The Full Monty on Naked Short Selling

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).

http://jpkoning.blogspot.ca/2014/09/getting-naked-in-praise-of-naked-short.html

“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.

About

View all posts by
Guest
brookside
2 years 7 months ago

A “complete” naked short sale is impossible. If the short seller does not deliver the securities to the buyer within the settlement period (3 days for stocks) the trade is a “fail” and the positions of the seller and the buyer are cancelled. If a bank is really doing a “naked short sale” when it makes a loan as Mr. Koning contends then the loan would never be funded. I am confused by Mr. Koning discussion currency and demand deposits but I think it gets down to saying that a demand deposit is not money.

Admin
2 years 7 months ago

In futures contracts it is entirely possible to initiate a position by selling, and hold this open position even though you do not own the underlying. Every futures contract requires a long and a short.

Even though a true naked short is not possible in our equity markets, the concept of a naked short is extremely useful, and useful in this discussion.

Guest
AH
2 years 7 months ago

You could make a synthetic naked short by buying a put and selling an option at the same strike price and date.

There is a relationship between shorts, forwards and interest rates because for covered interest parity to hold, people need to be able to short the underlying in arbitrage arrangements. It seems there should be something interesting to say about this, but I havn’t quite figured it out.

Guest
2 years 7 months ago

Hah, love the title. I’m going to read this over later, hopefully I can get caught up to the conversation by Wednesday.

Guest
JKH
2 years 7 months ago

One aspect not yet discussed is that of the optionality inherent in a demand deposit. Anyway you look at it, the bank is not really short CB notes and TR coins. It’s short a call option on those CB liability forms. The strike price is nominal equivalence and the nominal intrinsic and time values are zero. But the optionality is pronounced because the conversion rate would be pretty low.

This contrasts with a short position in Microsoft stock where the obligation is fixed rather than option dependent.

Guest
JKH
2 years 7 months ago

Hope so JP.

I’m not clear on all this yet.

But I find it very intriguing.

I basically wrote this post to prevent my head from exploding.

So the thought process at this point is somewhat “short” of pristine.

🙂

Guest
2 years 7 months ago

Yes, I think I see where you’re coming from. This post brings up memories of the discussion over Tobin’s Widow’s Cruse paper.

I do feel like reminding folks that the main point of my post wasn’t about banking per se, but about the stock market and the “evils” of naked short selling. If your ‘counterfactual’ and ‘factual’ amount to the same thing, naked short selling is still just “something that banks do to”… and therefore anyone who wants to maintain an anti naked short selling of stocks stance needs to to argue against banking, at least if they want to be consistent.

The fact that there is no difference between banks making loans by writing up a new deposit and making loans in the form of currency/reserves would seem to indicate that the MMT criticsm that ‘banks don’t lend out reserves’ is really not that important. Of course banks don’t lend out reserves, but we can think of them as if they were lending out reserves since the final resting point is the same.

Guest
JKH
2 years 7 months ago

One area where I disagree with the language of your post is the (one?) reference to shorting deposits. It’s a CB liability that is being shorted in the context of the post. Bank deposit liabilities are evidence of borrowing. Banks don’t borrow an evidence of borrowing already on their books and they don’t promise naked to deliver something they’ve already delivered.

Guest
JKH
2 years 7 months ago

I think that was a one time slip but it’s indicative of a difficult paradigm to apply to self-issued liabilities.

Guest
JKH
2 years 7 months ago

“The fact that there is no difference between banks making loans by writing up a new deposit and making loans in the form of currency/reserves would seem to indicate that the MMT criticsm that ‘banks don’t lend out reserves’ is really not that important. Of course banks don’t lend out reserves, but we can think of them as if they were lending out reserves since the final resting point is the same.”

A rabbit hole I fell into thinking this through. It’s a point that Nick R. makes from time to time. But the reserve form needs to be CB notes and Treasury coins for the point to hold.

Guest
2 years 7 months ago

“But the reserve form needs to be CB notes and Treasury coins for the point to hold.”

Why is that exactly?

Guest
JKH
2 years 7 months ago

Non-bank customers don’t hold CB balances apart from notes or coins directly or separately from the banks (or at all), and therefore can’t deposit them with banks with the effect of increasing system deposits – as happens when banks create loans and deposits simultaneously

Guest
2 years 7 months ago

Right, I thought you were going to say something like that. Agreed.

So we know that banks don’t actually lend out reserves, and we know that non-bank customers can’t actually hold central bank reserves, but we can’t go wrong assuming these things are true since we arrive at the same resting point as we would if we were to work out how the actual banking system functions. (In the same way that we needn’t analyze a naked short sale of stock any differently than a straight short sale).

Guest
JKH
2 years 7 months ago

Right

Endogenous rocks – even in the face of exogenous intrusion

🙂

Guest
2 years 7 months ago

So the traditional econ 101 analytical shortcut taken by economists (like Scott Sumner and Paul Krugman), whereby banks are mere intermediaries, lending out customer’s deposits — while not an accurate description of banking — is “good enough”, since in the end it aligns with the the more accurate view that “loans create deposits”? What about all the hard work being done to understand the intricacies and fine detail of central banking and commercial banking? Is it worth the effort?

Guest
Fed Up
2 years 6 months ago

“So the traditional econ 101 analytical shortcut taken by economists (like Scott Sumner and Paul Krugman), whereby banks are mere intermediaries, lending out customer’s deposits — while not an accurate description of banking — is “good enough”, since in the end it aligns with the the more accurate view that “loans create deposits”?”

Is that only true if the capital requirement is 100% or the effective capital requirement is 100%?

Guest
JKH
2 years 7 months ago

I don’t think it’s good enough. I would compare it to Paul Krugman’s reverse engineering of I S L M to describe the world. It’s doable, but it’s not ideal on a stand alone basis.

The intricacies consist almost entirely of basic accounting and corresponding operations. That’s where the core logic is – like the operating system of a computer. I think it’s interesting because you can confirm whether you’ve got it right a lot of the time. Then it’s a matter of slapping some behavioral functions on top of that mechanical foundation. That’s what Godley and Lavoie do.

Maybe economics as a field of thought just hasn’t gone digital yet. There’s a glut of free form thinking that is disembodied from what should be it’s foundations. The field should be the target of a reverse takeover by accounting and finance (factual rather than academic finance).

🙂

Guest
Nick Rowe
2 years 7 months ago

JKH: I am not 100% sure I am following you here. But let me try this:

1. I can’t see any (relevant economically meaningful) difference between naked shorting a microsoft stock and covered shorting a microsoft stock. In both cases one more unit of microsoft stock-equivalent has been created.

2. I can’t see any (relevant economically meaningful) difference between banks making loans by writing up a new deposit and making loans in the form of currency (if that currency then gets deposited at another bank). In both cases one more unit of currency-equivalent has been created.

3. 1 and 2 are the same.

4. Are you saying the same as me? (I think you are, but I’m not sure.)

Guest
JKH
2 years 7 months ago

Thanks for having a look at this Nick.

1. I stayed away from the case of Microsoft in what I wrote. I know you made that point originally at JP’s post and I think I agree with it.

2. I agree. This is the part where I recall the kinds of things you’ve written about the endogenous money theme. You’ve stated that part more succinctly than I have obviously. My point is that it doesn’t make a lot of sense to hypothesize a counterfactual system where all commercial transactions must take place using CB money – when that requirement seems to be superfluous to what is already achieved in the existing system.

3. I’m not sure on that one Nick. What I did say was that the idea of naked shorting is a derivative of regular shorting (i.e. borrowing and selling). And because regular shorting didn’t make a lot of sense to me in the banking context (my point in 2. above), its derivative of naked shorting didn’t make a lot of sense either. But having said that, what you say about connecting those two things is probably true and comparable to the Microsoft case. It’s just that the regular shorting of Microsoft stock occurs within a coherent existing monetary system framework, whereas I find the regular shorting hypothesis involving CB currency and bank deposit liabilities is a rather weird suggestion in a superfluous, non-real world way – as I described in 2. above.

Guest
2 years 7 months ago

JKH: I think we might be roughly on the same page. My mind is not clear yet.

I think the banking counterpart of regular shorting is this: the bank borrows currency (accepts a deposit) and then lends that currency (makes a loan by giving currency to the borrower).

And naked shorting is where the bank makes a loan by creating a deposit.

And (as I have said in the past) the same amount of deposits get created either way, ultimately, so it makes no difference, except maybe under a microscope.

But in banking, currency and deposits are not perfect substitutes. Sometimes currency is safer and more convenient, and sometimes deposits are safer and more convenient. I don’t know enough about real-world finance to know whether “synthetic” microsoft stock is a perfect substitute for “real” microsoft stock.

Guest
JKH
2 years 7 months ago

Nick,

I think I agree with all that

Although one nitpick –

A bank can lend currency without borrowing it first.

That was one of my points.

The idea that in a counterfactual system a bank can only create deposits by exchanging them for currency doesn’t mean the bank can’t lend first and then find the required inflow of currency afterward. There is no religion to what comes first in bank asset liability management and money market and reserve management operations.

So the bank can create the asset (loan) by giving out currency for it (from reserves) – without borrowing the currency first. And that’s eerily similar to creating a deposit from a loan.

A bank that does that is naked between the timing of the loan and the timing of covering the required currency inflow.

That gets into the difference between trade date timing and settlement date timing, which I noted in my comments at JP’s post. And I think that’s tied into this aspect of naked short selling more generally.

Guest
yhvd
2 years 7 months ago

One clear difference between the cases that I can see is that in the case of real stocks there can be a short squeeze were people that are short can be required to find actual stocks to exit the (time limited) short position. If there is less actual stock available than actual stock, then the naked short holders will be forced to buy back the stock at *any* price.

See the Porsche-Volkswagen example for a pretty dramatic example of this.

I don’t see how a true short squeeze would be possible for banking operations.

Guest
Nick Rowe
2 years 7 months ago

yhvd: A bank run is like a short squeeze. But as Tom says below, having the central bank step in as lender of last resort makes the commercial bank safe from a short squeeze. It is like Microsoft issuing new shares, to help the short sellers.

Guest
Fed Up
2 years 7 months ago

“It is like Microsoft issuing new shares, to help the short sellers.”

There is a big point here. MSFT is almost certainly not going to do that. The central bank is expected to increase currency if needed as long as the commercial bank(s) is/are solvent.

Guest
JKH
2 years 7 months ago

i.e. The case for analogizing deposit banking as naked shorting is based on the absence of a counterfactual (regular shorting of CB currency in deposit banking) that doesn’t make a lot of sense to me as a plausible counterfactual. That’s obviously not the case when contrasting naked shorting and regular shorting of Microsoft stock.

wpDiscuz