A dollar-denominated global economy means the Fed is at once the bankers bank and government bank, as well as both U.S. central bank and global central bank — managing that hybrid is the challenge of our time
Today global money is largely private credit money, the issue of a profit-seeking bank that promises ultimate payment in public money which is the issue of some state, quite possibly a different state from the one where the bank is chartered and does its business. Global money is also largely dollar-denominated, even when the ultimate users of that money lie completely outside the United States. The issue of dollar-denominated US Treasury bonds is just part of the huge stock of dollar assets and liabilities; the stuff of dollar hegemony is the private credit money dollar, not the issue of the state.
Although global money is substantially private credit money, the fact that it is denominated in dollars means that the Fed is de facto, if not de jure, the ultimate lender of last resort for global money. Therein lies the rub. De facto the Fed’s responsibility is global but de jure its authority is only local. The Fed is essentially hybrid, both government bank and banker’s bank, and also both US central bank and global central bank. The great challenge of the present time is the politics of managing the hybrid reality of the global dollar system.
In the normal course of business, banks meet their gross settlement obligations substantially through offset, and then meet any net deficit by borrowing from other banks in global money markets. Normally this settlement process functions noiselessly and without intervention by any public authority. But breakdown or dysfunction in interbank money markets, as occurred between 2007 and 2009, send net deficit banks to their own individual national central banks as backstop. In 2008, these national central banks revived a network of borrowing arrangements, so-called liquidity swaps, to acquire needed means of ultimate global payment for their client banks. It is the network of central bank liquidity swaps, centered on the Fed but functioning as a network, that today serves as global lender of last resort.
And it is the network of private foreign exchange swap markets that serves as global dealer of first resort, for a profit. In a recent speech at the World Bank, Hyun Song Shin of the BIS drew attention to two remarkable facts about that private system (summarized in his Graph 2). First, prices in the foreign exchange swap system fail to obey Covered Interest Parity. Second, the degree of deviation from the theoretical CIP ideal seems to be positively correlated with the exchange value of the dollar. Shin floats an idea about what might be causing these new facts, having to do with institutional investors hedging the dollar exposure in their asset portfolios in order to reduce mismatch with non-dollar pension and insurance liabilities. But a simpler, and more general, explanation seems to me also worth considering.
From a money view standpoint, deviation from CIP is simply the expected profit for private dealers that is required to entice them to take onto their own balance sheets the opposite side of market imbalances. The fact that global money is dollar private credit money means that the two facts Shin documents are actually symptoms of the same imbalance. Possibly hedging demand is an important source of today’s market imbalance, but other sources can also be imagined. The important point is that in today’s market the cross-currency basis swap operates as a sensitive barometer of market conditions, and central bank liquidity swaps operate to create the outside spread that puts bounds on how far price can deviate from CIP.
The origins of modern central banking can be traced back to the early days of the international gold standard in the 1870s and Bagehot’s famous Lombard Street (1873). Following Bagehot, in time the Bank of England came self-consciously to embrace de jure responsibility as lender of last resort for its own national banking system. Bagehot never addressed the problem of lender of last resort for global money, perhaps because the international gold standard centered on London was at that time yet in its infancy. But it turns out he did know a thing or two about global money, and about the global money markets in which private dealers operated as global dealer of first resort. We can learn from him for our own time.
In 1869, Walter Bagehot published A Universal Money, a collection of his ruminations on the desirability and possibility of a truly global money. The occasion was the invitation, at the International Monetary Conference of 1867, for Britain to join the France-led Latin Monetary Union, formed in 1865 with Belgium, Switzerland, and Italy. The original idea of the Union was to create a common 5-franc silver coin to facilitate trade. The new idea for an expanded Union was to create a common 25-franc gold coin to expand the reach of the Union, not only to Britain but also the United States and Germany, among others.
Bagehot took this invitation quite seriously, in part because he expected Germany to join. “Before long all Europe, save England, will have one money, and England be left outstanding with another money.” But more generally Bagehot thinks a universal money would be a genuinely good thing, if only it were possible. He argues that the French proposal is not in fact workable–too top-down, disregarding facts on the ground–but proposes his own alternative road to universal money.
In 1869, Bagehot’s alternative proposal was for an Anglo-Saxon Monetary Union between Britain and the US, which he thought Germany might find more attractive than the Latin Union. His idea was to link the US $5 gold piece with the British pound, while at the same time decimalizing the British currency by treating the farthing (1/4 penny) as 1000th of a new pound. The future he envisaged was a world of two Unions, the Teutonic and the Latin, between which other countries would be free to choose whether to join, and which to join.
In that case, there would be one Teutonic money and one Latin money; the latter mostly confined to the West of Europe, and the former circulating through the world. Such a monetary state would be an immense improvement on the present. Yearly one nation after another would drop into the union which best suited it; and looking to the commercial activity of the Teutonic races, and the comparative torpor of the Latin races, no doubt the Teutonic money would be most frequently preferred. In this case, as in most, the stronger would daily come to be stronger, and the weaker daily be in comparison if not absolutely weaker. Probably in the end the less coinage would merge in the greater, but at any rate it would be a great step to have but two moneys, and we could well make shift to do with that if we were sure, as we should be, that there never were to be any more.
It didn’t happen that way of course, at least not right away. The dollar was then off gold, as a consequence of Civil War finance. And by the time it got back to gold, Britain was well under way on a different project, constructing the international gold standard around the undecimilized pound, using its imperial reach as the core. (Decimalization would finally take place only in 1971, a full century after Bagehot was writing!) Thus the Bank of England became de facto central bank of the world even if de jure only central bank of England.
For a while the pound was universal money, but only for a while. World War I, then world depression, then World War II would finally result in replacing the pound with the dollar as the center of the system, and seventy years later the result is what we see today. Today the dollar, not the pound, is global money. And its principal rival is not the French-based Latin Union as Bagehot anticipated, but rather the German-based Eurozone.
But leave all that history aside for the moment. For our purposes, the most interesting passage of Bagehot’s book is the one where he talks about the determination of foreign exchange rates. Note in the passage below that there is no hint of the abstract logic of covered interest parity; instead the story is about the price of bills of exchange as determined in dealer markets.
An exchange calculation is really the cost of remitting money from one country to another. That cost is substantially the same, whether the country from which the money is exported and the country to which it is imported have the same currencies or different currencies. Australia and England have the same currencies; the sovereign is the main coin in both; but, nevertheless, there is an expense in remitting money to Australia. The remitting banks make a charge for selling their drafts, and this is the common exchange calculation in a new shape. If France and America had the same currencies as England, it would still happen as now, that bills on Paris or New York would be at a discount or a premium. The amount of money wishing to go eastward across the Atlantic, and the amount wishing to go westward, would then as now settle how much was to be paid in London for bills on New York, and how much was to be paid in New York for bills on London. The original element in exchange transactions—the remittance of money—would remain as now, and the two principal accessory difficulties would be just as great. In practical exchange business the rate of interest is to be considered, and the state of credit also. If you buy a bill at three months’ date you lose a certain sum in interest, depending on the rate for the day, and you rely on the credit, more or less good, of the parties to the bill. These main peculiarities of exchange business are fixed by its nature, and no change of currency can alter them.
Bagehot is of course talking about bills that finance trade of actual goods, and the physical cost of remitting money. In our world, by contrast, trade of financial assets completely swamps trade of real goods, and in world of electronic transfer the substantial cost is not physical but rather the risk exposure from open positions. Today the most important money market instrument is not the real bill but rather the repo. Indeed, Pozsar’s recent Credit Suisse paper asserts that the overnight GCF repo market “is the only functioning money market left standing today”, and suggests that the Fed’s new reverse repo facility is one side of the current outside spread that bounds repo rates from below.
Taken together, Shin and Pozsar sketch a map of the emerging system of global money. It is a system of dealer markets making key prices–FX swap basis and GCF repo–and a network of central bank backstops for those prices–liquidity swaps and reverse repo facilities.
I give the last word to Bagehot who, in his own way, seems to have foreseen something like this moment:
The real objection is that after all this plan does not combine; it leaves us with two moneys; but if all the nations of the world gradually joined either the Latin coinage league or the Teutonic coinage league, trade would be very easy; and the amalgamation of these two might be left to a future and more educated age.
That’s us he is talking about. Our amalgamation is first about the C6, the top six central banks now joined in a the central bank swap network. It took a global financial crisis to get us to this point. The next step is bringing in the periphery, starting maybe with the BRICS. Hopefully we are educated enough to do that without requiring another global financial crisis!