Technically the title of this post is wrong: the truth is that nobody could possibly know or predict what a bank run would looks like in details suffice to say that it would have terminal and devastating results on the global economy. One needs only remember what happened when the Reserve Fund broke the buck and the $3 billion money market industry was at risk of unwinding (for those who do not, Paul Kanjorski does a good summary here). What we do, however, wish to demonstrate is the tenuous balance between physical money - yes, just like precious metals, there is actual "physical money", better known as currency in circulation - and more abstract, confidence-based, "electronic money." Now when it comes to talking about systemic instability, pundits often enjoy bringing up the case of the $600+ trillion (recently discussed here in a different capacity) in synthetic derivatives, whose implosion would "wipe out the world." While that may indeed be the case (the memory of the CDS-precipitated AIG implosion is still all too fresh), since nobody really can comprehend the side effects of the collapse of global derivative system, which by some estimates is over $1 quadrillion when combining exchange and OTC based derivatives, it is largely based on pure conjecture. And, as we demonstrate below, one doesn't even need to do get that high up in the pyramid of credit money. The truth is that should there be an American bank run, what would happen is the conversion of all electronic dollars into physical dollars, as retail Americans rush to empty their checking and savings accounts, exit their money markets, while institutional America converts all "shadow" liabilities into hard dollar assets (Zero Hedge has a specific methodology of defining what liabilities make up the shadow banking system). The truth is that should there be a D-Day in the American banking system and there is a global scramble for physical paper (ignore gold) the conversion ratio for binary dollars into hard ones could be as high as 30 to 1. Which begs the question: should one apply a 90% discount when evaluating their electronic dollar exposure? That, and many other questions too...
When looking at actual "hard" dollars, there is just one place: the Fed's weekly H.6 statement which shows what the total amount of currency in circulation at any given moment is. The H.6 is the statement that breaks down the two forms of monetary stock tracked by the Fed: M1 and M2. Currency is at the very top. As a reminder, currency, together with Fed bank reserves are the only two actual forms of money "printed" into circulation. Yes, there is much polemic over the nature of bank reserves, but they, together with currency in circulation are the only two actual liabilities on the Fed's balance sheet, backed by such assets as Treasurys, Mortgage Backed Securities and, questionably, gold (questionably, because as Ron Paul has been crusading, the existence of gold on the Fed's asset side is taken on faith, and is based on promises by the Fed that it in fact exists, but nobody is allowed to actually see it).
So how many actual physical dollars are there? Well according to the H.6, as of June 27, there was $967.3 billion in currency currently circulating within the US economy, while the H.4.1 tells us that as of July 6 there was $1.66 trillion in bank reserves with the Fed, which if need be can be promptly released as currency to the wider public on demand (granted the dynamics of this release are completely unclear).This adds up to just over $2.6 trillion in "physical currency" (which also happens to be the "record" asset side of the Fed's balance sheet).
So that's what the the 'supply' side of money looks like in a dollar bank run. What about the demand. In other words, who will have the non-contractual "right" to pursue these $2.6 trillion in cold, hard cash?
Let's start with the M1, which is where the first tranche of electronic dollars is situated.
M1, in addition to currency in circulation, also contains demand and checkable deposits. The most recent number for these two is $982.9 billion. So far so good: if only demand and checkable deposits were pulled, the currency in circulation would be sufficient, although there would be a small impairment of just about 1.5%.
Next up, we go to the M2, which in addition to the M1 components, also contains such abstract concept as Savings Deposits, Small-Denomination Time Deposits and Retail Money Funds. The dollar values associated with these assorted claims on cash are $5,662.8 billion, $827.9 billion, and $698.7 billion respectively, or a total of just under $7.2 trillion. Add to this the roughly $1 trillion in non-cash M1 and you get $8.2 trillion. And this is where things start getting interesting. Because should every retail saver who has documented paper claims in America's checking, savings, time-deposits and money market pull their money, they would find that there is just $2.6 trillion in cash available to actually satisfy said claims.
But wait, there's more.
While the M2 conveniently ignores it, another major component of monetary aggregates is institutional money funds, which adds another $1.833.2 trillion in claims to physical fiat. Added across and we get just over $10 trillion.
But wait, there's more.
Remember how on March 23, 2006 the Fed discontinued the M3 because it was "too expensive" to keep track of all this "money." Well, courtesy of various replication loophole Zero Hedge has been able to track a far more comprehensive indicator of the broadest money stock in the US economy: the shadow banking system, which for all intents and purposes is the same as above: namely claims on actual money however more by institutional accounts than retail.
The breakdown, based on the most recent Z.1 (through March 31, 2011) is as follows:
- GSE Liabilities: $6,577.8 billion
- Agency Mortgage Pools: $1,166.3 billion
- Asset-backed securities Issues: $2,280.6 billion
- Securities Loaned by Funding Corporations: $709.0 billion
- Liabilities in Fed Funds and Security Repo agreements: $1,263.3 billion
- Total Outstanding Open Market Paper: $1,131.2 billion
Whipping out the calculator, and we get... $13.1 trillion in shadow banking system claims. Adding across with the M1 and M2 stock noted above and one gets $23.1 trillion. As a quick reminder, the physical money, in a best case scenario, is $2.6 trillion when adding reserves, and in a worst case, $967 billion. In other words, the paper to physical dilution is anywhere between 8.8 times and 24 times.
But wait, there's more.
Observant readers will recall that in our 2009 piece which before anyone else had even considered it, explained how the Fed bailed out the world with FX liquidity swaps, one of the key take home messages was that there was a synthetic short on the USD to the tune of about $6.5 trillion courtesy of the USD carry trade and other considerations. In other words, this is how many dollars would have to be conjured up into existence to satisfy existing electronic claims (and why the Fed had to scramble to implement the FX swaps when it did). One thing that is certain is that in an American (and thus global) bank run, all of the dollar shorts would cover in milliseconds as the carry trade would collapse instantaneously.
The take home is that courtesy of this latest and greatest demand on cash, there is up to another $6.5 trillion in potential claims on underlying hard dollars (and likely much greater as this BIS study was conducted at a time before ZIRP, and before the USD was the new, step aside JPY, carry currency of the world).
Summing it all up
Putting together all of the above, there are anywhere between $967.3 billion and $2.6 trillion in physical claim satisfying pieces of paper which everyone would scramble to grab if the sky was falling, and against these there are just under $30 trillion in paper claims on said hard paper. This can be seen visually on the indicative chart below:
Do readers see now why it is irrelevant to add X trillions or even quadrillions in derivatives? Because when just taking the plain vanilla electronic claims on circulating dollars there would have to be between a 11x and 31x haircut when everyone rushes to procure the suddenly all too precious pieces of paper with the picture of a dead president on the face.
For all intents and purposes this has been more of a thought experiment than any indicative scientific evaluation as there are many other nuances when analyzing all of the above. However, for the sake of esthetic purity, the truth is that no matter how one slices and dices it, there will be an unimaginable scramble to get out of electronic dollars and into physical ones. The amusing thing is that there are many who are worried that physical silver claims may be diluted by outstanding paper. This is true, however, ironically, it is very true when dealing with the heart of the fiat system. And recall that we refused to look at the $1 quadrillion in credit money at the derivative level. We believe that for illustration purposes, knowing that at best 10% of electronic money is covered in a worst case scenario should be sufficiently enlightening. As for those who say that all the Fed would need to do is merely hit the print button and not stop, remember: this money would simply flow to bank reserves. How it gets from there to outright currency in circulation is something the Fed has been bashing its head over the past 3 years, so far, unsuccessfully. And any money paradropped into circulation directly, would not do anything to alleviate the dilution factor as it would add to both sides of the "claim" and "deliverable" ledger (not to mention that it would also leads to instantaneous hyperinflation).
So, in the loosely paraphrased immortal words of Troy Mclure, now that you know, roughly, what a bank run would look like, don't do it.