Tobin vs. Yeager, my view
I really hesitated to write this post. For the last 10 days my blogging frequency has been close to nil as I re-read and gave a thought to James Tobin’s Commercial Banks as Creators of Money, Leland Yeager’s What are Banks, as well as several blog posts, including David Glasner’s (here, here and here). In the end, it looked to me that everything that could be said had already been said. But I’ve been asked to provide my opinion. Moreover, I believe that most of what has been said so far derived from economists’ point of view (which doesn’t mean this is wrong per se) and that a financial analyst/practitioner could potentially bring useful points to the debate.
As I said, I’ve been asked to provide my opinion. Let me give you a quick answer. I believe that both Tobin’s New View and Yeager’s Old View (?) are right. I also believe they are both wrong. Some of their arguments are true. Some others seem to contradict other points they make.
I believe that banks’ expansion is indeed limited by economic constraints. Perhaps just not the exact ones Tobin was thinking of. I also believe banks can and usually indeed do fully ‘loan up’. But perhaps not exactly the way Yeager was seeing it as I believe that reserve requirements (in particular, internally-defined ones, see below) are part of banks’ economic constraints.
A lot of the discussion surrounding the debate in the economic area/blogosphere rests on the nature of bank inside money and whether it is subject to the ‘hot potato’ effect. This is very clear in David Glasner’s posts. I encourage you to take a look at them (including their comment sections) although this can be a little technical for non-economists. But here I am going to take a slightly different approach, more finance theory-based.
First, I don’t believe in Tobin’s arguments that demand deposits’ yields are key. Tobin says:
Given the wealth and the asset preferences of the community, the demand for bank deposits can increase only if the yields of other assets fall. […] Eventually, the marginal returns on lending and investing, account taken of the risks and administrative costs involved, will not exceed the marginal cost to the banks of attracting and holding additional deposits.
I don’t know anyone who keeps his/her money in demand deposit accounts because of their yield (by ‘demand deposit’ I mean checkable deposits, not saving deposits available on demand, this is an important distinction). People maintain real balances in demand deposits to cover their cost of living (i.e. expected cash outflows over a given month). Here, on the ‘hot potato’ issue, I side with Yeager. Demand deposits provide a convenience yield. Nominal yields are pretty much unimportant. It is only when banks’ customers have surplus balances in their account that yields become important in the choice between using them for consumption or investment (and to pick the type of investment). But this does not concern demand deposits. It is hard to imagine anyone transferring most (or all) of his money from a demand to a saving account (or any other sort of investment) for yield purposes…
In addition, banks always pay close to nothing on demand deposits. In order to grow their deposit base, banks barely vary rates on demand deposits. Saving accounts and related financial products (retail bonds, certificates of deposit, etc.) that are not media of exchange are the ones used by banks to attract customers.
But the Tobin/Yeager debate focuses on demand deposits due to their very nature. And there seems to be some confusion in both Tobin’s article and Glasner’s posts. Tobin explicitly refers to demand deposits, and how competition from non-bank FIs would push up their interest rates. This is unlikely as described above. Saving deposits would be the ones to suffer. But they are not media of exchange and not the topic of the discussion. As a result, I find myself in general agreement with Bill Woolsey, who commented on Glasner’s posts.
Second, Tobin seems to downplay the role of reserve requirements. What he fails to see is that they also are part of banks’ economic constraints. He also contradicts himself when he declares that reserve requirements are only a legal constraint that kicks in before natural economic constraints prevent the bank from expanding, only to say a few pages later that reserves…aren’t a constraint on lending (which I believe is wrong as I have already said many times).
Some readers already know the simple bank accounting profit equation I referred to in a couple of previous posts:
Net Profit = Interest Income from lending – Interest Expense from deposits – Operational Costs
Let’s now move on to a slightly more complex version.
In the short-term, banks survive by making accounting profits. However, in the long-term, banks survive by making economic profits (= at least covering their cost of capital). And, unlike accounting-based financial statements, the cost of capital incorporates risk.
The risk I am particularly interested in today, and which is the most relevant for the Tobin/Yeager debate is liquidity risk. Let’s now modify the profit equation using White’s The Theory of Monetary Institutions. A bank’s economic profit equation thus becomes:
Economic Profit = II – IE – OC – Q, where Q represents liquidity cost.
Liquidity cost isn’t a ‘tangible’ cost and is therefore excluded from a bank’s accounting profit. Nevertheless, the less liquid a bank becomes, the riskier it becomes, and investors demand as a result appropriate compensation for bearing that extra risk, reflected in a higher demanded return on capital. Eventually, liquidity cost can also impact accounting profitability through higher cost of funding (i.e. interest expense).
Why I am referring to liquidity risk? Because liquidity risk becomes a direct economic constraint on bank expansion. Individual banks maintain a liquidity buffer to face redemptions and interbank settlements. The traditional view is that reserves play this role, through the exogenously-defined reserve requirements. Without them, banks would turn ‘wild’ and become way too risky. This view is inaccurate.
Indeed, banks in free banking systems during the 19th century used to maintain a high level of reserves, despite the absence of reserve requirements (though this level used to vary with technological advancement and demand for banks’ liabilities). With the development of capital markets, banks found another solution: swap some of their reserve holdings for highly-marketable and high quality securities, which effectively became reserve-replacements, or claims on reserves.
In order to maximise economic profitability, modern banks now usually maintain their (primary) reserves to a minimum while also holding secondary reserves (that is, in a non-interest on excess reserves world). In a way, banks are never fully loaned-up. From a primary reserves point of view, they indeed are. From a total reserves point of view, they are not. Banks sacrifice higher yields on loans for lower yields on safe and liquid securities. However, I have to admit that, from a risk-adjusted point of view, banks could be considered fully loaned-up*.
Seen that way, Tobin was right to believe that economic constraints would prevent bank expansion, but was wrong in believing that reserve requirements merely prevented the expansion from reaching its natural economic limits and that banks’ special status was only derived from this legal restriction**:
In a régime of reserve requirements, the limit which they impose normally cuts the expansion short of this competitive equilibrium. […] In this sense it is more accurate to attribute the special place of banks among intermediaries to the legal restrictions to which banks alone are subjected than to attribute these restrictions to the special character of bank liabilities.
If anything, economic constraints kick in before reserve requirements***, and internally-defined reserve requirements are part of economic constraints.
On the other hand, Yeager sometimes seems to overstate the effects of reserve requirements and underplay economic constraints in limiting expansion:
Proponents of this view are evidently not attributing “the natural economic limit” to limitation of base money and to a finite money multiplier, for that would be old stuff and not a new view. Those familiar limitations operate on the supply-of-money side, while the New Viewers emphasize limitations on the demand side.
To be fair, Yeager’s exact point of view isn’t entirely clear in his article. He seems to reject the ‘natural economic limits’ only to later endorse them, though this might be due to the facts he does not view liquidity cost as an economic constraint in his reasoning:
It is hard to imagine why a bank might find it more profitable to hold reserves in excess of what the law and prudence call for than to buy riskless short-term securities with them.
I also partially disagree with Yeager’s point that “the real marginal cost of expanding the system’s nominal size is essentially zero.” This is true… in the long run. But in the short-term extending credit often involves growing operational costs (i.e. hiring further loan officers for example), which weigh on banks’ accounting profitability in real terms, and growing liquidity costs, which weigh on banks’ economic profitability, before the nominal size of the system is finally expanded.
I might be missing something, but the ‘hot potato’ effect of banks’ inside money and the natural economic constraints on banks expansion do not look irreconcilable to me. This is also the view that White seems to take in his book. Furthermore, only this view seems to be able to explain the behaviour of free banks. Banks loan up to a degree that maximise economic profitability (which includes safety/risk) but any exogenous increase in reserves can also bring about an expansion of banks’ inside money (money multiplier effect) that results in a new nominal paradigm.
It’s funny but I have the feeling that this debate will make me think for much longer.
** In addition, the margin between interest income and interest expense isn’t the only way to generate revenues. In fact, many banks accept to extend credit to customers at a loss, in order to generate profits through cross-selling of other financial products (derivatives, insurance, clearing and custody services…). This effectively pushes back the economic limits of expansion (without taking account liquidity cost). A traditional (i.e. non-IB) bank’s economic profit equation would then become:
Economic Profit = II – IE + CS – OC – Q, where CS represents revenues from cross-selling.
*** Some of you might believe that this is in contradiction with my claim that reserve requirement policies have been successfully implemented in various countries. It isn’t. Primary reserves that are not in excess are essentially ‘stuck’ at the central bank. As banks fully ‘loan-up’ on them, increasing reserve requirements merely reduce lending expansion unless banks decide to slash their secondary reserves (which essentially replace excess reserves), which would in turn increase liquidity cost.