank liabilities clear “at par” (a one dollar deposit at Chase is valued the same as a one dollar deposit at Bank of America). The Fed makes sure that checks clear among banks and that depositors can use the ATM machines. That means banks must have reserves as required. So the Fed’s control is based on “price”, not “quantity”: it can set the interest rate at which it lends reserves to banks, but cannot determine the quantity.

Further, in practice, bank reserve requirements are calculated based on deposits created up to six weeks previously. So the quantity of reserves that banks are required to hold in systems like the one we use in the US (with required reserve ratios) is always a function of historical deposits. Again, the central bank’s control is price, not quantity. Finally, modern central banks work with interest rate targets (the Fed announces its fed funds rate target), which again means that the Fed can set the “price” then supplies reserves as needed to hit the fed funds rate target.

Still, it has taken US textbooks a long time to catch on, and Krugman is not the only textbook writer who still gets it wrong. (Notice I keep referring to the US—textbook writers outside the US did a better job. For example, Charles Goodhart in the UK wrote a textbook a long time ago that got all this right.)

But in his blog, Krugman’s mistake goes well beyond this common textbook error. He argues that banks really do not create money “out of thin air”—rather, they take in deposits and then lend them out. That is a mistake even the worst textbooks do not make.

Let me just quote three relevant passages from Krugman that summarize his view:

“the self-proclaimed true Minskyites view banks as institutions that are somehow outside the rules that apply to the rest of the economy, as having unique powers for good and/or evil…

First of all, any individual bank does, in fact, have to lend out the money it receives in deposits. Bank loan officers can’t just issue checks out of thin air; like employees of any financial intermediary, they must buy assets with funds they have on hand. I hope this isn’t controversial, although given what usually happens when we discuss banks, I assume that even this proposition will spur outrage…

Yes, a loan normally gets deposited in another bank — but the recipient of the loan can and sometimes does quickly withdraw the funds, not as a check, but in currency. And currency is in limited supply — with the limit set by Fed decisions. So there is in fact no automatic process by which an increase in bank loans produces a sufficient rise in deposits to back those loans, and a key limiting factor in the size of bank balance sheets is the amount of monetary base the Fed creates — even if banks hold no reserves…”

We’ll see below how much Minsky, himself, viewed banks as special. For now let us focus on Krugman’s own views: banks cannot issue checks “out of thin air” but rather must use “funds they have on hand”. And, further, banks are limited to the currency that “is in limited supply”, set by “Fed decisions”.

Krugman sets back monetary theory by at least a century.

Scott Fullwiler has written a beautiful analysis of Krugman here.

He goes through the balance sheets and counters Krugman point-by-point. There is no reason to repeat the detailed rebuttal here.

Instead, I want to give a general overview of Minsky’s approach to the nature of banking: what is it that banks do? Readers should first work through this, then go over to Scott’s piece for the details, and then finally decide for themselves whether it is Minsky or Krugman who ought to be trusted on banks.

But before proceeding look at it this way. A bank deposit is the IOU of the bank, showing up on the liability side of the bank’s balance sheet. Banks have trillions and trillions of dollars of these IOUs on their balance sheets (in the US we have two banks each of which alone has issued $2 trillion in IOUs, and several others are not that much smaller). The IOUs are “contingent liabilities” in the sense that the bank’s creditors can insist on “payment” or “conversion to cash” either on demand (“demand deposit”) or after some waiting period (“time deposit”) or after some specified event.

Where Krugman goes wrong is that he thinks banks operate like “money lenders” that stand on street corners in Chicago—taking in deposits of currency and then lending them out at a higher (usurious) interest rate. Banks supposedly then hold some of the cash as reserves to meet withdrawal of deposits. But since the Fed limits the quantity of cash, bank lending is limited.

But how could that be so? In the first place, the math won’t work. The total amount of cash in existence is less than a trillion dollars—and estimates put at least half of that outside the US, used largely to finance illegal activities in black markets, gun running, and drug smuggling (which is also largely what cash is used for within the US). So, only a small fraction of the total cash is available for banks to receive in deposits in order to make loans. And yet they’ve got trillions and trillions of loans on their balance sheets and have issued just as many IOUs.

Think about the last time you went to a bank. Did you take a wheelbarrow of cash in for deposit, so that your friendly banker could make some mortgage loans?

Honestly, I can remember only one time in the past five years when I made a significant deposit of cash. I had received some $4000 or so in cash travel reimbursement in Colombia (in a dark and somewhat scary underground money changer’s office), and walked around the streets of Bogota with a huge and suspicious bulge of Ben Franklins in my pocket. With some relief, I made it back home and immediately deposited it in my local branch.

Other than that, all of my cash transactions have gone the other way—withdrawals from the ATM. And if you think about it, you rarely see anyone making a large cash deposit in any bank. Unless you’ve got a large cash business (dealing drugs or picking up garbage), you just don’t receive enough cash to keep your bank supplied in green paper.

Actually what I see is that almost everyone who goes to the bank takes cash out! Banks supply cash, they do not receive it in order to make loans. So how could that work? Because whenever banks need cash to meet withdrawal, they do not turn to depositors, rather they call up the Fed. The Fed trucks cash to the banks to stock the ATMs. In turn, the Fed debits bank reserves held at the Fed (these are just the private banking system’s “checking account” held at the Fed).

Now what if a bank is short reserves—will the Fed refuse to send the cash? No. The Fed lends reserves to cover the cash needs. Otherwise the bank would have to close its doors—refusing to meet demands for cash—which would scare the bejeezus out of other depositors and lead to runs on banks. So except for occasional hiccups you do not find the ATM machines shut down or bank doors closed due to cash shortages. Indeed, all money and banking texts that I know of insist that the nonbank public determines the supply of cash—since banks promise to supply it on demand, the Fed provides banks with all they need to meet withdrawals.

It is the Fed that brings the wheelbarrows of cash to the banks—NOT depositors. And the Fed supplies cash NOT so that banks can make loans. Rather, the cash is to cover withdrawals from deposits.

Oh, where does the Fed get the reserves it credits to bank “checking accounts” at the Fed? Out of thin air—keystrokes. Where does the Fed get the green banknotes it trucks to the ATMs? Out of thin air—keystrokes instruct the printing press to print more.

Do you notice a pattern here? Money is always created out of “thin air”.

So Krugman has got to have this banking business all wrong. Whatever deposits you make into banks are almost entirely deposits of bank IOUs. It is all bank money. Where did it come from? Well, from banks. Where did they get it? They created it. How? Thin air.

Look at it this way. You can write an IOU to your neighbor: “I owe you five bucks”. It is your financial liability and your neighbor’s financial asset. Where did it come from? Thin air.

Did you have to get cash first to write the IOU? No. Do you have to have $5 in cash in your pocket to write the IOU? No.

Now, you do have to “redeem” your debt at some point. Your neighbor presents your IOU to you for redemption and you cough up the cash, or you write a check on your bank deposit, or you provide something else of value that is mutually acceptable. When you satisfactorily redeem yourself, your neighbor hands back your IOU and you tear it up.

In this process, you “created money” out of “thin air”; the “money” was your IOU denominated in dollars. (The money you created is destroyed when you repay your debt.)

Now, you might object: but how can that be money? It was just my debt held by my neighbor. It didn’t circulate. The neighbor could not buy anything with it. Yes, that could be true.

On the other hand, it is conceivable that you are well-known and trusted across your entire neighborhood. In that case, the neighbor holding your IOU certainly might be able to pass it in payment for her own IOU to another neighbor (a “third party”). In that case, this other neighbor can present it to you for redemption. Or, your neighbor might hire a local kid to mow the lawn—and then the kid presents it for redemption. So, at least in theory, your IOU could circulate to pay debts or to buy services.

As Minsky always said: anyone can create money; the problem is in getting it accepted.

What we are getting at is degree of “moneyness”. I am making no claim that your IOU is as good as a bank’s IOU—clearly that is not the case. Banks are special. Except for government’s own currency, nothing fulfills money’s functions as well as bank deposit IOUs.

But the dividing lines have always been blurred and are much more blurred today with all the innovations by banks and other financial institutions—what we now call “shadow banks”. For example, except in crisis, money market mutual funds—issued by shadow banks—are almost equivalent to bank deposits. In crisis, however, the government backstop provided to banks in the form of FDIC insurance and Fed lender of last resort promises make bank IOUs much better bets than shadow bank IOUs.

So, unless you’ve got Uncle Sam standing behind you, your IOUs will be “less liquid” and thus inferior “money” in comparison to bank deposits.

(To be sure, there are other reasons banks are special including their specialization in underwriting—determining credit-worthiness, although it is apparent that all too-big-to-fail banks have forgot how to do that. We won’t go into all that now.)

All of that is a preamble to Minsky. I first learned my money and banking from John Ranlett—who used his own textbook (still a good resource if you can find it). And by coincidence, my colleague Stephanie Kelton also studied money and banking with Ranlett. While Ranlett did teach the “deposit multiplier”, he understood this as a relation between reserves and deposits, not that banks literally accept deposits and then lend them out. Further, alone among most recent textbooks, his showed that government spending increases bank reserves (not a topic for today, but a point recognized by MMT) that, too, are “created out of thin air”.

If only Krugman had studied Ranlett he’d never have made such a serious error.

Fortunately, I then went on to study with Minsky—who probably had the deepest understanding of banking of any academic economist.

Sometime around 1983 he handed me a paper by Basil Moore on “contemporaneous reserve accounting” (later published in the Journal of Post Keynesian Economics) and asked me to read it and tell him what I thought. The paper went through the Fed’s attempt at targeting the money supply growth through control of reserves, and argued that even with its change to the procedures used to calculate required reserves (discussed briefly above) the Fed would fail. Basil argued that the Fed must accommodate bank demand for reserves, and went on to argue that we should think of the supply of reserves as “horizontal” at the interest rate target. Later, in 1988, he published a book with the title Horizontalists and Verticalists. Under Minsky’s direction, I went on to write a dissertation and book (Money and Credit in Capitalist Economies) with a somewhat different approach to the same topic.

No need to go into that now. But what is important is that an entire literature developed that carried the views of people like John Ranlett forward—so that today among economists as well as policy makers the “endogenous money” view is dominant. It is easy to find quotes by economists at the Fed, the BIS, and the other international agencies involved in money and banking to back this up.

What Krugman is presenting is the equivalent to 19th century physics—so out-of-date that it is not only misleading but downright embarrassing. His views on central bank control of bank lending through quantitative constraints on reserves has been rejected over the past three decades by almost everyone working in this field. And his view that banks take in cash then make loans was abandoned sometime in the 19th century.

(Finally, according to Krugman, “in the end, banks don’t change the basic notion of interest rates as determined by liquidity preference and loanable funds — yes, both, because the message of IS-LM is that both views, properly understood, are correct.” Yet the ISLM model’s synthesis of liquidity preference and loanable funds theory was shown to be a logical confusion by Keynes back in 1937. In 1975 Duncan Foley showed that the model’s stock-flow relations are suspect. Even John Hicks—the creator of the model—abandoned it as confused nonsense in the 1980s. And mainstream macro finally dropped the LM curve altogether in the “New Monetary Consensus” developed since 1990. So he’s way behind the times on that, too. But all of that is too wonky for a blog.)

Let us turn to a summary of Minsky’s view of money and banking. In many of his writings he emphasized six main points:

  1. a capitalist economy is a financial system;
  2. neoclassical economics is not useful because it denies that the financial system matters;
  3. the financial structure has become much more fragile;
  4. this fragility makes it likely that stagnation or even a deep depression is possible;
  5. a stagnant capitalist economy will not promote capital development;
  6. however, this can be avoided by apt reform of the financial structure in conjunction with apt use of fiscal powers of the government.

Central to his argument is the understanding of banking that he developed over his career. Here I will focus on Minsky’s general approach to financial institutions and policy; I will not provide specific recommendations for policy reform.

According to Minsky, “A capitalist economy can be described by a set of interrelated balance sheets and income statements.” The assets on a balance sheet are either financial or real, held to yield income or to be sold or pledged. The liabilities represent a prior commitment to make payments on demand, on a specified date, or when some contingency occurs. Assets and liabilities are denominated in the money of account, and the excess of the value of assets over the value of liabilities is counted as nominal net worth.

Indeed, all economic units—households, firms, financial institutions, governments—can be analyzed as “banks” since they all take positions in assets by issuing liabilities, with margins of safety maintained for protection. One margin of safety is the excess of income expected to be generated by ownership of assets over the payment commitments entailed in the liabilities. Another is net worth—for a given expected income stream, the greater the value of assets relative to liabilities, the greater the margin of safety. And still another is the liquidity of the position: if assets can be sold quickly or pledged as collateral in a loan, the margin of safety is bigger.

If the time duration of assets exceeds that of liabilities for any unit, then positions must be continually refinanced. This requires “the normal functioning of various markets, including dependable fall-back markets in case the usual refinancing channels break down or become ‘too’ expensive.”

If disruption occurs, economic units that require continual access to refinancing will try to “make position” by “selling out position”—selling assets to meet cash commitments. Since financial assets and liabilities net to zero, the dynamic of a generalized sell-off is to drive asset prices toward zero, what Irving Fisher called a debt deflation process.

Specialist financial institutions can try to protect markets by standing ready to purchase or lend against assets, preventing prices from falling. However, they will be overwhelmed by a contagion, thus, will close up shop and refuse to provide finance. For this reason, central bank interventions are required to protect at least some financial institutions by temporarily providing finance through lender of last resort facilities. As the creator of the high powered money, only the government—central bank plus treasury—can purchase or lend against assets without limit, providing an infinitely elastic supply of high powered money.

These are general statements applicable to all kinds of economic units. This is what Minsky meant when he said that any unit can be analyzed as if it were a “bank”, taking positions by issuing debt.

Yet financial institutions are “special” in that they operate with very high leverage ratios: for every dollar of assets they might issue 95 cents of liabilities; their positions in assets really are “financed” positions. Further, some kinds of financial institutions specialize in taking positions in longer term financial assets while issuing short term liabilities—that is, they intentionally put themselves in the position of continually requiring refinancing.

An extreme example would be an early 1980s-era thrift institution that holds 30 year fixed rate mortgages while issuing demand deposits. Such an institution requires continuing access to refinancing on favorable terms because the interest rate it earns is fixed and because it cannot easily sell assets. This can be described as an illiquid position which requires access to a source of liquidity—Federal Home Loan Banks or the Fed.

Still other kinds of financial institutions—investment banks–specialize in arranging finance by placing equities or debt into portfolios using markets. They typically rely on fee income rather than interest. In normal circumstances they would not hold these assets directly, but if markets become disorderly they can get stuck with assets they cannot sell (at prices they have promised) and thus will need access to financing of their inventories of stocks and bonds. Some might hold and trade assets for their own account, earning income and capital gains, or might do so for clients.

Thus there are many kinds of financial institutions. Minsky distinguished among traditional commercial banking, investment banking, universal banking and public holding company models.

A traditional commercial bank makes only short term loans that are collateralized by goods in production and distribution. The loans are made good as soon as the goods are sold—this is the model the old Real Bills doctrine had in mind. The bank’s position is financed through the issue of short term liabilities such as demand and savings deposits (or, in the 19th century, bank notes). The connections among the bank, the “money supply”, and real production is close—the sort of relation the quantity theory of money supposed.

Essentially, the firm borrows to pay wages and raw materials, with the bank advancing demand deposits received by workers and suppliers. When the finished goods are sold, firms are able to repay loans. Banks charge higher interest on loans than they pay on deposits—with the net interest margin supplying bank profits.

Note that banks do not sit and wait for deposits in order to lend. Rather the process is precisely the reverse: the bank accepts the IOU of the firm that needs to pay for wages and raw materials, then creates a deposit (or in the old days, a bank note) that the firm uses for its purchases. As endogenous money types say: “loans create deposits”—not in some metaphysical sense but in the sense that the bank “buys” the IOU of the firm by issuing its own deposit IOUs. It “finances” its position in the firm’s IOU by issuing its own deposit IOU.

Once the firm finishes production and sells the output, it receives deposits and uses these to retire the short term loan. Following our discussion above, the firm “redeems” itself by bringing back to the bank the bank’s own IOUs; repayment of the loan “destroys” the bank deposits (the loans and deposits on the bank’s balance sheet are simultaneously debited).

Now it is likely that many of the sales the firm makes are to consumers who bank at some other bank, so the firm receives bank checks (or, in the old days, banknotes) drawn on other banks—and submits these to its own bank. But the checks are cleared (either through the Fed or a private interbank settlement system) at par. So it is not necessary for the firm to return to its bank that bank’s own IOUs—the IOUs of any bank will do.

If deposits are to maintain parity (with each other and with cash), losses on assets must be very small because the commercial bank’s equity must absorb all asset value reductions. It is the duty of the commercial banker to be skeptical; as Minsky loved to say, a banker’s cliché is “I’ve never seen a pro forma I didn’t like”—borrowers always present a favorable view of their prospects. This is why careful underwriting is essential.

While it is true that loans can be made against collateral (for example, the goods in the process of production and distribution), a successful bank would almost never be forced to take the collateral. A bank should not operate like a pawn shop. As Martin Mayer says, banking has always been a business where profits come over time as borrowers pay principal and interest. He alludes to the morality of a loan officer, whose success depends on the success of the borrower. It goes without saying that betting on the failure of one’s borrower is inimical to the duties of a commercial bank. (Memo to Goldman Sachs: pay attention.)

The banker holds the key—he is the “ephor of capitalism”, as Minsky’s original dissertation advisor, Josef Schumpeter, put it—because not only do entrepreneurs have to be sufficiently optimistic to invest, they must also find a banker willing to advance the wage bill to produce investment output.

By financing the wage bill of workers in the investment goods sector, commercial banks are promoting the capital development of the economy even if they do not actually provide finance for position-taking in investment goods. Hence, we can separate the issue of producing capital goods from ownership of them.

For Schumpeter, and for Minsky, the “ephor of capitalism” breaks the simple circuit of production and consumption of wage goods—in which banks simply finance production of consumer goods by workers whose consumption exactly exhausts the wage bill required to produce them. In other words, the ephor allows generation of profits by financing spending of those not directly involved in producing consumption goods.

To go further would get us into complicated matters. But the next step would be to discuss the role of the investment banker, who finances the long term positions in capital assets. This is a quite different activity, which allows savers to choose between holding liquid (financial) assets or positions in real assets (either directly by owning a firm, or indirectly through ownership of shares). Glass Steagall maintained a separation of the investment banking and commercial banking functions. Lines were blurred when we first allowed bank holding companies to own both types of banks, and then gutted and finally repealed Glass Steagall. For those interested in Minsky’s views on all of this, see my piece at the Levy Institute.

Let’s recap.

  1. Banking should not be described as a process of accepting deposits in order to make loans.
  2. Rather, banks accept the IOUs of borrowers, then create bank deposit IOUs that the borrowers can spend.
  3. Indeed, often the bank simply accepts the IOU of the borrower and then makes the payment for the borrower—cutting a check in the name of the car dealer, for example.
  4. Like all economic units, banks finance positions in their assets (including IOUs of borrowers) by issuing their own IOUs (including demand deposits).
  5. Banks use reserves for clearing with other banks (and with the government—a topic not covered here). Banks also use reserves to meet cash withdrawals by customers. Bank reserves at the central bank are debited when they need cash for withdrawal.
  6. In some systems, including the US, the central bank sets a required reserve ratio.  But this does not provide the central bank with any quantitative controls over bank loans and deposits. Rather, the central bank supplies reserves on demand but sets the “price” at which it supplies reserves when it targets the overnight interest rate. In the US the main target is the fed funds rate. Fed control over banks is all about price, not quantity, of reserves.

That gets us up to a 20th century understanding of banking. Krugman could benefit by standing on the shoulders of a giant like Minsky.


5 Responses to “Krugman vs Minsky: Who Should You Bank On When It Comes to Banking?”

RSS feed for comments on this post. TrackBack URL

  1. Comment by Greg Nowell — April 3, 2012 at 1:44 pm   Reply

    Macleod gave a very lucid account of endogenous money creation in the 1890s. But he was stuck on gold capitalization of banks and couldn’t quite see his way to a pure endogenous theory.

    There is one point in this presentation that isn’t quite right. It is true that most of us take printed money out and never put it back in. That does not mean it is a one way trip for the physical dollar. The printed money accumulates in places like Wal-Mart and the local football stadium and is carted back to banks in Brinks trucks. At one point Japan had a negative interest rate on deposits in part because, having no use for much cash in a liquidity trap, they had to pay to have the cash hauled to the central bank, and passed the charge on to customers.

    Aside from these very minor comments I read this with great interest.


  2. Comment by Dan Sadler — April 13, 2012 at 11:58 pm   Reply

    Nice article! Krugman doesn’t understand the multiplier effect. Neither do most people. I saw a youtube presentation about how banks create money, and I was really surprised. I didn’t understand fractional reserve banking, and I have now read and watched several presentations on how it works.
    I tried to explain how this works to my father, who did not believe me.
    I tried to explain that banks create money out of thin air, but he thought I was crazy. He then told me to ask my father-in-law, who is an was an economist. I asked my father-in-law, who confirmed that I was right and then explained the multiplier effect to me.
    However, my father-in-law is kinda old, and he has trouble explaining economics to me. He is now 83 years old, and doesn’t talk very well anymore.
    Nice article, thanks a lot.
    Sincerely, Daniel L Sadler, son in law of John G Ranlett, author of “Money and Banking, An Introduction to Analysis and Policy”

  3. Comment by LRWray — April 17, 2012 at 11:13 am   Reply

    Hey Dan, Thanx and please say Hi for me to your father-in-law Professor Ranlett! He taught me about banks.

  4. Comment by Patrick — June 27, 2013 at 4:17 pm   Reply

    How do you explain initial capitalization? Banks don’t simply appear, they have to have initial capital reserves in order to access the fed funds window. While they may lend out vastly more than the initial capital, there has to be something to base it on. Its the Fed certification that they meet the basic requirements, governance, and regulation to be a bank. For the neighbor down the street, the initial capital is reputational, or drawn on their obvious assets and ties to the geographic location. That’s why banks build big, substantial structures, to show they’ll be around. This analysis pre-supposes banks existence, but they come from somewhere, they have specific regulations that control their initial existence.

Leave a Reply