Prepared by

The Northern Trust Company

The September M3 money supply data were released today. On a year-over-year basis, this broadest of the Fed's money supply definitions increased by 8.6%–the largest increase since February 1987's 8.6% increase. Using the core CPI as the deflator, real M3 grew at an annualized rate of 7.2% in the third quarter versus the second quarter. In the second quarter, real M3 growth was only 4.3%. You have to go back to the second quarter of 1984 to find real M3 growth comparable to that of the just-completed third quarter. Both the recent acceleration in real M3 growth and its magnitude suggest continued relatively strong economic growth in the quarter ahead.

Although M3 growth started accelerating sharply in 1995, as pointed out to me by a colleague, growth in nonfinancial sector debt has been decelerating in recent quarters. With regard to the future implications for economic growth and inflation, should we pay more attention to the acceleration in M3 growth or the deceleration in nonfinancial sector debt growth? My answer is M3 growth. I'll explain why in this commentary.

The year-over-year behaviors of real M3 and real nonfinancial sector debt are shown in Chart 1. Because the core CPI is not a consistent historical series, I have used the core Personal Consumption Expenditure deflator to calculate real values of M3 and nonfinancial sector debt. As shown in Chart 1, growth in real nonfinancial sector debt since 1992 peaked in 1996:Q1 at 3.1%. By 1997:Q2, the latest data point available for this series, real debt growth had slowed to 2.5% growth. In contrast, real M3 growth has generally trended up in this period, reaching 5.1% in 1997:Q2–its fastest growth since the 6.6% posted in the first quarter of 1985.

The same two series are plotted in Chart 2, but over a longer time period. Notice several things in Chart 2. First, real M3 growth tends to lead real nonfinancial sector debt growth. M3 is, in effect, a subset of nonfinancial debt. Why? Because M3, accounting for most of the liabilities of the banking system, is a proxy for the assets of the banking system by virtue of the fact that assets equal liabilities plus net worth. Most of the assets of the banking system are bank credit–loans and securities. Bank loans to the nonfinancial sector and bank holdings of securities issued by the nonfinancial sector are part of nonfinancial debt. Implicitly then, a good part of M3 is represented in nonfinancial debt. The fact that M3 leads nonfinancial sector debt suggests that M3 “causes” nonfinancial debt, not the other way around.

The second thing to notice is that from the early 1960's through the early 1970's, real M3 growth generally exceeded real debt growth. This was a period of relatively strong real economic activity. But from 1982 through almost the end of 1995, real nonfinancial debt growth consistently exceeded real M3 growth. This was a period of relatively slower economic growth.

Third, notice that from the early 1960's through the early 1980's, there were several instances when real M3 growth was rising rapidly from below real debt growth to above it. Late 1967 to early 1968 was one of these instances. For now, just note this. What happened to GDP growth shortly after these instances will be discussed a little later in this commentary.

Next, however, I want to explore the reason for the recent slowdown in real nonfinancial sector debt growth. Chart 3 helps us in this regard. As the chart shows, a sharp deceleration in net federal government debt issuance in 1996 and in the first half of 1997 is largely responsible for the slowing in total real nonfinancial debt growth. And why has government debt been growing more slowly of late? Primarily because federal tax revenues have been growing so rapidly. And why have tax revenues been growing so rapidly? Because economic growth has picked up and because the stock market has been soaring, generating strong capital gains tax payments. Also, real direct government spending on goods and services has slowed, reducing the need for government borrowing. Despite this slowdown in direct government spending, real GDP growth strengthened in 1996 and 1997. So, the slowdown in government debt and spending growth evidently hasn't hampered economic growth. Real private debt growth has slowed, too, but not nearly to the extent that government debt has. Could the acceleration in real M3 growth have contributed to the strength in economic activity and the rally in the stock market, both of which were responsible for the fall in the federal budget deficit, and, thus, the slowdown in federal debt issuance?

As we mentioned, there have been instances, similar to the current situation, when real M3 growth moves up from below real nonfinancial sector debt growth to above it. What happened to real GDP growth after this? If real GDP growth accelerated, then this would suggest that real M3 growth is more important in determining the pace of economic activity than real debt growth. To investigate this hypothesis, I constructed a variable I'll call “relative real M3 growth” defined as real M3 growth minus real nonfinancial sector debt growth. Chart 4 shows the behavior of “relative real M3 growth” and real GDP growth. The shaded areas in Chart 4 are instances when M3 growth has passed from being below real nonfinancial sector debt growth to above it–i.e., when “relative real M3 growth” has moved from negative territory to positive territory. What has happened to real GDP growth in these instances? It has accelerated. This finding, plus the finding that real M3 growth leads real nonfinancial sector debt growth strongly suggest that real M3 growth is more important in predicting real GDP growth than is real nonfinancial sector debt growth. So the fact that real nonfinancial sector debt growth has slowed recently is not particularly relevant with regard to future real GDP growth. The fact that real M3 growth is accelerating suggests that real GDP growth will stay relatively strong. If the level of real GDP is rising relative to the level of potential real GDP, and if this GDP “gap” is positively related to future inflation, then the acceleration in real M3 growth is suggesting that the risks are growing that inflation will be accelerating ahead. Again, the deceleration in real nonfinancial debt has little relevance to the future behavior of inflation.

*Real M3 growth minus real nonfinan. debt growth

Why does the behavior of M3 dominate the behavior of nonfinancial sector debt in terms of explaining economic activity and inflation? As I indicated above, M3 growth is a proxy for bank credit growth. When the banking system creates credit, on net, economic activity is unambiguously stimulated. Why? Because the recipient of the net increase in bank credit uses the credit to buy something. No one else in the economy has to cut back on his or her spending. So, economic activity increases. When the banking system creates a net increase in credit, it is akin to a counterfeiter printing up a batch of new greenbacks. The counterfeiter can buy something with her newly “minted” money. When she does, economic activity increases because no one else will cut back on his or her spending.

When nonbank credit declines, it is ambiguous as to whether economic activity declines. Suppose I pay back a loan to you, an individual. Where do I get the funds to make this repayment? I could cut my current spending. But if you use the loan repayment to increase your current spending, then net, net, spending in the economy does not decrease because of my loan repayment. Perhaps I reduce my holdings of bank deposits in order to repay you. (Note that bank deposits in the aggregate do not fall as a result of this, they just get redistributed from me to you.) In this case, I do not cut back on my current spending. If you increase your current spending with the proceeds from the loan repayment, then spending in the aggregate goes up. Only if I repay my loan to you by cutting my current spending and you wish to hold the proceeds in the form of higher bank deposits will net spending go down with a fall in nonbank credit. So, sometimes a decrease in nonbank credit leads to an increase in net spending in the economy, sometimes it leads to no net change in spending and sometimes it leads to a net decrease. But a net increase in bank credit unambiguously leads to an increase in net spending. And that's why the behavior of M3 dominates that of nonfinancial sector debt with respect to future economic activity and inflation.

To paraphrase Robert Frost, two aggregates have diverged, M3 and nonfinancial sector debt. M3 is the aggregate “less traveled” these days. But I will take it, i.e., M3. And I think it will make all the difference in my forecast of economic activity and inflation.

October 16, 1997Paul L. Kasriel

The Northern Trust Company

50 South LaSalle Street, Chicago, Illinois

Financial Commentary

COMMODITY FUTURES FORECAST WEEKLY REPORT | DEAN WITTER FUTURES COMMENTARY
THE LATEST COMMITMENT OF TRADERS ANALYSIS FOR CURRENCY CONTRACTS
MYERS ON FUTURES | NIKKO MARKET COMMENTS
SCHAEFFER'S RESEARCH REVIEW
FED PRICES SLIDE BACK TO $65
MONEY VERSUS DEBT TWO AGGREGATES DIVERGED
STOCKMARKET CYCLES | WEEKLY OUTLOOK
INTEREST RATE WATCH | THE ALLENDALE ADVISORY REPORT
NIKKO MARKET COMMENTS | THE REAPER

Consensus National Futures and Financial On Line Index

Copyright 1997, by Consensus Inc.  All American and Pan American rights Reserved. editor@consensus-inc.com

Hosted by:
One Crossroads Place
610 West Maple Ave, Suite WWW
Independence, MO 64050
(816) 252-4080
sysop@kcmo.com

wmeubank@ocp.kcmo.com