Get documents about "Editor's Introduction from the symposium Mutual Funds and
Document Sample
‘1
Editor’s Introduction
On March 29, 1994, the Federal Reserve Bank toward bond and equity mutual funds, With
of St. Louis hosted a symposium on the implica- offering rates on deposits decreasing, prospective
tions of rapid mutual fund growth for monetary returns on bond and equity funds often appeared
policy. From 1990-93, household holdings of to be three- or four-fold greater.
shares in bond and equity mutual funds increased The subsequent slow growth of M2 during a
at a record pace. During the same period, the period when many analysts perceived monetary
Federal Reserve’s primary monetary aggregate, policy as becoming increasingly expansionary
M2, grew much more slowly than suggested by led to doubts about its usefulness as a policy
its historical relationships to economic activity indicator. The monetary aggregates were offi-
and opportunity costs. Does the confluence of cially dc-emphasized as policy guides in Federal
these events suggest that M2 has become less Reserve Board Chairman Greenspan’s July 1993
useful as an indicator of the stance of monetary Humphrey-Hawkins Act testimony before
policy? Should it be replaced with a new aggre- Congress. At about the same time, a group of
gate that includes these mutual funds? economists at the Board of Governors completed
Financial innovation and advances in two studies evaluating whether M2 might usefully
technology change the structure of financial be replaced by a redefined aggregate that included
markets, alter the indicator properties of monetary bond and equity mutual fund shares. l’hese
aggregates, and give rise to pressures for their studies, and the five commentaries that appear
redefinition. When Regulation Q capped deposit here, were presented at the St. Louis symposium.
offering rates during the 1970s, for example, many Written by economists closely involved in policy
households learned that money market mutual analysis, the studies provide a unique perspective
funds provided an attractive alternative to holding on the range of issues that arise whenever it is
bank and thrift deposits. As a result, money suggested that a monetary aggregate be redefined.
market funds were included in M2 when it was Before a monetary aggregate may be used, it
redefined in 1980. During the 1980s, households must be measured. In the first article, Sean Collins
became increasingly familiar with financial and Cheryl Edwards discuss the measurement of
institutions other than banks and tbrifts. Mortgage a monetary aggregate M2+ that includes both M2
loans were increasingly originated by mortgage and shares in bond and equity mutual funds.
brokers, auto loans extended by finance compa- They first describe how interest in redefining
nies, and retirement funds held in self-managed an aggregate arises when its growth differs from
IRA and Keogh accounts. At the same time, the that suggested by its historical behavior, Although
mutual fund industry benefited from technical necessary, such deviant behavior may not be
progress that reduced the cost of servicing large sufficient unless there also has been significant
customer lists and managing portfolios of mar- innovation or technical progress in financial
ketable securities. markets since the previous redefinition of the
Thus, participants on both sides of the financial aggregates. The latter imparts an a priori reason-
markets seemed poised to react swiftly during ableness to suspicions that the array of money
the 1990s to the combination of a sharp decrease substitutes available to households and firms
in the overall level of market interest rates and a has expanded, or that the transaction costs of
record widening of maturity-related yield spreads. substituting among various alternatives has
Available survey and anecdotal evidence, as well decreased.
as negative statistical correlations in the aggregate The current M2 monetary aggregate is designed
data, suggest that households shifted savings to measure household and firm holdings of liquid
away from traditional depository institutions assets that are either available for spending now
NOVEMBER/DECEMBER 1994
4
or will become so in the near future. Retaining opportunity cost, The authors conclude that the
this focus in a new monetary aggregate that nonbank public’s holdings of bond and equity
includes bond and equity mutual funds requires fund shares seem to respond to both the spread
separating institutional holdings of mutual fund between the return on the mutual fund and the
shares from those held by firms and households, Treasury bill yield, and the change during the
as Collins and Edwards discuss in the latter half previous period in the overall level of market
of their article. Further, data on several items rates or equity prices.
that are not included in the new aggregate, such
In the latter part of their article, Orphanides.
as the M2-type assets owned by the bond and
Reid and Small examine the stability of the
equity funds and the amount of mutual fund demand for M2+ and its value as a leading indi-
shares held by households as illiquid retirement
cator for nominal GDP. Working within the linear
balances, are also required. These amounts are
error-correction framework developed by Board
subtracted from the new aggregate.
staff in previous money demand studies, they find
Collins and Edwards discuss several unique a reasonably good overall fit to the data. Yet, the
problems that arise while building M2÷ that estimated semi-elasticities of M2+ demand with
have no direct parallel in the current monetary respect to market yields and various measures of
aggregates. One is the inclusion of assets its opportunity cost appear highly sensitive to the
denominated in foreign currencies. All compo- form in which these variables enter the regression.
nents of the current official monetary aggregates Perhaps more disappointing, however, is their
(Ml, M2, MS and L) are denominated in U.S. conclusion that M2+ has not been a generally
dollars. Mutual funds that invest in foreign better indicator than M2 of movements in
currency-denominated assets, however, have nominal GDP growth during the lOgOs,
been among the most rapidly growing type of
In their commentary, William Barnett and
funds in recent years. Second, current M2
Ge Zhou interpret the definition of M2+ as a
includes only assets that are capital-certain or,
dynamic index number problem. When the
in other words, only assets whose value does
financial assets in an economy can be partitioned
not vary with the level of market interest rates.
into two non-overlapping groups—those that
Capital gains and losses are a significant factor
provide monetary (transaction) services and those
in changes in the value of bond and equity
that do not—the economy’s money stock is
funds, and present a thorny problem for both
correctly measured by summing the quantities
the construction and interpretation of the
of the assets in the former group. It is the essence
M2+ aggregate.
of financial innovation, however, to blur the
In the policy arena, monetary aggregates distinction between these groups and allow assets
may be valuable as either targets or indicators, in the latter category to provide monetary as well
with requirements for the former generally as non-monetary services. The authors show,
more stringent than for the latter. In both cases, however, that sequentially redefining a broad
the aggregate must have a reliable empirical monetary aggregate may cause the aggregate to
relationship to future economic activity. In move further away from, rather than closer to,
addition, to be useful as a policy target, the the economy’s true money stock, They laud the
demand for the aggregate must be a stable func- authors of the symposium’s two major papers
tion of a relatively small number of variables for grappling with the difficult issue of including
and the Federal Reserve must be able to control capital-uncertain assets in a monetary aggregate.
the aggregate’s growth. In the second article,
Jacob Dreyer doubts, however, that a new
Athanasios Orphanides, Brian Reid and David
M2+ monetary aggregate would be useful to
Small judge the M2+ monetary aggregate
policymakers. The necessity of estimating
relative to each of these criteria.
expected holding period yields likely precludes
Modeling the demand for a monetary aggre- obtaining useful estimates of a demand equation
gate requires identifying its close substitutes for the aggregate. He also argues that turnover
and, in turn, the opportunity cost of holding the rates suggest that bond and equity mutual funds
components of the aggregate rather than alterna- lack the necessary degree of “moneyness” for
tive assets. The inclusion of capital-uncertain them to reasonably be included in a monetary
assets in M2+ complicates calculation of its own aggregate along with the current components of
rate of return and identification of an appropriate M2. Although acknowledging that the transaction
FEDERAL RESERVE BANK OF ST. LOUIS
5
costs of buying and selling bond and equity funds indicator. In his view, the inverse correlation
have fallen sharply, he regards this change as between movements in market interest rates and
creating only the illusion, rather than the the value of the capital-uncertain assets included
substance, of moneyness. in M2+ likely precludes formulating any policy
In his commentary, John Duca attributes the feedback rules based on M2+.
slower-than-anticipated growth of M2 during the George Pennacchi suggests that households’
l990s to a combination of changes in banking increased holdings of mutual fund shares might
regulation and the pattern of market interest rates, be interpreted as a reaction to the narrower rate
Duca notes that significant shifts in financial spreads between liquid and time deposits at banks
intermediation also occurred in the lO7Os and during the l99Os. Reductions in transaction costs
l980s when government regulation interacted and improvements in computer technology have
with large movements in market interest rates. increased the liquidity of bond and equity mutual
Nevertheless, he concludes that the unprece- funds, making them more competitive with liquid
dented steep slope of the yield curve during a deposits in such an environment. He suggests
period when offering rates on bank and thrift that the lower turnover rates of bond and equity
deposits fell to their lowest levels in decades fund shares should not be taken as evidence that
likely was the primary motivation for households households have not responded to, and do not
to substitute holdings of bond and equity funds value, the increased liquidity of the funds.
for M2-type assets during the l990s. This Moreover, although the funds are subject to
substitution does not, by itself, call for a liquidity (interest rate) risk, economic theory
redefinition of M2, however. suggests that the expected holding period yields
Josh Feinman suggests that the dramatic of these mutual funds should not differ system-
shrinkage of the federal subsidy to the depository atically, on balance, from other market returns.
sector since 1990 seems to have played the larger
role in depressing M2 growth. He cites higher Finally, I want to recognize the economic analysts
deposit insurance premiums, new capital stan- in the Research Department of the Federal Reserve
dards and stricter supervision as examples. Bank of St. Louis who provided invaluable help
These changes reduced the incentive for deposi- in reviewing references and data for the sympo-
tories to pursue their traditional lending and sium papers: Heidi L. Beyer, Heather Deaton,
deposit-taking activities, resulting in an increasing Kelly M. Morris and Richard D. Taylor.
proportion of intermediation being conducted
through market instruments such as corporate
equity and commercial paper, popular investments Richard G. Anderson
of bond and equity mutual funds. He doubts that St. Louis, Missouri
M2+ could ever be useful as a policy target or November 1, 1994
NOVEMBER/DECEMBER 1994
