Edward Renshaw
Professor of Economics
State University of New York at Albany
Until the 1970s commercial banks were the only financial institutions that were allowed to provide their customers with checking accounts and were also prohibited from paying interest on those accounts.
In 1970, however, saving and loan associations were allowed to make preauthorized nonnegotiable transfers from savings accounts for household- related payments. In 1972 negotiable order of withdrawal (NOW) accounts that pay interest on checkable deposits were first offered by state- chartered mutual savings banks in Mass. and NH. In 1975 telephone transfers were allowed for savings balances at commercial banks and the Fed began reporting monetary growth objectives to Congress quarterly in response to a Congressional resolution. In 1976 depository institutions in Connecticut, Maine, Rhode Island and Vermont were authorized to offer NOWS and state- chartered mutual savings banks and S&Ls in New York state were authorized to offer demand deposits.
The Full Employment and Balanced Growth Act of 1978, known as the Humphrey-Hawkins Act, required the Fed to make semi-annual reports to Congress on its annual targets for growth of the monetary aggregates and to report on any variance. In 1979 the Fed announced a policy shift where it would directly control nonborrowed reserves to keep M1 within target ranges and let interest rates vary. Wide fluctuations in the federal funds rate ended the specter of double digit inflation by helping to tip the US economy into two recessions in less than two years.
The Depository Institutions Deregulation and Monetary Control Act of 1980 approved providing for a phaseout of interest rate ceilings on time and savings deposits and the introduction of NOW accounts at all depository institutions. In 1981 interest bearing NOW accounts prompted depositors to transfer funds from savings accounts in M2 into NOWs which are now included in the more conventional monetary aggregate, M1. M1, which had been increasing at a fairly steady rate of 6.8 percent from 1979-81, accelerated to a record increase of 16.9 percent in 1986 before plunging to only .9 percent in 1989.
In 1982 money market deposit accounts (MMDAs) were authorized by the Depository Institutions Deregulation Act. In the same year the Fed shifted policy again and dropped M1 as a policy target after the aggregate became unreliable as an indication of economic growth and inflation. For the fifth time since 1970 the Fed redefined the monetary aggregates and divided them into M1, M2, M3 and L. In 1984 the debt of domestic nonfinancial sectors was added to the Fed's money stock, liquid assets, and debt measures statistical report.
In 1987 the phaseout of interest rate ceilings on time and savings deposits that began in 1980 was concluded. In the same year the Fed stopped reporting growth ranges for M1 in its Humphrey-Hawkins report to Congress. M2 became the most important monetary aggregate in policy implementation.
In 1989 consumers and business enterprises, in a somewhat belated response to rising short term interest rates, shifted large amounts of funds from demand deposits to money market deposit accounts, which are not included in M1, and slowed the growth of M1 to only .9 percent. This slump in the growth rate for M1 caused the demand for money equations in Table 6.2 to predict an increase in long term interest rates at a time when the yield on Moody's high grade bond index was actually declining.
From March 1989 to July 1990 the federal funds rate was allowed to drop more than 1.50 percentage points from 9.85 percent to only 8.29 percent. This easing of monetary policy, however, was not sufficient to prevent the U.S. economy from slipping into another recession.
During the recessionary trough year of 1991 investors, in response to low deposit rates that failed to provide a positive after tax and inflation return on their savings, began shifting large amounts of funds from certificates of deposit found in M2 into long term stock and bond funds which are not included in the monetary aggregates.
In 1992 the Fed's discount rate was reduced to 3 percent, its lowest level in more than 30 years. A 1.7 percent growth rate for M2 fell below the Fed's target range of between 2.5 and 6.5 percent when nominal GDP was increasing at a 5.5 percent rate.
In 1993 Fed Chairman Greenspan announced that M2 had been downgraded as a reliable indicator of financial conditions in the economy and "that no single variable has yet been identified to take its place".
In February 1994 the Fed began a policy of gradually raising the federal funds rate to slow the economic growth rate to a more sustainable pace and prevent an acceleration of the inflation rate.
In September 1995 chairman Greenspan admitted that the Fed has only one tool (the federal funds rate) for regulating the US economy and was sympathetic to the idea that the Fed's primary objective should be to maintain price stability.
At that point in time higher short term interest rates were again encouraging banks to automatically "sweep" or shift funds from reservable checking accounts to non-reservable money market deposit accounts. Banks benefit from lower reservable deposits and hence lower reserve requirements at the Fed. On a Dec.-Dec. basis M1 declined 2.1 percent during 1995 at a time when M2 was accelerating from only .4 percent in 1984 to 4.2 percent. This was the biggest contraction in the more conventional money supply since the recessionary trough year of 1937.
The yield on Moody's high grade corporate bond index bottomed out at 6.66 percent in September 1993 and was increasing before the Fed launched its preemptive strike against the possibility of accelerating inflation in February 1994. It then spiraled upward to a high of 8.68 percent in November 1994 before drifting downward to 6.82 percent in December of 1995. The conflicting behavior of long and short term interest rates caused the demand for money models in Table 6.2 to again predict an increase in long term interest rates that did not occur. Financial innovation, in any event, has made the demand for money more complicated than it used to be.
For many years the square root rule, borrowed from inventory management and promoted by Baumol in (1952), provided a useful framework for explaining the demand for money. The financial innovations of the 1980s, however, have made M1 more substitutable for other types of savings and may have created a need for another role model. The inspiration for an alternative model that was usually more in tune with interest rate changes and the demand for cash balances from 1987-95 than the Baumol model can be obtained from a (1960) paper by Latane. He found that the interest rate on long term high grade corporate bonds was approximately equal to 1.3 times the velocity of money minus half a percentage point over the 1909-58 period. In the analysis which follows we will ignore the relatively small constant term and assume that:
i = kV (1)
The data presented in column (5) of Table 5.1 indicate that k has drifted downward from an all time high of 2.02 in 1982 to 1.18 in 1995 or the same value which prevailed in 1960 when the Latane article was first published. The more recent estimates of k, however, are not very different, on the average, from the regression coefficient of 1.3 which Latane obtained for the 1909-58 period.
In his 1952 paper Baumol derived the square root formula:
.5 -.5
C = 2bT i (2)
In contrasting the predictions of Latane and Baumol we will follow Goldfeld (1973) and interpret C as being equivalent to some measure of real cash balances (C = M/P) and let T represent some measure of economic activity (T = Q) such as real GNP or GDP. Converting equation (2) to logarithms and noting that small changes in logarithmic values are approximately equal to percentage changes one can obtain the prediction that the percentage change in the interest rate should be about equal to the growth rate for real GDP plus two times the inflation rate minus two times the growth rate for the money supply if the all encompassing brokerage fee, b, is constant.
From 1969-90 this formula correctly explained the direction of the average yield change for Moody's high grade corporate bond index 85 percent of the time and sometimes came very close to predicting the magnitude of the actual yield change (Renshaw 1992, Table 1.82). In column (6) of Table 6.2, Baumol's predictions are reestimated using figures for the US Commerce Department's new kid on the block, GDP, for the 1982-95 period. These predictions can be contrasted to the simpler predictions of the Latane model in column (7) which presumes that the percentage change in the interest rate should be equal to the growth rate for nominal GDP minus the growth rate for the money supply if k is indeed a constant. When these predictions are compared to the actual changes in the long term interest rate in column (8) the conclusion is that the Latane model outperformed the Baumol model by a ratio of 2-1 from 1982-96.
Instability in the demand for money is not a new phenomenon. It has given rise to such classic articles as Stephen Goldfeld's (1976) paper titled "The Case of the Missing Money". It may have also been instrumental in encouraging the Institute of Business Forecasting in the user's guide to their Easy Forecaster Plus software package to admonish users to never trust a model. "Keep a close eye on the performance of a model. If and when it stops working, discard it." The Fed, in managing the nation's money supply, examines a wide range of information and now has a long history of ignoring demand for money equations when their predictions seem out of tune with what is happening to the U.S. economy.
The Latane model is not as rich from a theoretical point of view as the Baumol model. It does have the advantage of being simple enough to be discussed at the principles level and also provides another anchor or focal point for discussing monetary policy in a sea of financial change.
Instability in the demand for M1 has led some economists to rearrange the Latane and Baumol equations and focus their attention on the implied impact of changes in money, output, inflation and other variables on interest rates.
Studies by McNees (1986 and 1992) indicate that the Federal Reserve's interest rate reaction function has been forward looking. Changes in the federal funds rate and other short term interest rates suggest that the Fed dislikes inflation but has reacted even more strongly to increases in the unemployment rate (Trahan and Renshaw 1990).
After drifting downward from a post World War II high of 10.8 percent in December 1982 to only 5.0 percent in March 1989, the civilian unemployment rate then began to edge upward to a recessionary high of 7.7 percent in June 1992. During that prolonged period of rising unemployment the Federal Reserve followed a rather steady policy of lowering the federal funds rate. See columns (1), (3) and (5) of Table 6.3.
A decrease in short term interest rates, however, will not be very effective at preventing a recession or stimulating an economic recovery from a recession unless long term interest rates also decline. From March 1989 to June 1992, when the federal funds rate was reduced from 9.85 percent to only 2.92 percent, the yield on ten year Treasury bonds only declined by a little over two percentage points from 9.36 percent to 7.26 percent.
The data in the last two columns of Table 6.3 raises an interesting question as to whether the Federal Reserve hasn't lost some of its ability to control the direction of long term interest rates. During the 1990s quarterly changes in the yield on ten year Treasury bonds have moved in an opposite direction from changes in the federal funds rate, more often than not. This can be contrasted to the 1950s, 60s, 70s and 80s when ten Treasury bond yields only moved opposite to the federal funds Rate 14.3, 35.0, 25.0 and 20.0 percent of the time respectively.
There are a number of reasons why changes in the federal funds rate may now have less influence on long term interest rates than used to be the case. One reason that is often cited is a globalization of the market for long term debt. When monetary policy and economic activity in the more industrialized countries of the world are not in sync, multi national corporations are likely to borrow in the cheapest country.
An explosive growth of interest rate derivatives and the profits to be obtained from international arbitrage may have also lessened the power of the Fed to control the direction of long term interest rates in this country.
Another development that ought to be given some consideration is the possibility that speculators in the bond market are now more "foreword looking" than the Board of Governors of the Federal Reserve. One cannot help but be impressed by the fact that long term interest rates turned up before the Fed began its preemptive strike against the possibility of accelerating inflation in February 1994 and also turned down before the federal funds rate was allowed to decline in 1995.
During the first four months of 1996 long term interest rates spiked upward by more than one percentage point in response to higher oil prices and a dramatic increase in grain prices. The Board of Governors of the Federal Reserve, after lowering the federal funds rate a bit to make sure the economy would not slip into another recession, remained rather complacent about the outlook for inflation. Let us hope that the Fed's complacency turns out to be right in this case.
There are two classes of forecasters according to John Kenneth Galbraith: "Those who don't know--and those who don't know they don't know." Tom Herman (1993) has gone even further and reported: "Looking for an easy way to make money betting on interest rates? Find out the consensus forecast among leading economists--and then bet against it. That strategy would have paid off handsomely, according to two separate analyses of fearless forecasts from economists polled semiannually since 1981 by The Wall Street Journal."
"It's crazy," says Jack Miller, a New York investment specialist. "These guys get paid enormous sums to make these calls, but it turns out you would actually have been better off" betting the other way.
Robert Beckwitt, a portfolio manager at Fidelity Investments in Boston has found that by betting against the consensus 30 year Treasury bond yield forecast every six months since December 1981, an investor would have earned a sparkling average annual return of 13.7 percent.
The yield on new issues of 91 day Treasury bills has fluctuated in a procyclical manner in the post World War II period. Economic forecasters, however, have not had good luck at forecasting the direction of short term interest rates. During the 14 year period from 1982 to 1995 the consensus forecast compiled by the Blue Chip Economic Indicator survey only outperformed a no change in the last offering yield during four years (Table 6.4). Two of these successful cases followed years containing a recessionary peak in business activity.
It is not easy to find variables that do a good job of explaining future changes in interest rates. There is some tendency, however, for participants in the long term bond market to over react to the threat of inflation. In the post 1955 period there has been a weak inverse relationship between following Dec.-Dec. changes in the yield on ten year Treasury bonds and the current year change in T-bond yields minus the current year change in the all item CPI inflation rate. See Table 6.5.
The ability of this indicator to explain future changes in the T-bond yield can be improved a little by also including in a multiple regression the Nov.-Nov. percentage change in residential building permits. At the end of 1995 these indicators were predicting an increase in long term interest rates for 1996.
One reason, perhaps, why economists have not had much luck at forecasting interest rates is that the Board of Governors of the Federal Reserve does take a hard look at economic forecasts and is rather forward looking in setting interest rate policy (McNees 1986 and 1992). Another reason for poor forecasts is that changes in the inflation rate--the most important determinant of interest rates in the long run--are often difficult to predict unless the economy gets mired down in a recession.
Profitability in the banking industry and the welfare of investors in bank stocks has been sensitive to the spread between the interest rates paid by banks to depositors and the rates that they can obtain on loans and other types of investments.
The federal funds rate is sometimes referred to as the base rate to which other money market rates are anchored. In column (3) of Table 6.6 the December values for the federal funds rate are subtracted from the prime rate. The following year percentage changes in bank stock price indexes in column (4) are not inconsistent with the hypothesis that bank stocks perform better, on the average, in a high spread environment than in a low spread environment.
It should be noted, however, that the character of the banking industry has been changing. Though still impeded by regulatory constraints many commercial banks are now providing their customers with a much wider range of financial services than was the case in the 1950s and 1960s. With more "off balance sheet activity" taking place and with the prime rate now linked more closely to changes in the Federal funds rate it is by no means clear that this type of spread will continue to be as useful at identifying good years to own bank stocks.
Fluctuations in the earnings per share associated with the Standard and Poor's index of major regional banks in column (6) of Table 6.6 provide a nice departure point for appreciating the problem of bank failures which emerged after major changes in the regulatory climate in the 1980s. It should be noted, however, that industry averages can hide a lot of variability and that the profitability and financial returns for some individual commercial banks has diverged significantly in recent years. Some banks have remained very conservative in their lending policies and have not experienced major loan losses or as much variance in their earnings as might be indicated by the S&P earning index.
Baumol, W. (1952). "The Transactions Demand for Cash: An Inventory Theoretic Approach," Quarterly Journal of Economics, November.
Federal Reserve Bank of New York (1989). U.S. Monetary Policy and Financial Markets.
-----, (1994). "Understanding the M's in Monetary Policy". Provides a condensed history of the M's and financial innovations.
Goldfeld, Stephen (1973). "The Demand for Money Revisited," Brookings Papers on Economic Activity, 4(3), 577-638.
Herman, Tom (1993). "Why Investors Shouldn't Listen to Economists," The Wall Street Journal, January 22, Cl and C15.
-----, (1976). "The Case of the Missing Money," Ibid., (3), 683-730.
Keim, Donald and Robert Stambaugh (1986). "Predicting Returns in the Stock and Bond Markets," Journal of Financial Economics, (17), 357- 90.
Latane, Henry (1960). "Income Velocity and Interest Rates: A Pragmatic Approach," Review of Economics and Statistics, 42(November), 445-49.
McNees, Stephen (1986). "Modeling the Fed: A Forward-Looking Monetary Policy Reaction Function," New England Economic Review, Nov./Dec., 3-8.
-----, (1992). "A Forward-Looking Monetary Policy Reaction Function: Continuity and Change," Ibid., 3-13.
Mehra, Yash (1992). "In Search of a Stable, Short-Run M1 Demand Function," Federal Reserve Bank of Richmond, Economic Review, 78(May/June), 9-23.
Renshaw, Edward (1992). Editor, The Practical Forecaster's Almanac(Business One Irwin).
Trahan, Emery and Edward Renshaw (1990). "Presidential Elections and the Federal Reserve's Interest Rate Reaction Function," Journal of Policy Modeling, 12(1), 29-34.
Year Nominal M1 Velocity Yield Aaa k = i/V Change
GDP Dec. of M1 Bonds or 4/3 in k
(Billions of Dollars)
(1) (2) (3)n (4)n (5) (6)n
1960 526.6 140.7 3.74 4.41 1.18 - .03
1961 544.8 145.2 3.75 4.35 1.16 - .02
1962 585.2 147.8 3.96 4.33 1.09 - .07
1963 617.4 153.3 4.03 4.26 1.06 - .03
1964 663.0 160.3 4.14 4.40 1.06 .00
1965 719.1 167.9 4.28 4.49 1.05 - .01
1966 787.8 172.0 4.58 5.13 1.12 .07
1967 833.6 183.3 4.55 5.51 1.21 .09
1968 910.6 197.4 4.61 6.18 1.34 .13
1969 982.2 203.9 4.82 7.03 1.46 .12
1970 1,035.6 214.4 4.83 8.04 1.66 .22
1971 1,125.4 228.3 4.93 7.39 1.50 - .16
1972 1,237.3 249.2 4.97 7.21 1.45 - .05
1973 1,382.6 262.8 5.26 7.44 1.41 - .04
1974 1,496.9 274.3 5.46 8.57 1.57 .16
1975 1,630.6 287.4 5.67 8.83 1.56 - .01
1976 1,819.0 306.3 5.94 8.43 1.42 - .14
1977 2,026.9 331.1 6.12 8.02 1.31 - .11
1978 2,291.4 358.4 6.39 8.73 1.36 .06
1979 2,557.5 382.8 6.68 9.63 1.44 .07
1980 2,784.2 408.8 6.81 11.94 1.75 .31
1981 3,115.9 436.5 7.14 14.17 1.98 .23
1982 3,242.1 474.5 6.83 13.79 2.02 .04
1983 3,514.5 521.1 6.74 12.04 1.78 - .24
1984 3,902.4 552.1 7.07 12.71 1.80 .02
1985 4,180.7 619.8 6.75 11.37 1.68 - .12
1986 4,422.2 724.4 6.10 9.02 1.48 - .20
1987 4,692.3 749.8 6.26 9.38 1.50 .02
1988 5,049.6 786.9 6.42 9.71 1.51 .01
1989 5,438.7 794.2 6.85 9.26 1.35 - .16
1990 5,743.8 825.8 6.96 9.32 1.34 - .01
1991 5,916.7 897.2 6.59 8.77 1.33 - .01
1992 6,244.4 1,024.4 6.10 8.14 1.33 .00
1993 6,553.0 1,128.6 5.81 7.22 1.24 - .09
1994 6,935.7 1,148.7 6.04 7.97 1.32 .08
1995 7,253.8 1,124.9 6.45 7.59 1.18 - .14
1996 7,580.0 1,076.9 7.04 7.37 1.05 - .13
(3)n. Column (1) divided by column (2).
(4)n. Average yield on Moody's Aaa corporate bond index.
(6)n. First differences in the value of k or column (4) divided by column (3) shown in column (5).
Source of basic data: Economic Report of the President.
Yield Average Growth Rate Percentage Change in Moody's
Year Aaa ------------------- Dec.-Dec. Aaa Corporate Bond Yields
Corp. Nominal Real IPD Growth of ----------------------------
Bonds GDP GDP GPD M1 Predicted Change Actual
Baumol Latane Change
(1) (2) (3) (4) (5) (6)n (7)n (8)
1982 13.79 4.1 -2.1 6.3 8.7 -6.9 -4.6B -2.7
1983 12.04 8.4 4.0 4.2 9.8 -7.2B -1.4 -12.7
1984 12.71 11.0 6.8 3.8 6.0 2.4 5.0B 5.6
1985 11.37 7.1 3.7 3.4 12.3 -14.1B -5.2 -10.5
1986 9.02 5.8 3.0 2.6 16.9 -25.6B -11.1 -20.7
1987 9.38 6.1 2.9 3.1 3.5 2.1 2.6B 4.0
1988 9.71 7.6 3.8 3.7 4.9 1.4 2.7B 3.5
1989 9.26 7.7 3.4 4.2 .9 10.0 6.8B -4.6
1990 9.32 5.6 1.3 4.4 4.0 2.1 1.6B .6
1991 8.77 3.0 -1.0 3.9 8.6 -10.4 -5.6B -5.9
1992 8.14 5.5 2.7 2.8 14.2 -20.1 -8.7B -7.2
1993 7.22 4.9 2.2 2.6 10.2 -13.0B -5.3 -11.3
1994 7.97 5.8 3.5 2.3 1.8 4.5B 4.0 10.4
1995 7.59 4.6 2.0 2.5 -2.1 11.2 6.7B -4.8
1996 7.37 4.5 2.5 2.0 -4.3 15.1 8.8B -2.9
(6)n. The predicted change is equal to the growth of real GDP in column (3) plus two times the IPD inflation rate in column (4) minus two times the growth of M1 in column (5).
(7)n. The predicted change is equal to the growth of nominal GDP in column (2) minus the growth of M1 in column (5).
B identifies the prediction which is closest to the actual percentage change in Moody's Aaa corporate bond yield in column (8).
Source of basic data: Economic Report of the President.
Quarter Unemp. CPI Federal 10 Year First Differences
Rate Inflation Funds Treasury Fed Treasury
Rate Rate Bond Yield Funds Bond Yield
(1) (2) (3) (4) (5) (6)
1988-1 5.7 3.93 6.58 8.37 - .19 - .62
-2 5.4 3.96 7.51 8.92 .93 .55
-3 5.4 4.17 8.19 8.98 .68 .06
-4 5.3 4.42 8.76 9.11 .57 .13
1989-1 5.0L 4.98 9.85H 9.36H 1.09 .25
-2 5.3 5.17 9.53 8.28 - .32 -1.08
-3 5.3 4.34 9.02 8.19 - .51 - .09
-4 5.4 4.65 8.45 7.84 - .57 - .35
1990-1 5.2 5.23 8.28 8.59 - .17 .75*
-2 5.1 4.67 8.29 8.48 .01 - .11*
-3P 5.7 6.16H 8.20 8.89 - .09 .41*
-4 6.2 6.11 7.31 8.08 - .89 - .81
1991-1T 6.8 4.90 6.12 8.11 -1.19 .03*
-2 6.8 4.70 5.90 8.28 - .22 .17*
-3 6.7 3.39 5.45 7.65 - .45 - .63
-4 7.2 3.06 4.43 7.09 - .02 - .56
1992-1 7.3 3.19 3.98 7.54 - .45 .45*
-2 7.7H 3.09 3.76 7.26 - .22 - .28
-3 7.5 2.99 3.22 6.42 - .54 .16*
-4 7.3 2.90 2.92L 6.77 - .30 .35*
1993-1 7.0 3.09 3.07 5.98 .15 - .79*
-2 7.0 3.00 3.04 5.96 - .03 - .02
-3 6.7 2.69 3.09 5.36L .05 - .60*
-4 6.5 2.75 2.96 5.77 - .13 .41*
1994-1 6.5 2.51 3.34 6.48 .38 .71
-2 6.1 2.49 4.25 7.10 .91 .62
-3 5.9 2.96 4.73 7.46 .48 .36
-4 5.4 2.67 5.45 7.81H .72 .35
1995-1 5.4 2.85 5.98 7.20 .53 - .61*
-2 5.6 3.04 6.00H 6.17 .02 -1.03*
-3 5.7 2.54 5.80 6.20 - .20 .03*
-4 5.6 2.54 5.60 5.71 - .20 - .49
1996-1 5.5 2.84 5.31 6.27 - .29 .56*
-2 5.3 2.75 5.27 6.91 - .04 .64*
-3 5.2 3.00 5.30 6.83 .03 - .08*
-4 5.3 3.32 5.29 6.30 - .01 - .53
Footnotes for Table 6.3
H identifies a cyclical high value for the series.
L identifies a cyclical low value for the series.
P identifies a quarter containing a recessionary peak in economic activity.
T identifies a quarter containing a recessionary trough in economic activity.
*identifies quarters when the change in the end of the quarter yield on ten year Treasury bonds was opposite of the change in the federal funds rate in column (5).
Source of basic data: Economic Report of the President.
Year Blue Chip Last Offering Actual
Consensus Yield for the Average
Forecast Preceding T-Bill
(January December Offering
Survey) Yield
(1)n (2)n (3)
1982T 10.8(- .1)B 11.7(-1.0) 10.7
1983 7.7( .9) 8.0( .6) 8.6
1984 8.8( .8) 8.9( .7) 9.6
1985 8.6(-1.1) 7.8(- .3) 7.5
1986 7.1(-1.1) 7.0(-1.0) 6.0
1987 5.4( .4) 5.7( .1) 5.8
1988 6.0( .7)B 5.7( 1.0) 6.7
1989 7.8( .3) 8.2(- .1) 8.1
1990 7.1( .4) 7.8(- .3) 7.5
1991T 6.3(- .9)B 6.8(-1.4) 5.4
1992 4.2(- .7) 4.1(- .6) 3.5
1993 3.5(- .5) 3.2(- .2) 3.0
1994 3.4( .9)B 3.1( 1.2) 4.3
1995 6.2(- .7) 5.6(- .1) 5.5
Mean Absolute Error
1982-95 .68 .61
Footnotes for Table 6.4
(1)n. The January survey is conducted very early in the month. The figures in parentheses are the forecasting errors obtained by subtracting the Blue Chip consensus forecast from the actual average yield on new offerings in column (3).
(2)n. The average last offering yield on 91-day Treasury bills for the preceding month of December published in Moody's Municipal & Government Manual. The figures in parentheses are error terms obtained by subtracting the last offering yield from the actual average new offering yield in column (3).
B identifies years when the forecasting error for the Blue Chip consensus was smaller absolutely than the no change forecasting error in column (2).
T identifies years containing a recessionary trough in business activity.
Year ---Dec.-Dec. Changes--- Col. (1) Following
T-Bond CPI Minus Change in
Yield Inflation Col. (2) T-Bond Yield
(1) (2) (3) (4)
1975 .57 -5.4 6.0 -1.13
1981 .88 -3.6 4.5 -3.18
1980 2.45 - .8 3.2 .88
1991 - .99 -3.0 2.0 - .32
1994 2.04 .0 2.0 -2.10
1982 -3.18 -5.1 1.9 1.29
1971 - .46 -2.3 1.8 .43
1958 .65 -1.1 1.7 .83
1967 .86 - .5 1.4 .33
1983 1.29 .0 1.3 - .33
1959 .83 - .1 .9 - .85
1961 .22 - .7 .9 - .20
1976 -1.13 -2.0 .9 .82
1964 .05 - .6 .6 .44
1986 -2.15 -2.7 .5 1.88
1972 .43 .1 .3 .38
1969 1.62 1.5 .1 -1.26
1988 .12 .0 .1 -1.27
1963 .27 .3 .0 .05
1992 - .32 - .2 - .1 -1.00
1957 - .38 - .1 - .3 .65
1984 - .33 .1 - .4 -2.24
1960 - .85 - .3 - .5 .22
1965 .44 .9 - .5 .22
1955 .45 1.1 - .7 .63
1970 -1.26 - .6 - .7 - .46
1962 - .20 .6 - .8 .27
1993 -1.00 - .2 - .8 2.04
1977 .82 1.8 -1.0 1.32
1978 1.32 2.3 -1.0 1.38
1989 -1.27 .2 -1.1 .24
1996--- - .31------- .8---- -1.1------ ?
1990 .24 1.5 -1.3 - .99
1966 .22 1.6 -1.4 .86
1968 .33 1.7 -1.4 1.62
1987 1.88 3.3 -1.4 .12
1995 -2.10 - .2 -1.9 - .31
1956 .63 2.6 -2.0 - .38
1985 -2.24 - .1 -2.1 -2.15
1974 .69 3.6 -2.9 .57
1979 1.38 4.3 -2.9 2.45
1973 .38 5.3 -4.9 .69
Source of basic data: Economic Report of the President.
Following Year
December Values Percentage Changes
--------------- ------------------ High
Year Prime Federal Spread Bank S&P Earnings P/E
Rate Funds Stocks 500 per Share Ratio
(1) (2) (3)n (4)n (5) (6)n (7)n
1954T 3.00 1.26 1.74 3.9 26.4* ---- ----
1955 3.50 2.48 1.02 -3.8# 2.6* ---- ----
1956 4.00 2.94 1.06 -11.5# -14.3 ---- ----
1957 4.50 2.98 1.52 35.1 38.1* 3.31 12.54
1958T 4.00 2.42 1.58 16.8 8.5 3.46 14.08
1959 5.00 3.99 1.01 -.9# -3.0 3.87 14.74
1960 4.50 1.98 2.52 45.2$ 23.1 4.32 13.19
1961T 4.50 2.33 2.17 -16.9 -11.8* 4.11 20.69
1962 4.50 2.93 1.57 11.5 18.9* 3.96 19.19
1963 4.50 3.38 1.12 -.3 13.0* 4.15 18.90
1964 4.50 3.85 .65 -8.7# 9.1* 4.55 17.91
1965 4.92 4.32 .60 -2.2# -13.1 4.86 15.55
1966 6.00 5.40 .60 -1.9# 20.1* 5.49 13.08
1967 6.00 4.51 1.49 46.1 7.7 5.85 12.29
1968 6.60 6.02 .58 -14.0# -11.4* 6.71 15.07
1969 8.50 8.97 -.47 2.5# .1 6.96 14.19
1970T 6.92 4.90 2.02 6.1 10.8* 7.62 11.20L
1971 5.25 4.14 1.11 28.4 15.6 7.90 11.96L
1972 5.79 5.33 .46 -15.7# -17.4 8.48 14.37
1973 9.75 9.95 -.20 -38.8# -29.7* 9.56 12.29
1974 10.50 8.53 1.97 17.7 31.5* 10.43 10.72L
1975T 7.26 5.20 2.06 28.1 19.1 11.49 8.10L
1976 6.35 4.65 1.70 2.8 -11.5 11.13 10.11L
1977 7.75 6.56 1.19 7.4 1.1 12.42 8.76L
1978 11.55 10.03 1.52 5.8 12.3* 15.44 7.58L
1979 15.30 13.78 1.52 9.4 25.8* 17.29 6.93L
1980T 20.35 18.90 1.45 20.9 -9.7 17.54 6.82L
1981 15.75 12.37 3.38 9.2$ 14.8* 17.57 7.53L
1982T 11.50 8.95 2.55 30.5$ 17.3 15.66 7.54L
1983 11.00 9.47 1.53 12.6 1.4 11.16 11.12L
1984 11.06 8.38 2.68 52.0$ 26.3 .45 ....
1985 9.50 8.27 1.23 18.1 14.6 6.42 17.17
1986 7.50 6.91 .59 -5.3# 2.0* 3.12 40.67
1987 8.75 6.77 1.98 11.4 12.4* -9.03 def
1988 10.50 8.76 1.74 -10.2 27.3* 1.35 83.79
1989 10.50 8.45 2.05 -34.8 -6.6* 10.37 13.31
1990 10.00 7.31 2.69 37.5$ 26.3 8.67 13.91
1991T 7.21 4.43 2.78 52.0$ 4.5 5.26 26.24
1992 6.00 2.92 3.08 29.4$ 7.1 10.85 15.79
1993 6.00 2.96 3.04 1.1$ -1.5 15.74 12.79
1994 8.50 5.45 3.05 44.8$ 34.1 18.73 10.41L
1995 8.65 5.60 3.05 ?$ ?
Footnotes for Table 6.6
(3)n. Column (1) minus column (2).
(4)n. The following year percentage change in bank stocks is based on the S&P weekly index of banks outside York City from 1954-72. Since 1972 the percentage changes are based on the NASDAQ bank stock index.
(6)n. The earnings per share associated with Standard and Poor's index of major regional banks.
(7)n. The high price to earnings ratio associated with Standard and Poor's index of major regional banks.
T identifies years containing a recessionary trough in economic activity.
L identifies years when the high price-earnings ratio was under 12.00. These years have so far always been followed by a gain in bank stock prices in column (4).
# identifies following year percentage changes in the bank stock index where the spread in column (3) was less than 1.10 percentage points. Except for 1969 all of these percentage changes are negative.
$ identifies following year percentage change in the bank stock index where the spread in column (3) was equal to 2.25 percentage points or more. All of these percentage changes have been positive (so far) and in most cases greater than the percentage change associated with the S&P composite stock price index in column (5).
* identifies following years when the S&P composite stock price index performed better than the bank stock indexes in column (4).
Source of basic data: the cyclical indicator section of The Survey of Current Business, Standard & Poor's Security Price Index Record, Its Over the Counter and NASDAQ Daily Stock Price Records and Its Analysts' Handbook. Daily values for the NASDAQ bank stock indicator are now published in The Wall Street Journal and many other newspapers.
Go on to Essay 7:
Return to the Introduction