The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
March 23, 1984 Anatomy of the 1981-83 Disinflation Monetary models that rely on current and past M 1 growth overpredicted inflation significantly in 1982 and 1983, while forecasting models based upon the effects of economic slack performedquite well. For 1984, a monetary model forecasts an inflation rate of 9.2 percent in the personal consumption deflator, excluding food and energy, but both a "slack" model and a consensus of professional forecasters predict inflation of 5 V2to 6 V2percent. Our purpose in this Letter is to explore in some detail why the monetarist model "broke down" in 1982. more money balances relative to its income because the return on securities becomes less attractive. Therefore, the income velocity of Ml will fall. However, this well-established positive relationship between interest rates and velocity cannot fully explain the failure of monetarist models to account for the recent drop in inflation. If the drop in inflation and nominal interest rates were the only explanation of the sharp decline in velocity, then the decline in velocity cannot be used to explain the large decline in inflation, or else the reasoning becomes circular. Economic slack and monetary growth Economic slack models stressthat high monetary growth boosts inflation only if it raises real aggregate demand enough to lower the rate of unemployment and thereby put upward pressure on wages and prices. In the past, monetary models have tended to capture the short-run association between monetary growth and economic slack, but beginning in 1982, the monetary model overpredicted inflation badly. Evidently, the increase in economic slack in late 1981 and 1982 was larger than could have been predicted solely on the basis of the past slowing in monetary growth. In fact, monetarist equations already capture the changes in velocity caused by lower or higher inflation, making the estimated response of inflation to changes in past monetary growth greater than one-to-one over any but the longest periods. Therefore, some non-monetary shock must have produced a larger amount of economic slackand therefore larger declines in inflation, nominal interest rates, and velocity-than could have been predicted on the basis of the past deceleration in monetary growth. The main source of this shock lay in the foreign trade sector. The proximate reason for why monetary models broke down in 1982 was the unusual decline in the income velocity of mohey or, in other words, its rate of turnover. Between the fourth quarter of 1981 and the first quarter of 1983, the income velocity of M1 declined at a 5;5-percent annual rate in contrast to its long-term positive growth trend of 3 percent. What caused this decline in velocity? One reason was the sharp drop in the inflation rate, and nominal interest rates along with it, that occurred between 1981 and 1982. This changed the relationship between monetary growth and aggregate demand. At lower interest rates, the public will wantto hold An extraordinary decline in net exports of $25.3 billion (in 1972 dollars), plus simultaneous inventory adjustments, accounted for all of the decline in U.S. production during the 1981-82 recession (see chart). Although the recession officially began in July 1981, real final sales (GNP less inventory investment) had already flattened out at the beginning of that year; they declined by $3.6 billion (in 1972 dollars) over the two-year period. However, real final sales are augmented by export sales and reduced by imports. The underlying stre'lgth of the demand of domestic purchasers is thus equal to real final sales less net exports. Net exports and the exchange rate in thi newslctkr do not the views of Bank of the Board of COH'lnors of the Federal ,wnl"''';';O'' 01 '-"'; "",.,. ."'"._----_.,_ .. _._,._ .. -_ _. _- - - - .... . '----_ _-- .. -_ ..._...... We have estimated an econometric tion embodying the monetary model of inflation over the period from 1 964 through 1 980. This equation does not capture even the temporary nexus between monetary growth and the real exchange rate because, in the first place, until 1 973, exchange rates were fixed under the Bretton Woods system and bore no relationship to year-to-year movements in money. Moreover, in the period between 1 973 and 1 980, when exchange rates beween major currencies were a Ilowed to float, a statistical test reveals that changes in the real exchange rate were not significantly related to current and past monetary growth-the influence of other factors dominated. And although a negative relationship does exist between the level of the real exchange rate and monetary growth, it fails to hold into the forecast period beyond 1 980. Consequently, our monetarist equation for forecasting inflation cou Id not have predicted the effect of the sharp rise in the real exchange rate that actually occurred. From the beginning of 1 981 to the rec:ession trough at the end of 1 982, real final sales to domestic purchasers actually increased by $21.7 billion. Therefore, if real net exports had not declined, real final sales would have grown by $21 .7 billion; and the recession would have been much less severe or perhaps even avoided. It is widely agreed that a sharp appreciation in the foreign exchange value of the dollar wast he fundamental cause of the decline in net exports. Between 1 980 and 1 982, the real value ofthe dollar (adjusting for changes in foreign price levels relative to the u.s. price levelL rose by more than 30 percent on a trade-weighted basis. The most important reason for this extremely large rise was the increase in U.s. real interest rates (nominal rates adjusted for inflation) relative to real interest rates abroad. Higher real U.s. interest rates made investment in this country attractive to foreigners, who bid up the real value of the dollar in foreign exchange markets. After a period of time, real appre- : ciation of the dollar reduced the quantity of U.S. exports and increased the quantity of imports. Lesser, but possibly significant, additional factors strengthening the dollar have been recent changes in U.s. tax law that give more favorable treatment to capital investment in the United States and increased risk on investments in other parts of the world. While the slowing in nominal monetary growth that occurred prior to 1982 contributed to a temporary increase in real interest rates (real M 1 declined by over 4 percent a year between 1 979 and 1981) that made the dollar so strong, it is only one of the contributing factors. The reduction in the growth of nominal M 1 accounted for less than half of the reduction in real M 1 growth during this period. The dominant influence on real M 1 was an acceleration of inflation, strongly fueled by shocks from food and energy prices. Money not the only cause Monetarist models failed to capture the effect of amuch stronger dollar in generati ng economic slack by reducing net exports that, in turn, contributed to the large decl ine in inflation. M1 growth slowed from 8.1 percent in 1 978 to 7.4 and 7.2 percent in 1 979 and 1 980, respectively, and to 5.1 percent in 1 981 . Such a slowing should tend to raise real short-term interest rates, and hence the real exchange rate, temporarily by first reducing the growth of real money balances. , As the price level adjusts in the long-run, however, the real stock of money, real interest rates, and the real exchange rate should return to normal. More importantly, the strength of the effect of higher real short-term interest rates on the real exchange rate depends upon how long high real rates are expected to last, or, equ ivalently, whether they are transmitted to real long-term interest rates. If a real interest rate differential on a 1-year bond of 3 percentage points in favor of the United States is expected to last for a year, it could boost the real exchange rate above its long2 . _-_. _-----_ run value by only 3 percent. That is, the expected return of the exchange rate to its long-run value would just offset the extra yield on the bond. But if this differential were anticipated to last for 10 years, the real exchange rate should rise by 30 percent. Two readily identifiable factors contributed to market expectations that high real shortterm interest rates would last for some time. First, because the Administration supported the Federal Reserve's efforts to reduce the rate of monetary growth, international investors renewed their confidence in the ability of the United Statesto pursue a course of stable monetary policy. Such a policy reduces thelikelihood that real interest rates could be temporarily depressed by excessive monetary stimulation in the future. Second, after passageof the Economic Recovery Tax Act in the Summer of 1981, it soon became clear that the structural, or cyclically adjusted, federal budget· deficit would grow very substantially in the future. The demand of the federal government for credit would therefore be'expected to keep future real short-term interest rates high. Conclusion The solution to the puzzle of why monetarist equations for forecasting inflation have broken down recently lies in the foreign .. _ - - trade sector. During the 1981 the drop in real final demand was led by a drop in real net exports. Slower monetary growth contributed in expected degree to weakness in such sectors as consumer durables, housing, and business fixed investment. But the evidence indicates that an appreciation in the real foreign exchange value of the dollar, which was the key factor in the sharp drop of net exports, could not have been accurately predicted from the previous slowing in monetary growth. The most important influence on the real exchange rate was the expectation of continued high real interest rates created by renewed international confidence in U.s. monetary stabi lity and by the anticipation of large federal budget deficits. The decline in net exports contributed greatly to the increase in economic slack, that was mainly responsible for the subsequent drop in inflation. The reduction in real aggregatedemand caused by the drop in net exports helped to produce a decline in the income velocity of M 1, at first directly, and later indirectly as less inflation led to lower nominal interest rates. Because of this quite independent effect of net exports upon aggregate demand and economic slack, the decline in inflation was considerably greater than predicted by monetarist forecasting equations. Adrian W. Throop The 1981·1982 Recession Trough of Recession Real Final Sales Real GNP ...... Real Final Sales to Domestic Purchasers .LSl:Il.::I uOll'lUI4SPM.4Pln • uol'laJO • l'pl'AaN • 04l'pl !!l'Ml'H l'! UJOjlll':). PUOZ!J\! • P>jSl'IV 'J!I!!::> 'mspu!!J:J U!!S lSL 'ON OIVd:I9VlS Od 's'n llVW SSVU lsm: J <BHIOS:I}ld BANKING DATA-TWELFTH FEDERAL RESERVE DISTRICT (Dollar amounts in millions) Selected Assetsand liabilities Large Commercial Banks Loans, Leases and Investments 1 2 Loans and Leases1 6 Commercial and Industrial Real"estate Loans to Individuals Leases U.S. Treasury and Agency Securities2 Other Securities 2 Total Deposits Demand Deposits Demand Deposits Adjusted 3 Other Transaction Balances4 : Total Non-Transaction Balances6 Money Market Deposit Accounts-Total Time Deposits in Amounts of $100,000 or more Other Liabilities for Borrowed MoneyS Weekly Averages of Daily Figures ReservePosition, All Reporting Banks Excess Reserves (+ )/Deficiency (-) Borrowings Net free reserves (+ )/Net borrowed ( -) Amount Outstanding 3/7/84 176,691 156,418 46,636 59,276 26,947 5,011 12,184 8,088 185,698 43,364 29;102 12,467 129,867 -1,010 993 373 46 2 5 3 - 13 295 - 872 403 462 114 40,524 Change from year ago Percent Dollar Change from 2/29/84 151 37,971 19,327 Weekended 3/7/84 - - - - 666 1,063 673 377 296 50 322 75 5,298 5,872 2,229 307 882 - - . 10.7 927 - 113 894 - 193 3,679 Weekended 2/29/84 NA NA NA - NA NA NA 1.7 3.1 6.7 2.9 5.1 4.6 11.8 4.2 12.7 54.4 32.5 11.0 3.1 - 2.3 73.0 Comparable year-ago period NA NA NA 1 Includes loss reserves, unearned income, excludes interbank loans Excludes trading account securities Excludes u.s. overnment and depository institution deposits and cash items g ATS, N OW, Super N O W and savings accounts with telephone transfers S Includes borrowing via FRB, TT&L notes, Fed Funds, RPsand other sources 6 Includes items not shown separately 2 3 4 Editorial commentsmay be addressed the editor (GregoryTong)or to the author .... Freecopiesof to Federal Reservepublications can be obtained from the Public Information Section, FederalReserve Bank of San Francisco, 7702, SanFrancisco94120. Phone(415) 974-2246, .