JP Koning has written an interesting post that depicts banks as engaging in “naked short selling” when they simultaneously create new loan assets and new deposit liabilities. For example, he compares the idea of systemic deposit creation with the case of selling Microsoft stock without owning the stock (shorting the stock) and also without borrowing the stock (naked shorting the stock).
“Just as an equity short seller will borrow and then sell Microsoft with the intention of repurchasing it at a better price, bankers borrow and then sell dollars with the intention of repurchasing them at a better price… In the previous paragraph you may have noticed that I described banking as the borrowing of depositors’ dollars in order to lend those dollars out. Because the dollars were borrowed prior to sale, this would qualify their activity as regular short selling, not naked short selling. But hold on, this isn’t at all how banks function. Bankers don’t wait for physical Federal Reserve dollars to be deposited by the public before selling them away… Banks are engaged in naked shorting pure and simple: they sell a financial instrument that they never actually had in their possession… How do they do this? The key here is that banks don’t actually sell Fed paper dollars short, rather, they sell dollar-linked IOUs (i.e. deposits) short.”
The post essentially says (I think) that banks engage in a sort of naked shorting exercise when they create deposits “ex nihilo” – that is without actually borrowing some form of funds in that process. Banks just issue new deposits to customers who take on new loans for example. Banks “acquire” a new loan asset without having “borrowed” “actual” “funds” to do so. So in a sense they have “sold” money (in exchange for a loan asset) without having borrowed it in the first place. Hence they are “naked short” in that sense. I think that’s the intended meaning, roughly.
An interesting question then is – naked short relative to what precisely?
Consider the counterfactual where bankers do wait for those dollars to be deposited before selling them away. According to the post, banks are naked short by comparison to such a counterfactual monetary system.
In this counterfactual system, it would seem there are at least two ways in which new deposits suddenly appear.
First, a bank customer can transfer or be the recipient of a transfer of money that already exists in deposit form in another bank. That transfer doesn’t change the level of system bank deposits. It happens through the bank reserve clearing system, and that is really no different than what happens today in the existing system.
Second, a customer can deposit central bank money in the form of banknotes. That DOES expand the level of system commercial bank deposits. And that is what would be different about this counterfactual monetary system. The level of system deposits in the counterfactual system – it would first appear – cannot expand in direct conjunction with new loan creation, but it can expand through new deposits of central bank money. This suggests a picture of central bank money circulating with a velocity that allows for banking system deposit expansion, assisted by a pace of central bank money expansion that boot straps that velocity. Conversely, in the monetary system that we actually have, customers who are granted new banking system deposits aren’t necessarily required to bring new funds into their bank to do so – in particular, a deposit customer doesn’t need to bring central bank money (banknotes) into a bank branch in order to receive a new deposit (i.e. a deposit that is both new to the bank and incremental to the level of already existing system bank deposits) – provided that customer can simultaneously borrow the funds from the bank.
But is this distinction between factual and counterfactual system actually meaningful?
The question at this point becomes how are new loans created in such a system? We know they don’t come into existence by simultaneous loan/deposit creation at a single bank.
One way a new loan can be made is by a commercial bank paying out central bank money (in the form of banknotes that it keeps in reserve) when it makes a loan.
What happens to that central bank money? What happens to those banknotes? Well, they are used in commerce until somebody deposits them back in a bank. And then they end up in that bank’s reserve account, along with other banknotes and reserve balances. Other things equal, the lending bank will have a reserve deficiency and the deposit taking bank will have a reserve excess. Those aberrations will be sorted out as necessary by each bank taking steps in the money markets in order to balance their reserve positions as desired.
But that is what typically happens in the existing system. The borrower will typically use the new funds in commerce and those funds will find their way to another bank. It is the same result. A new loan and a new deposit have appeared on the banking system balance sheet, and the result looks like the same type of “endogenous” money creation at the system level that already happens in the monetary system we actually have, once positions associated with the original loan and its associated deposit have been cleared. The counterfactual system simply skips the initial step in which a borrower has both a loan and a deposit with the same bank. But otherwise, endogenous money creation carries forward at a systemic level.
And what happens if the deposit that is created in this counterfactual system ends up landing back at the same bank that issued the new loan? Well, essentially that lending origination bank now has a new loan and a new deposit. The only difference from the case of the existing monetary system is that the loan and the deposit are issued to different customers now. But otherwise, the net balance sheet result is exactly the same, including the fact that the bank’s reserve account hasn’t changed on a net basis. A new loan and a new deposit have appeared on the banking system balance sheet, and in this case on the balance sheet of a single bank – similar to what happens in the existing system. So the net result looks like the same type of “endogenous” money creation that already happens in the monetary system we actually have at the level of a single bank.
Let’s take this one step further. For example, what is to prevent a borrowing customer in the counterfactual system from depositing central bank funds he receives from a new loan directly back as a deposit with the same bank he borrowed them from – even if temporarily? Is that going to be precluded as a counterfactual system constraint? This seems absurd.
So what purpose has been served by this counterfactual restriction whereby customers must deposit central bank money in order for system deposits to expand?
I see no constructive answer to that. This counterfactual system makes very little sense as a useful modification to the existing system.
A counterfactual system which precludes immediate loan/deposit expansion is really no different in effect from the existing system – even if you want to outlaw a borrowing customer depositing his own central bank money proceeds back with his commercial bank. Moreover, the idea that a bank with multiple, frequently transacting depositors needs to identify specific deposit monies before lending them out is a nonsensical impediment to standard bank liquidity management. Given the intra-day dynamics of money markets, the order of depositing and lending is largely irrelevant in this whole process of bank balance sheet management.
Conversely, the existing monetary system bypasses a questionable if not useless counterfactual constraint requiring all bank customers to operate with central bank money – by allowing the simultaneous creation of loans and deposits.
“Naked shorting” as a conceptual framework for banking only lives logically when “regular shorting” can be visualized by comparison. That context requires a counterfactual monetary system. And that counterfactual monetary system would seem to be a pointless tweaking of how the existing system operates. Indeed, “naked shorting” is fundamentally equivalent to “regular shorting” in the same context – as suggested above – just by visualizing a virtual flow of central bank money that connects the two key points of joint origination of balance sheet expansion – loans and deposits.
In summary, there is no problem in referring to bank assets as long positions and bank liabilities as short positions – in a relatively simple measurement paradigm of long and short gross balance sheet items from the perspective of the bank. (This is essentially the terminology of hedge funds.) An appropriate methodology (if desired) can then be chosen for netting such gross measures to a coherent summary of net “longness” or “shortness” for various types of exposures. At a high measurement level, a bank is net flat by virtue of a balanced balance sheet. But there are all sorts of ways of drilling down and becoming more granular in the description of effective net exposure. This is what hedge funds do in the case of third party positions in both assets and liabilities. Moreover, this is also the essence of bank asset-liability management across all types of risk – including liquidity risk, structural interest rate risk, structural foreign exchange risk, and all types of trading book market risks. It’s a matter of specifying what the measurement methodology is for gross and net exposures. But it seems to me that given the inevitable reality of endogenous money creation and the sensible clearing short cuts inherent in that process, the underlying “non-naked short” idea from which the idea of “naked short” must be derived is itself quite dubious, making the associated concept of “naked short” a somewhat questionable although very interesting characterization of bank deposit creation.