Saturday, October 16, 2010

QE2 and Fixed Exchange Rates

One avenue by which quantitative easing could raise money expenditures on domestic product is through a lower exchange rate. The lower exchange rate will tend to result in an expansion of exports and a decrease in imports. The decrease imports should be associated with an expansion in the demand for the products of import competing industries.

If the foreigners respond to the quantitative easing with their own quantitative easing, then exchange rates will not change. Quantitative easing will still raise purchases in each and every country, including purchases of foreign products. Both imports and exports would rise as real output and real income recover. (This, of course, assumes that real expenditures is less than the productive capacity of the economy, which is the primary reason to undertake quantitative easing. If real expenditures are equal to the productive capacity of the economy, then money expenditures are at the appropriate level.)

Suppose that rather than responding with their own quantitative easing, foreign central banks prevent their exchange rates from rising by accumulating assets denonominated in the currency issued by the central bank underaking the quantitative easing, while selling off other types of assets. F or example, imagine that China restricts new loans to Chinese firms, and as old loans are repaid, purchases U.S. Treasury securities. In other words, China generates whatever net capital outflow is needed to avoid any increase in its exchange rate.

Unlike the scenario where competitive quantitive easing expands demand all over, and so results in expansions in both exports and imports for each country, here the pathway by which a lower exchange rate expands demand is closed, and there is no foreign quantitative easing to expand exports. The only pathway by which quantitative easing expands money expenditures is domestic purchases on domestic product. This simply means that more quantitative easing--a larger expansion in the quantity of money--is necessary to return money expenditures to the desired level.

Another way to see the "problem," is that the efforts of the foreign countries to avoid exchage rate appreciation without expanding money expenditures in their own countries--that is, sterilization--involves a decrease in the natural interest rate for the country undertaking the quantitative easing. The net capital inflow generated by the foreign central bank(s) is an addition to saving, and requires a lower level of interest rates for total saving to be balanced with investment.

My view is that a clear committment to a prompt return to a reasonble growth path for money expenditures will raise the natural interest rate. Any increase in expected inflation would further raise the nominal interest rate consistent with any natural interest rate. And so, the most likely scenario is that efforts by foreign central banks to keep their currencies from appreciating would simply dampen the increase in the nominal market interest rates consistent with money expenditures returning to and then remaining on the targetted growth path.

However, if those effects are ignored, so that the additional quantity of money simultaneously involves an expansion in the total demand for bonds, then an alternative perspective would be that a larger increase in the quantity of money also entails a larger decrease in some type of nominal interest rate. If it is assumed that quantitative easing involves the central bank purchasing short and safe assets, such as T-bills, and further, that the foreign central banks are also purchasing T-bills to keep their currencies from appreciating, and still further, that the yield on those T-bills has been driven to approximately zero, then perhaps the foreign central banks might be blamed for the failure of quantitative easing. If only the foreign central banks didn't purchase all of those T-bills, then their yield would be above zero and quantitative easing would work. Of course, so would conventional interest rate targeting.

But proposals for quantitative easing usually involve the purchase of assets that are not so so short or safe, in particular, longer term government bonds whose yields are not close to zero. And so, the bottom line is that the efforts of the foreign central banks to keep their currencies from appreciating by purchasing some kind of assets requires that the central bank undertaking the quantitative easing purchase a larger quantity of longer term or riskier assets than otherwise would be needed. With longer term assets, like 10 year government bonds, at the very least bearing more interest rate risk, the efforts of the foreign central banks are requiring the central bank undertaking quantitative easing to bear more risk than would otherwise be necessary.

To me, the obvious solution is for the yield on money to fall, and if necessary, to become negative. Since I strongly favor a growth path for money expenditures consistent with a stable price level on average, negative nominal interest rates on money is the obvious answer. If foreign central banks want to accumulate large amounts of safe and short assets, then the nominal yield on such assets should perhaps be negative. And, of course, that creates problems with the issue of zero-interest hand-to-hand currency.

To put it simply, if the central bank promises to issue zero-interest hand-to-hand currency on demand, and it is committed to a rule for noninflationary growth of money expenditures, then, it is possible that it will have to bear, at the very least, interest rate risk by borrowing short (issuing zero interest currency) and lending long. If foreign central banks accumulate assets, then this problem is exacerbated. If they can be brow-beaten into allowing their currencies to appreciate, then this "problem" is less severe.

Friday, October 15, 2010

QE2 and Begger Your Neighbor

I favor quantitative easing combined with a target for the growth path of money expenditures, in particular, final sales of domestic product. The Fed has seemed poised to undertake some quantitative easing along with an at least temporarily higher target for inflation. While I oppose having any target for inflation, it seems likely to me that a prompt return to any reasonable growth path for money expenditures would be associated with a temporary increase in inflation.

It is also likely that quantitative easing and these expectations of inflation (or greater real growth) will cause a depreciation of the dollar on foreign exchange markets. That the dollar did depreciate in response to Federal Reserve discussions of quantitative easing and inflation targets is no surprise.

Ceteris paribus, one of the pathways by which quantitative easing can result in an increase in money expenditures on domestic product is through a decrease in the exchange rate. Imports become more expensive, which tends to raise domestic demand for import-competing goods and services. Exports become cheaper to foreigners, raising their demand for domestically -produced goods and services. The "cost of living" is likely to increase, since that includes the prices of imported goods, but the rate of inflation on domestically produced goods will only rise due to anticipated excess demands for those products. That, of course, is the possible unfortunate side effect of returning money expenditures to a target growth path.

Again, ceteris paribus, this pathway for expanding money expenditures on domestic product has an adverse impact on money expenditures on foreign goods and services. As foreign products become more expensive in dollar terms, fewer of them are purchased. The foreigners export less. Similarly, as exports expand due to lower prices in terms of foreign currencies, at least some of those sales come at the expense of foreign producers.

Suppose the foreigners "retaliate" by their own quantitative easing. The exchange rates don't change after all, and so the "ceteris paribus" condition doesn't hold. Changes in the prices of imports and exports, and decreased imports and expanded exports are not a means by which money expenditures rise in any country. Instead, excess money balances are spent on all goods and services (domestic and foreign) and so every country would tend to both import more as its residents spend on foreign goods as well as domestically produced goods, and export more as foreigners spend more too.

Now, if there are some countries that already have sufficient money expenditures, then they should not undertake quantitative easing to keep their exchange rate from rising. No, they should only do what is needed to keep their money expenditures on target. An increase in their exchange rate, resulting in reduced exports and greater imports would help prevent the development of excess money expenditures in their economy. It would help relieve inflationary pressures.

So, why the worries about "beggar your neighbor?" Suppose it is 100 years ago, and the typical country is on the gold standard and small relative to the world economy. The key goal of monetary policy is to keep the domestic paper money redeemable for gold. If gold is flowing into a country, there is no problem in maintaining currency at par. The only "problem" that might develop (keeping in mind that inflation, unemployment and everything else is irrelevant to the one golden goal of monetary policy,) is that gold may flow out of the country, gold reserves may disappear, and then redeeming paper money with gold may become impossible.

If this sad state of affairs should develop, then the obvious and practical response is raise interest rates. These higher interest rates should attract foreign short term investment (a net capital inflow) which will slow, stop, or reverse the outflow of gold. The goal of monetary policy can be maintained.

Of course, using monetary policy to raise interest rates tends to slow the economy and perhaps results in a recession--money expenditures grow more slowly or shirk, sales fall, and production and employment fall. This unfortunate side effect of the policy to maintain redeemability can help, because by slowing real growth, the demand for foreign goods will tend to fall and so reduce the gold outflow.

The long run solution to the problem however, is for money incomes, particularly wages, to grow more slowly, or perhaps even shrink. This will improve the competitiveness of domestic industry. The result is an export led recovery, because unit costs for export industries are lower compared to the rest of the world, and further, as economic recovery leads to growing demand, less of that demand will be for imported goods because, again, the lower unit costs provides domestic producers with an improved competitive edge.

Now, rather than go through this painful approach described above, suppose a country decides to give its domestic industries the desired competitive edge by devaluation--raising the price of gold in terms of the domestic currency. Of course, this sacrifices the core goal of monetary policy--keeping the currency tied to gold at par. But will it provide the improved competitiveness?

By devaluing the currency, imports become more expensive and, from the point of view of foreigners, exported goods become cheaper. Domestic industries are given a competitive edge without slowing the growth of wages or reducing them. There is no need to raise interest rates and attract foreign investment. There is no need for the economy to slow.

Obviously, there are complications. With less gold being received for exports and more gold being paid for imports, it is possible that the gold outflow would accelerate. There is the J curve. And, of course, expectations of a devaluation will result in a gold outflow. But leave these issues aside for a moment.

What if other countries retaliate? Suppose they devalue their currencies as well? Then this competitive edge created by devaluation does not materialize. And so, devaluation does not result in the gold outflow being slowed, stopped, or reversed. The exchange rates between countries don't change, and so nothing much happens.

Ah.. but the "goal" of monetary policy, keeping the paper money on par has been violated, with no good purpose. Right? The world has fallen into a "beggar your neighbor" policy seeking to develop a favorable balance of trade (and nongold capital flows,) to build gold reserves at the rest of the world's expense, and it doesn't work. And each currency is devalued more and more. The horror!

But suppose there is no gold standard. And there is no goal of maintaining the gold redeemablity of paper money. If an exchange rate depreciates, there is no goal of obtaining gold or other foreign exchange to maintain redeemability. No, the goal is to maintain money expenditures on domestic product. In that world--the real world of today--there is no reason to worry that other countries might retaliate to quantitative easing by doing some themselves. Their quantitive easing helps by increasing their domestic purchases, at least some of which are your exported goods.

P.S. Later I will discuss the possibility that trading partners keep their currencies from appreciating by increasing their rate of accumulation of short term debt. (In other words, China responds to U.S. quantitative easing by accelerating their rate of purchase of U.S. Treasury bills.)

Wednesday, October 13, 2010

Please! Professor Meltzer!

Allan Meltzer has criticized the Fed's proposal for a new round of quantitative easing.

He claimed that the Great Inflation was caused by the Fed choosing to target higher inflation in order to reduce unemployment.
Increasing inflation to reduce unemployment initiated the Great Inflation of the 1960s and 1970s. Milton Friedman pointed out in 1968 why any gain in employment would be temporary: It would last only so long as people underestimated the rate of inflation. Friedman's analysis is now a standard teaching of economics. Surely Fed economists understand this.

That isn't the way I understand the history. On the contrary, the Fed targeted short term interest rates in order to lower unemployment, or at least to prevent it from rising, and claimed that inflation was caused by "cost push" factors. So, for example, the Fed would observe rising inflation, and refrain from raising short term interest rates because unemployment was too high, or at least, wasn't too low.

The implicit model was that the Fed controls the short term nominal interest rate, which impacts real expenditure. Real expenditure impacts real output, employment, and unemployment. Inflation depends on "cost push" factors. If the Fed's target for the unemployment rate is too low, then this "model" can easily result in progressively worsening inflation--something like the Great Inflation.

I am no fan of the "new Keynesian" approach, but the model today would be that higher expected inflation reduces real interest rates on short term, safe assets that already have nominal yields near zero. This directly raises real expenditure, but also creates a shift towards riskier and longer term securities that have higher real yields, lowering those real yields as well. This also results in greater real expenditures. The higher real expenditures raises real output and employment and lowers the unemployment rate. Inflation rises consistent with the higher expectations of inflation.

While this approach does imply a higher inflation rate, and perhaps it won't work to reduce unemployment, there seems to be little danger of a return to the Great Inflation of progressively worsening inflation. If inflation rises "too much" then the new Keynesian approach calls for an increase in short term, low risk, nominal interest rates by more than the increase in expected inflation.

While the Fed and the new Keynesians remain wedded to interest rate targets, Milton Friedman won the war about the Fed's responsibility for inflation. Treating a target for unemployment as a goal of monetary policy (say, keep unemployment at 4%,) has next to no support.

Some of Meltzer's other comments are even more disappointing. He claims that there is plenty of liquidity. However, the proper standard isn't how much was apparently needed in the past, but rather relative to the current demand to hold liquid assets. With money expenditures well below their trend from the Great Moderation, it is apparent that there isn't enough to meet the demand.

Surely Meltzer hasn't adopted the "market clearing" approach which implies that the current level of prices and wages must be at the level that keeps the real quantity of money equal to the real demand for money, and simultaneously, the flow of real expenditures equal to the productive capacity of the economy. Sure, there is always enough liquidity to meet the demand if that is true, but how realistic is that?

Similarly, real interest rates, especially on short and safe assets, are very low by historical standards, but apparently they are not low enough to coordinate saving and investment. Perhaps the Obama administrations policy of green jobs, health care reform, and making the rich pay their "fair" share of taxes to fund out of control spending depresses investment. But real interest rates should coordinate saving and investment given expectations of future profitability, not only if business people feel comfortable with the current political leadership.

I am very puzzled by Meltzer's concern that the Fed cannot reduce the excess reserves in the banking system when the demand for credit recovers. The Fed gets weekly data on bank deposits and loans. While the Fed would probably need to raise its target for the Federal Funds rate in the face of growing loan demand, the quantity of credit offered by the banking system doesn't have to shrink, but could even grow. The notion that they must passively allow the quantity of money to expand enough so that all of the excess reserves in the banking system today become required, or perhaps, drained off into currency holdings, is absurd.

Of course, to the degree the Fed holds bad securities (say, mortgage backed securities that are claims to lots of mortgages in default,) the Fed might run into problems. Further, their failed new Keynesian approach of dealing with the zero bound on short term and safe assets by promising to keep interest rates low for a good long time, is a bit of a problem.

Meltzer really does seem to remain committed to some kind of quantity of money rule. And if the demand to hold money rises, so that a quantity of money rule results in monetary disequilibrium, and reduced money expenditures, then the problem is the economy. They shouldn't demand so much money.

No! The quantity of money should adjust to accommodate the demand to hold money, with a nominal anchor of a stable growth path for money expenditures. With such a policy, real output, employment, unemployment, the price level, and inflation all depend on market forces. Still, there is no way that such a policy would result in another Great Inflation. Well, unless the the target for the growth path of money expenditures is in double digits.

Like Meltzer, I am a critic of the Fed. But I think QE2 (another round of quantitative easing) is appropriate because money expenditures are far below any reasonable trend.

I agree with Meltzer that a higher target for the inflation rate is a mistake, but not because I think the current target of 2 percent is optimal. On the contrary, inflation targeting should be replaced by a target for the growth path of money expenditures consistent with zero inflation in the long run.

And what I would like to see from Meltzer is evidence that he remembers the distinction between money and credit and doesn't consider the Federal Funds rate, or any other interest rate, as providing any clear signal of the stance of monetary policy.

Can the Fed come to its Senses?

At the last Fed meeting, somebody mentioned a target for the growth path of money expenditures.....

Participants noted a number of possible strategies for affecting short-term inflation expectations, including providing more detailed information about the rates of inflation the Committee considered consistent with its dual mandate, targeting a path for the price level rather than the rate of inflation, and targeting a path for the level of nominal GDP.
Last but not least?

I might also complain a bit about seeing money expenditure targeting as a means to raise short term inflation expectations (and so, lower real short term interest rates and the output gap.) The actual point would be to raise expectations of real sales, and so increased investment at any level of real interest rate as well raise expectations of employment, and real consumption expenditure. The increase in real expenditures should increase real output (reduce the output gap) and employment. That this might result in higher inflation is an undesirable side effect. While people expecting this unfortunate side-effect shouldn't count as a cost, generating such an expectation is hardly the point.

More importantly, a target for the growth path of money expenditures--if it is permanent--protects against any temporary inflationary impact from creating expectations of permanently higher inflation. Once money expenditures return to an appropriate target, persistently higher inflation would involve firms pricing themselves out of sales. Since firms would be motivated to avoid that sort of suboptimal behavior, expecting such inflation would not be rational.

Most fundamentally, the reason for targeting the growth path of money expenditure is that it provides the least bad macroeconomic environment for microeconomic coordination. Memoryless inflation targeting has proven a failure in the face of a large negative shock to monetary expenditures. Price level targeting would surely be a disaster in the face of any significant adverse supply shock and perhaps bubble prone if there were a favorable aggregate supply shock. Short term interest rate targeting has failed, once again, this time in the face of a severe financial shock that involved a shift away from holding more risky assets to holding safer assets.

It is time for something new.

Well, at least it is on the Fed's radar now.

HT to David Beckworth

P.S. OK, so Scott gets to be Frodo. I have always liked Pippin, and I am clearly not the faithful follower type, but Sam did become Mayor.

Monday, October 11, 2010

Money Expenditures in Japan

What was up with Japan's "lost decade?" Will the U.S. suffer a similar fate?

The monetarists look at some measure of the quantity of money. The neo-Keynesians look at expected inflation and short term interest rates. And what about monetary disequilibriumites? (O.K. As bad as quasi-monetarist sounds, perhaps it is the least bad option.)
I date the "Great Moderation" from 1984 to the third quarter of 2008. I measure money expenditures using Final Sales of Domestic Product. For Japan, I am instead using GDP (which is nominal GDP. Later, I will try to construct Final Sales of Domestic Product for Japan.)

From 1984 to 1990, money expenditures in Japan remained very close to a growth path that increased at a 6 percent annual rate. From the perspective of a "quasi-monetarist," there was a regime change at that time, and a shift to a much lower .12 percent growth path.

There were substantial fluctuations around that trend, rising to as much as 6 percent above trend in 1997 and 3 percent below trend in 2003.

In 2008, the "Great Recession" hit early in Japan, with money expenditures shrinking at a 5 percent annual rate in the second quarter, and then continuing to shrink at that rate until the first quarter of 2009, when they dropped at the remarkable 18 percent annual rate.
Money expenditures remain about 8 percent below the growth path of 1990-2008. (If you look at the 6 percent growth path from early in the Great Moderation, money expentures are 65 percent below that trend. )