Supply-Side University Lesson #10
Memo To: Website Students
From: Jude Wanniski
Re: Fixing Exchange Rates
This lesson is similar to one I presented almost a year ago. Those of you who remember it, will benefit this second time around, especially because it has just as much relevance today and is critically important in any discussion about monetary policy. The Europeans are only weeks away from fixing their independent monetary systems to a single currency, the euro. The United States remains in the same monetary deflation we described a year ago. And I continue to believe it may be necessary for a generalized fixing of currencies next year in preparation for the Year 2000 computer problem (Y2K). This is on the assumption that the only reason the world financial system does as well as it does with floating exchange rates is that supercomputers are able to manage the chaos.
We left the fixed exchange-rate system provided by the Bretton Woods gold standard in 1967-71 because the professional economists erred in their policy advice on how to maintain the dollar/gold peg. It was believed that it only could be done by raising interest rates sharply, which the political leaders believed would result in a recession and cost President Lyndon Johnson (in 1968) his re-election and President Richard Nixon (in 1972) his re-election. It was and still is taught that a country can improve its trade balance by altering its exchange rate with another country. There are very few economists who understand the issue, because they have not taught the subject or thought about it.
To begin with, let's be clear on what an exchange rate is. We normally think of the dollar/yen exchange rate as the number of yen it takes to exchange for a dollar or of the dollar/peso exchange rate as the number of pesos it takes to exchange for a dollar. For the purpose of this discussion, let's start by thinking of the exchange rate of an apple to an orange. If the price of an apple is an orange, the exchange rate is "one" whether you ask for the apple/orange rate or the orange/apple rate. If it takes two apples to buy an orange, the apple/orange rate is 2 and the irange/apple rate is 0.5. If you introduce the concept of "money" into this system, then it will always be true that the dollar amount or the yen mount or the peso price of an apple will usually be half that of an orange. Why usually? Because if you are buying apples or oranges in Tokyo, they come from different places and cost more or less because of "the cost of carriage." If apples keep better in transport from California than oranges, the exchange rate might change to the final consumer as a result.
Still, if you are told that the dollar/yen exchange rate is 120, this does not mean that a dollar is "better" than the yen as a unit of account. It only means it takes 120 units of yen to exchange for one unit of dollar. When I was a boy, I always thought the British pound inherently was somehow superior to a dollar, because it took five dollars to buy a pound. In fact, for an American tourist, the dollar in the postwar years was far superior to the pound even at that fixed exchange rate. Internationally traded 5 goods that had equivalent "costs of carriage" would be equally priced to the wholesale consumer in New York or London. But the people were so impoverished they were pricing their own services at low levels. In other words, a fixed exchange rate doesn't mean prices will be equal in all things. A New York dollar is the same as a Nevada dollar when it comes to buying a plane ticket to Brazil, but for most goods and services the dollar goes a lot further in Nevada because taxes are much lower and the cost of real estate lower still.
To understand these distinctions by thinking of an apple/orange exchange rate, it is easier to see that if you are going to change the rate at which apples and oranges exchange, it is not likely you can do so by changing the rate of interest. We here must remember that an interest rate is the price of credit, not the price of money. The price of money is its purchasing power. By that I mean the price of buying a dollar is an orange, or perhaps two apples. Money is an asset used to acquire goods and services in real time, right now. Credit is an asset used to acquire money. These are critical differences. We can see that changing the interest rate does not create money and thus has nothing to do with changing an exchange rate between apples and oranges. The terms of trade can be changed between apples and oranges by a failed apple crop or a freeze of orange trees. It can't be changed by changing the price of credit or by changing the purchasing power of money. When the dollar was losing its purchasing power against gold in 1967-71 it was because there was more liquidity in the banking system than was being demanded by the world of commerce. The only way to protect the dollar's purchasing power was to remove the surplus liquidity by selling interest-bearing debt.
In a demand model, there is only one way to "tighten" or "ease" monetary policy — and that is to raise or lower the one interest rate more or less controlled by the Federal Reserve — the overnight "federal funds" rate. The mechanism of control is the result of the need each bank in the Federal Reserve system has for closing out its books at the end of each business day. The Fed requires every bank to hold a percentage of its deposits in reserve — the reserve requirement. At the end of each day, some banks have more in reserve than they need, others are in deficit. They borrow from each other to even things up. If the Fed sets the funds rate at 5%, which is where it is today, the "open-market desk" in New York City will add reserves if the rate pushes up from 5% as the banks bid for reserves in order to close their books on target. They add liquid reserves by buying U.S. Treasury bonds, which the private banks hold as part of their reserves against deposits. This "creates money." The banks can actually take the liquid dollars on their books and buy actual cash from the U.S. Mint.
If the funds rate falls below 5%, the open-market desk (which buys or sells U.S. Treasury bonds in the open market for the Fed's account in the open market), sees this as a surplus of liquidity, over and above the needs of the banks. It sells bonds on the open market to the banks, taking in the reserves that pay no interest. This destroys money, in the sense that U.S. debt in its non-interest-bearing form is converted back into interest-bearing bonds. "Money"disappears and bonds appear. There is no place for it on the Fed's balance sheet, as it was also created out of thin air when the Fed bought the bonds from the banks in the first place. Do you see the power of the Fed and Alan Greenspan in particular? As a result of the Federal Reserve Act of 1913 it has the power to buy assets from the banks with a check book that has nothing behind it but the authority of Congress — which has the Constitutional power to coin money. In 1913, Congress gave this power to the Fed, but always could withdraw it or set new conditions on the exercise of that power.
To repeat: The Fed tightens or eases by raising or lowering the fed funds rate in order to expand or contract reserves, which makes it harder or easier for the banks to lend money to borrowers for one sort of investment or another.
In a supply model, this operating mechanism is extremely inefficient. Years ago, when it was first explained to me by a banking editor at the WSJournal, I thought of an automobile which is driven from the back seat by a driver with a remote control instead of hands on wheel, foot on brake. Instead of having 12 men and women on the Federal Open Market Committee sit around a table every six weeks trying to decide what the fed funds rate should be for the next six weeks, how much simpler it is to instruct the open-market desk to buy bonds when the price of gold seems to be falling and sell bonds when the price of gold seems to be rising. All that need be done is decide on what the dollar/gold price should be, and the FOMC need never meet again. When I explained this methodology to Paul Volcker when he was Fed chairman, sometime in the 1982-85 period, he said: "You want to turn me into a robot," and I said that was exactly what I wanted to do. Why should the whole world rely on a few men and women to determine the value of money by pondering an interest rate target — when it could let the whole world market make that decision? Is it better for 12 people to decide the laws of government, which is how monarchy or politburos worked. Or is it better to have the whole population in the broad marketplace make the decisions, as a democracy works?
In explaining how a gold standard works, Alexander Hamilton said it a different way in a letter to the first U.S. Congress. He said if the bank issues more money than is demanded at the guaranteed gold price, the money will return upon the bank.
The stamping of paper money is an operation so much easier than the laying of taxes, that a government, in the practice of paper emissions, would rarely fail, in any such emergency, to indulge itself too far in the employment of that resource, to avoid, as much as possible, one less auspicious to present popularity. If it should not even be carried so far as to be rendered an absolute bubble, it would at least be likely to be extended to a degree which would occasion and inflated and artificial state of things, incompatible with the regular and prosperous course of the political economy.
Among other material differences between a paper currency, issued by the mere authority of Government, and one issued by a bank, payable in coin, is this: That, in the first case, there is no standard to which an appeal can be made, as to the quantity which will only satisfy, or which will surcharge the circulation; in the last, that standard results from the demand. If more should be issued than is necessary, it will return upon the bank.
Hamilton understood that at any given point in time, the population only requires so much liquid money, preferring to hold the rest of the national debt in bills and bonds that pay interest. In the United States today, if the gold price were set at $350, the Fed would pay no attention to the overnight fed funds rate. If it issued too much "money," i.e., bank reserves, the banks would never have to make what appear to be risky loans. The surplus money would in the first instance buy gold, putting the gold price nearer the ceiling of the acceptable gold band. If the band were five dollars on each side of $350, the open-market desk immediately would have to sell bonds as gold began to touch $355. This would make dollars scarce relative to gold and the gold price would retreat from its upper band. The market would determine the gold price between $345 and $355. If gold approached $345, the open-market desk would buy bonds, putting more dollars into the system and making gold scarce relative to dollars.
Remember we are talking about two methods of pegging an exchange rate: Raising or lowering interest rates, or expanding or contracting the central bank's balance sheet. The Fed's balance sheet expands when it buys bonds with its magic checkbook. It contracts when it sells bonds, the money it acquired disappears just as magically as it appeared. The Fed hates to contract its balance sheet. It hates to get smaller. And it always assumes that if it destroys "money," banks will become illiquid and fail. This is because a great many banks failed in the early 1930s when the Fed contracted its balance sheet in order to maintain the price of gold at $20.67 per ounce. Economists are horrified at repeating this "mistake," but in fact the banks failed because of the dramatic increase in tariffs and taxes in the Hoover administration. Milton Friedman to this day refuses to accept the tarifFtax argument I presented in The Way the World Works, preferring instead his argument that the Fed could have prevented the Depression if it had printed paper money when there was no demand for it relative to interest-bearing bonds.
In pegging an exchange rate, normally you think of an exchange rate as one currency exchanging for another. We are not talking about these exchange rates, but the exchange rate between the paper dollar and gold specie. The price of gold in dollars is an exchange rate. Three hundred and fifty dollars exchanges for one ounce of gold. It is not in the interest of Americans to have the Fed peg the yen rate, although it could, just as Thailand had been pegging its currency, the baht, to the dollar. For us to peg the dollar to the yen means we will give the Bank of Japan control over our currency. If the BoJ fouls up by letting the yen price of gold fall, we will be forced to deflate with it, by raising interest rates or selling bonds. And vice versa.
Alan Greenspan also has the blind spot on how to peg a currency to gold. In 1993, after several years of keeping the gold price close to $350 by pegging interest rates, gold climbed to $385 beginning in September of that year. As we observed at the time, the demand for liquidity had fallen as a result of the Clinton tax increases of 1993. Greenspan had warned that the Fed would not "accommodate11 the tax increases by lowering interest rates. Instead, he tried to drive down the gold price by raising overnight interest rates, which he did six times over a period of a year and a half. The gold price did not budge. Throughout the whole process, Polyconomics advised its clients that it was a hopeless endeavor to try to bring down the gold price and inflation expectations by raising interest rates. Only the selling of bonds to withdraw liquidity and make dollars scarce relative to gold would work. The Fed could not do this unless it changed its operating procedures.
Why wouldn't raising interest rates cause the gold price to fall? The gold price only would fall if there were a scarcity of liquidity — either because of increased demand relative to supply or reduced supply relative to demand. By raising interest rates, the Fed further reduced the demand for credit and the reduced supply simply matched that reduced demand. The very idea of raising rates to slow the economy is a perverse one, as it automatically changes the dynamic between the supply and demand for liquidity. When we were going through this experience in 1993-96, there was a conscious agreement among Fed and Treasury officials that it was necessary to slow the economy to prevent the return of inflation. Greenspan never considered the possibility that a draining of liquidity, a shrinking of the balance sheet, was an option. If he had, he still did not have an operating mechanism. All he had was interest rates, and every hike caused declines in the financial markets, but gold would not budge. Only when the prospect of a cut in the capital gains tax began to show itself after the November 1996 elections did it happen that there was an increase in the demand for liquidity.
When the Fed did not supply that demand, the price of gold began to decline. It is now roughly 25% below the level of November 1966, at $295 compared to $285. The Fed now has twice lowered the fed fimds rate by a quarter point, to 5% from 5.5%, and the price of gold remains unchanged. The lowering of the rate simply increases the demand for liquidity and the added liquidity simply matches the demand, where a surplus is required to produce a rise in the gold price and an end to the deflation.
In another clear example of pegging a currency, consider Mexico in 1993. The price of gold in pesos was rising and Mexicans and foreigners were scrambling to unload pesos in favor of dollars, which were rising, but at a lower rate. All the Bank of Mexico need have done was to sell peso bonds out of its portfolio, taking in and destroying pesos. Why wouldn't they do it? The government was afraid this would cause an increase in interest rates. Just before the devaluation, three-month peso notes were fetching 16% at an annual rate, nervously up from 12% a year earlier. Because of a fear of raising interest rates, then, they refused to sell peso assets bearing interest to mop up surplus pesos. And where did interest rates go on 3-month paper? For starters, 65% at an annual rate.
At the beginning of the lesson, we noted that the Bretton Woods system blew up in 1971 because the Federal Reserve did not even consider selling dollar bonds from its portfolio in order to withdraw surplus liquidity from the banking system. It was afraid interest rates would rise, slow the economy, and lead to the defeat of President Nixon in 1972. Instead, it ended the dollar's tie to gold at $35, which meant a decline in the purchasing power of the dollar as gold quickly jumped to $70. Interest rates rose, commodity prices climbed, but Nixon held other prices down with wage-and-price controls until he was safely re-elected. The inflation pressures could not be contained, especially after the oil-producing countries refused to accept devalued dollars for payment and quadrupled the oil price. Economists had predicted the devalued dollar would end the U.S. trade deficit with Japan, but the deficit went even higher. As classical, supply-side economists predicted, the terms of trade would not change by changing the value of the unit of account.
In the classical model, had the Federal Reserve sold bonds from its portfolio in 1971, making dollars scarce relative to gold, interest rates would have fallen, not climbed. The only economists in the world who were making this argument at the time were Robert Mundell and Arthur Laffer, both counted by their academic fellows as being eccentric. The argument is counter-intuitive, as it seems to make sense that if you take money out of the banking system, making it scarce, its price will rise. But that is exactly what we wanted to do, increasing its purchasing power in terms of real goods, first being gold. The interest rate, remember, is the price of credit, not the price of money. Even if short-term interest rates do rise, this would only reflect an increase in the demand for credit — which would strongly imply economic strength, not weakness.
In our lesson on exchange rates last year, we discussed Japan's puzzle. It has been in a long recession that began in 1990, when it increased the effective tax on capital gains. It now has lowered interest rates on overnight credit to 0.25% at an annual rate. Its gold price, we estimate, should be close to ¥44,000 per oz., yet it is closer to ¥34,000. Japan's economists make the grave mistake of confusing low interest rates with "easy money," when in fact they are having to starve the banking system of liquidity in order to keep interest rates that low. The dollar/yen exchange rate, in and of itself, tells us nothing about the relative positions of the two economies. It is today 123 yen per dollar. The number itself is meaningless. A supply-sider must know what is the gold price in dollars and what is the gold price in yen. We also must know what is the optimum gold price in dollars and the optimum gold price in yen -- where debtors and creditors are in balance ~ before we can tell you if the exchange rate is an appropriate one. Today, our estimate is that both central banks are making major errors in monetary policy and the exchange rate is the intersection of those errors.
The Federal Reserve meets this coming week to decide whether or not to cut the funds rate again or to leave it at 5%. Watch what happens to the gold price in dollars and the dollar/yen exchange rate. Also read carefully the Fed's statement on why it is doing what it is doing, as that will provide more information the market will digest as it rearranges the values of all assets. We'll review the action in next week's lesson. Questions of course are encouraged.
* * * * *