Fiat currencies trade at discounts to par: Corollary 1 – Any expansion of the Fed’s balance sheet dilutes the dollar
I’ll discuss an interesting corollary of the fact that fiat currencies trade at discounts to par; namely, that any expansion in the balance sheet of (say) the Fed necessarily entails ‘dollar dilution’. That is, each dollar becomes ‘good for’ less stuff.
In case some people are thinking; ‘Duh! Money printing obviously dilutes the dollar!’, I should point out that there exists a large group of financial commentators that believe that QE is a (somewhat benign) asset swap. I have misgivings with this notion; my understanding of currencies flies in the face of this argument.
The brief intuitive explanation is this: the discount from par means that the Fed can’t go into the market to purchase something and increase their assets as much as they increase their liabilities.
But what does this really mean?
The above can be conveyed with a simple example. Suppose a 17th century goldsmith issues $100 against the 100 ounces of gold in his vaults. The currency would be redeemable, so anyone with a $1 note would be able to get 1 ounce of gold for his note at any time he pleases. Let’s suppose that this goldsmith engages in a debauched practice; he uses the gold in his possession for speculative activities. [Huerta de Soto describes why this practice is debauched in Money, Bank Credit and Economic Cycles]
Say that his speculative activities prove unsuccessful and he loses 50 ounces of gold in his possession. This means that only 50 ounces of gold back the $100 in circulation, and somebody is going to have to take a haircut. The goldsmith has several options:
1) He continues redeeming, and eventually is shut down by a bank run. So, the early-comers salvage their gold whereas the late-comers lose everything. The goldsmith ends up closing his doors.
2) He revalues the dollar. Meaning, he says that $1 can henceforth be redeemed for 0.5 ounces of gold rather than 1 ounce. The haircut is delivered uniformly across depositors in this scenario. The goldsmith would likely end up closing his doors.
3) He debases the underlying gold in his possession, hoping that depositors won’t notice. For example, he might mix copper with gold and pretend that he has 100 ounces of gold. The goldsmith would likely end up closing his doors.
4) He could stop redeeming in gold (with or without government decrees on his side).
We’ll consider 4), because it describes the essence of a fiat currency. Fiat currencies are the carcasses of formerly redeemable currencies.
I contend that – upon stopping redemption – the $ price of gold would eventually rise to at least $2/ounce. Why? Well, firstly there are only 50 ounces of gold to back $100. This implies that the maximum $1 is immediately ‘good for’ is 0.5 ounces of gold, and hence the maximum $2 is immediately ‘good for’ is 1 ounce of gold ($2/ounce). But that is not all; you can no longer redeem anything upon demand. There is a time component involved; owning a $1 note is now some potential claim on some quantity of gold in the future (if at all). So there would be a discount involved with this time component on the market. Furthermore, the question of the goldsmith’s actions and behavior might be discounted on the market (questions like: will the goldsmith lose any more of the gold in his possession? Will the goldsmith print dollars? etc).
All of this would imply that dollars would trade with gold on the market at over $2/ounce. For this example, let’s suppose that gold trades at $2.5/ounce.
What happens when the goldsmith tries to expand his balance sheet?
Let’s see what happens when the goldsmith tries to print dollars to put 50 ounces of gold onto his balance sheet:
Using the heroic assumption that the market price of gold stays at $2.5/ounce. He would have to print $125 to buy 50 ounces of gold. This would mean that he owned 100 ounces of gold, but had $225 circulating against it. Do you see where I’m going with this?
He started off with 50 ounces of gold backing $100, and now has 100 ounces of gold backing $225. Previously, the maximum $1 was ‘good for’ was 0.5 ounces, now the maximum $1 is ‘good for’ is 0.44 ounces of gold. In other words, the gold price would have been at least $2 before, now it has to be at least $2.25!
We can approach the reality of the Federal Reserve by noting that our heroic assumption does not hold. There is a colossal industry that is in the business of watching the Federal Reserve. The Fed announces what it’s going to do before it does it, and indeed, it is the image of public sway and consensual ‘logic’; thus it is anticipatable. What I’m getting at is that if the goldsmith were to go to market to purchase 50 ounces of gold with printed dollars, he would likely have to pay in excess of the $2.5 previously traded. His actions have a dilutive effect on those dollars, and so the market would move to anticipate his actions. This would, in turn, make his actions all the more dilutive.
We arrive at a golden opportunity for the speculator; a self-sustaining trend. If the Fed plans to increase their balance sheet, that will entail dollar dilution (by definition). Since the Fed is anticipatable, this dilution will be factored in by markets. This process of factoring-in the Fed’s purchases will – in itself – increase the dilutive effects of the Fed’s purchases. These increased effects will be factored in by markets, and so on… This is useful to the speculator because it represents a left-field self-reinforcing trend that may seem paradoxical to some (NOTE: the self-sustaining part works the other way also: when markets discount less dilutive purchases, this entails less dilutive purchases,…). The paradoxical short-term trade is this: if you believe that the Fed is going to print money (consensually believed to be inflationary), then you want to be levered and long the bond market!
Applying the above to recent events:
First, let’s take a look at the balance sheet of the Fed to see what a dollar is ‘good for’ these days:
Glancing at this, we can see that the dollar is ‘good for’ – by and large – long term treasury notes and bonds, mortgage backed securities, and some things like gold.
Looking at some bond charts below, we see that the prices of US 10 year notes and 30 year bonds spiked up in anticipation of the QEs. These spikes were followed by equally dramatic collapses in prices. Incidentally, the recent QE2 extreme on the upside (October 2010) was corroborated by Jim Rogers coming in to buy the dollar. As far as I can tell, we’re in the process of squeezing out short dollar, long things (bonds, gold, equities, EM etc.) positions. However, because of the enormous mountain of IOU money (credit, demand deposits etc) that can go up in smoke, this could turn into something deadly. Every time the Fed acts to induce ‘inflation trades’, the implication is that this long-term consensual position becomes even more crowded. When trades become profoundly crowded, every downtick has the potential to induce mass selling. If long-term ‘short dollar long things’ positions get squeezed, a widespread credit collapse could ensue.
Recommended: Charting the Federal Reserve's Assets - 1915 to 2012