(quarterly averages)
GLOBALIZATION AND M3
WITH CRITICS LIKE THIS WHO NEEDS STRAWMEN?
Prepared by
The Northern Trust Company
All I can say is thank goodness for friends. Here it is Thursday afternoon, and I don't really have a topic for this week's PEC. Then a friend faxes me a commentary he had just received. And voila! A topic.
The commentary purported to show that a higher real U.S. fed funds rate causes the M3 money supply to grow faster. Why? According to this analyst, the higher real fed funds rate increases the demand for U.S. Dollar-denominated assets, and, somehow, this increases the supply of M3. Thus, according to this analyst, the faster M3 growth would be a sign of falling inflationary pressures, not rising inflationary pressures. So much for the notion that inflation is the result of too much money chasing too few goods. We truly are in a new era! Or are we?
We think that the analyst should have plotted the behavior of the dollar against the behavior of M3. He didn't in his commentary. We have. And after you look at our chart, we think that you will agree with us that there is no consistent relationship between the behavior of the dollar and that of M3. So, we think that the analyst doesn't even have his correlations right. Moreover, we do not think he understands how the supply of M3 is determined.
Plotted in Chart 1 are the quarterly observations of the real fed funds rate (the nominal funds rate minus the year-over-year percent change in the CPI) and the year-over-year percent changes in M3. Yes, starting in 1989, the real fed funds rate trended lower and so too did M3 growth. But notice that between 1987:Q4 and 1989:Q1, the real fed funds rate rose by 220-basis points while M3 growth at the end of the period was approximately equal to what it was at the beginning. What was the dollar doing during this period? As shown in Chart 2, it was rallying. In 1987:Q4, the trade- weighted dollar was down 13.8% from its year-ago level. By 1989:Q1, it was up 6.6%. Why didn't M3 growth pick up with the rise in the real fed funds rate and the rise in the dollar? Between 1995:Q3 and 1996:Q4, the real fed funds rate fell by 107- basis points. But the failing real fed funds rate did not produce a decline in M3 growth. On the contrary, M3 growth went from 5.5% to 6.8%. The dollar did rally, but evidently not because of the behavior of the real fed funds rate.
Chart 1
The Real Fed* Funds Rate Versus M3 Growth**
(quarterly averages)
*nominal fed funds rate minus yr./yr. pct. chg in CPI
**yr./yr. pct. chg. in M3 money supply
Let's take a closer look at Chart 2, which contains quarterly observations of year-over-year percent changes in the trade weighted dollar and M3 growth. Is there really a consistent relationship between the behavior of the dollar and M3 growth? Between 1987:Q4 and 1989:Q2, the dollar rallied big time, but M3 growth actually slowed. Same for the period from 1990:Q4 and 1991:Q3. Ditto for 1992:Q3 to 1993:Q3. Starting in the third quarter of 1995, there does seem to be a positive correlation between the appreciation of the dollar and growth in M3. Correlation, however, as any Stat 101 student knows, does not mean causation. And contrary to the analyst's “theory,” over most of this recent period, the real fed funds rate was falling, not rising.
Chart 2
Trade-Weighted Dollar Versus M3
(yr./yr. pct. chg. in qtrly. avg.)
Alas, this almost looks like another case where some “ugly” facts destroy a beautiful theory. We say “almost” because we don't think the theory would win any beauty contests. How can an increase the demand for U.S. Dollar-denominated assets increase the US M3 money supply? U.S. M3 is a close approximation for the liabilities side of the U.S. banking system. Standard double- entry accounting says that an increase in liabilities, i.e., M3, would have to be matched by an equal increase in assets–bank loans, investments and cash. Why should U.S. bank lending increase because investors want more dollar assets? If the demand for dollar assets is increasing, and increasing because U.S. real interest rates are rising relative to other countries', this means that global investors want to lend in the U.S., not borrow in the U.S. They would want to borrow from banks in those countries where real interest rates were lower. If anything, an increased demand for dollar assets would, all else the same, reduce U.S. M3, not raise it. As global investors lend more in the U.S., this would tend to drive down U.S. open market rates of interest. With a constant fed funds rate–the marginal cost of funds to banks–banks would have less incentive to buy more securities or make more loans. Because bank asset growth would slow, so too, would bank liabilities growth–i.e., M3 growth.
Currently, we are monitoring the behavior of U.S. M3 to give us an indication of whether the renewed weakening in the Japanese economy will lead to a weakening in the U.S. economy. Japanese government 10-year bond yields are now below 2.00%–the lowest yield for any major industrialized country since the Great Depression of the 1930's. If global investors start to buy more higher-yielding U.S. bonds rather than lower- yielding Japanese bonds, this could weaken the U.S. economy at a constant fed funds rate. If global investors want U.S. bonds, they have to buy them with U.S. Dollars. This would cause the dollar to appreciate, which would hurt U.S. export growth. Netting it all out, the world would want more of our bonds and fewer of our Bonnevilles. As mentioned above, as global investors lend more in the U.S., the consequent drop in open market interest rates would induce banks to cut back on their earning asset acquisitions. It would slow M3 growth.
How would a slowdown in U.S. bank lending and M3 lead to a slowdown in U.S. economic activity if nonbank lending is being substituted? Any member of the Austrian school of economics could answer this question. Nonbank lending is akin to what von Mises called “transfer credit.” The lender is cutting back on his current spending so that the borrower can increase his current spending. Spending is being transferred from the lender to the borrower. So, net, there is no new spending created by this transfer of credit. In our Japanese example, the rest of the world is cutting back on its current purchases of U.S.-produced goods, choosing to lend in the U.S. instead. This foreign lending to U.S. spenders at lower interest rates “displaces” U.S. bank lending. Bank credit is what von Mises referred to as “created credit.” When banks create credit, which they can do only with the “blessing” of the central bank, the recipient of that bank credit can increase his current spending, and no one else has to give up his current spending. In the case of created credit, and in contrast to transfer credit, net current spending will increase. Again, in our Japanese example, the consequent cutback in bank credit and M3 means that someone has cut back on his current spending. That someone is the rest of the world. So, the one-for-one substitution of nonbank credit for bank credit implies a slowdown in spending on U.S.- produced goods.
The author of the faxed commentary places the UK in the same camp as the U.S.–high real interest rates, appreciating currency and high broad money growth. He forgot one other attribute in common between the U.S. and the UK. They have the highest inflation rates in the G-7.
October 2, 1997Paul L. Kasriel
The Northern Trust Company
50 South LaSalle Street, Chicago, Illinois
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
