View PDF

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

DRAFT

BOARO OF GOVERNORS
O F THE

F E D E R A L R E S E R V E S Y S TE M

Office Correspondence
To_

Staff

T L sir j£ ae'jecM

+W.

Subject- Analysis of San Francisco

Mr. Kichline

From.

Date June 3. 1977
Federal Reserve Bank Study
; _______ "The Case for M^"_________

Summary .

E-rnriCilCo
S c*

I +*

»-ep/ue

^i ~
uj-gik
1 0 0 % a1
/2

.

S o-S o

The San Francisco Federal Reserve Bank staff proposes

w/m
th a t in

of

<m
b
ncvTO^^ctp <55/Ue
*
a




th e y e a r ahead th e FOMC s h o u ld

and M2

and p la c e

t a r y a g g re g a te
t h is

abandon i t s

100 p e r c e n t w e ig h t on M2

ta rg e ts.

50-5 0 w e ig h t in g

i n d e t e r m in in g mone­

They a d vance s e v e r a l argum ents

to s u p p o rt

reco m m end ation :
F ir s t ,

e x a m in in g

th e r e s u l t s

th e r e l a t i o n

o f t h e ir

r e d u c e d fo rm a p p ro a c h f o r

b etw e en GNP and th e m o n eta ry a g g re g a te s

t h a t M2

has p r o v id e d c l o s e r

S e co n d ,

th e y p o i n t o u t th e a p p a r e n t bre akd o w n o f th e demand f u n c t i o n

for

Mj, s in c e m id -1 9 7 4 .

s h ift s

in

of

M2 .

w it h

T h ir d ,

s a v in g s d e p o s it s

t im e " com ponent o f
F in a lly ,

fo re c a s ts

o f GNP th a n M^ s in c e

suggest

th e y a rg u e

t h a t i n t e r e s t r a t e - in d u c e d

te n d t o be o f f s e t b y s h i f t s

le n d in g a r e l a t i v e

th e y a rg u e

th e M2

is

ju s t

1975Q1.

s t a b ilit y

to

in

th e " o t h e r

th e g ro w th

as c o n t r o l l a b l e

as M^

a m o n e ta ry b a se in s t r u m e n t .
T h i s memorandum b e g in s w it h

a t t r ib u t e s

o f M^ and M2 .

p r e s e n t e d by

a d is c u s s io n

of

A d i s c u s s i o n o f th e v a r io u s

th e c h a n g in g
argum ents

th e San F r a n c i s c o F e d e r a l R e s e r v e Bank s t a f f

B e g in n in g w it h

a c r it iq u e

of

t h e i r p r im a r y a rg u m e n ts,

is

fo llo w s .

a b r ie f

summary

en um erated b e lo w .

of

our p r in c ip le

*

C o n t r i b u t o r s to t h i s memorandum i n c lu d e J o h n K a lc h b r e n n e r , D a v id
L in d s e y , R ic h a r d P o r t e r , D a v id W y ss, F l i n t B r a y t o n , J e r r y E n z l e r ,
and D a v id P i e r c e .

f in d in g s




- 2 -

1.
w h ic h ,

u n lik e

show v e r y
The

We r e p o r t a l t e r n a t i v e

red uced

fo rm e q u a t io n s

f o r GNP

th e San F r a n c i s c o F e d e r a l R e s e r v e B a n k 's e q u a t io n s , ,

lit t le

d e t e r io r a t io n

in

e it h e r

th e

o r M2

v a r ia n t s .

te n d e n cy o f e i t h e r e q u a t io n t o u n d e r p r e d ic t GNP d oes n o t emerge

u n t il

1976,

a lt h o u g h

th e u n d e r p r e d ic t io n

is

s lig h t ly

g re a te r

fo r

th e M^ e q u a t io n .
2.

H ow ever,

th e a p p a r e n t s t a b i l i t y

e q u a t io n may be i l l u s o r y .

o f t h e ir

t o GNP

A d i f f e r e n t h i s t o r i c a l a d m i n is t r a t i o n

o f R e g u la t io n Q need n o t h ave p ro d u c e d as s t a b le a r e l a t i o n b etw e en
M2

and GNP.

M o reo v er, because

th e p ro b le m o f
M2

e q u a t io n s

th e y d i d n o t a d e q u a t e ly d e a l w it h

th e e n d o g e n e it y o f q u a r t e r l y m o n eta ry g ro w th ,

fu rth e r o v e rsta te

th e c lo s e n e s s

o f th e ^

t h e ir

t o GNP r e l a ­

t io n s h ip .
3.
M^ i n

On b a la n c e ,

f a v o r o f M2 .

c o n s id e r a t io n s
4.
c o n s id e r a t io n s

d oes n o t seem a d v is a b le

T h is p o s i t i o n

is

t o abandon

s u p p o r te d b o th by t h e o r e t i c a l

and e m p ir ic a l e v id e n c e .
As f i n a n c i a l i n s t i t u t i o n s

c o n t in u e

t o e v o lv e ,

s u g g e s t t h a t th e n a r r o w e s t m o n e ta ry a g g re g a te

m easu re a l l nonbank p u b l i c
fo rm o r

it

lo c a t io n .

t r a n s a c t io n s

b a la n c e s r e g a r d le s s

t h e o r e t ic a l
s h o u ld
of

t h e ir




- 3 Changing Attributes of

and

What distinguishes

from other monetary aggregates?

The traditional answer to this question was that
circulating medium of exchange in the economy.
the demand for

is the principle
It followed that

ought to depend importantly on the level of trans­

actions in the economy and, therefore, be closely related to GNP.
Until recently, empirical estimates of the demand for

were quite

stable, as were the separate estimates of the demand for the compo­
nents of M^, demand deposits

and currency.

theory linked the quantity of
rate and wealth channels.

A highly developed

to aggregate demand through interest

Large scale econometric models success­

fully incorporated empirical estimates of these relationships.
In view of these considerations, it made sense to set monetary policy
in terms of
Recent events--such as the introduction of NOW accounts
and credit union share drafts and other transaction-related balances
which may not be measured in M^--could be expected to have altered the
traditional relationships between

and GNP.

All or at least some

large portion of the balances held in these new forms are close
substitutes for

and, therefore, could be added to

the effective medium of exchange.
balances

are

in computing

Because several of these new

interest bearing, the net interest elasticity of demand

for such an extended transactions-related monetary aggregate may




- 4 -

differ from that of M^.

Nevertheless, the theoretical mechanism

linking this extended monetary aggregate to GNP is much like the
traditional theory linking

to GNP.

The research problem suggest­

ing itself involves accurately measuring the amounts of the various
components outstanding, estimating their demand elasticities with
respect to various interest rates and to GNP, and determining the
demand and supply characteristics of the total extended aggregate.
From the standpoint of theory, the relevance of M 2 is
less obvious, particularly in the currently evolving financial
environment.

Analytical considerations suggest that a stable relation­

ship between M 2 and GNP would only be observed by coincidence.
The first consideration is that the demand for several of the compo­
nents of M 2 is only peripherally related to GNP.

Admittedly, passbook

savings accounts may partly serve as transactions balances as a
result of telephone transfers and automatic bill payments and may,
hence, depend on GNP as well as various interest rates through the
public's demand function.

But the other two components of M 2 "

consumer-type certificates of deposit and large nonnegotiable CD's
at weekly reporting banks together with all large CD's at other
banks--less importantly share

this transactions-related character.

Demand for these components still depends on various interest rates
but depends more on the public's portfolio allocation response to
changes in wealth than to changes in GNP.




-

5

-

The foregoing argument implies that the degree of stimulus
inherent in any specific M 2 level depends crucially on the mix of
the M2 components.

Although M2 may be controllable, each of its

separate components cannot be simultaneously controlled with a
single reserve aggregate or Federal funds rate tool.

Thus, for

example, changes in the mix of other time and savings deposits caused
by a shift out of passbook accounts into nonnegotiable CD’s could
leave M 2 unchanged but nevertheless produce lessened economic stimulus.
Another consideration undermining the idea of a stable
relationship between M 2 and GNP is the similarity between components
of other time and savings deposits and various accounts excluded
from M 2 .

The public likely regards passbook accounts and consumer-

type CD's at MSB's and S&L's as very close substitutes to similar
balances at commercial banks.

Likewise, large negotiable CD's at

weekly reporting banks are close substitutes for all large CD's
at other banks.
Hence, shifts in the mix of similar accounts between
commercial banks and thrift institutions would change M 2 without
any accompanying change in monetary stimulus. If the Federal Reserve
acted to restore the original level of M 2 , it may inadvertently
be altering the degree of monetary stimulus.

Redefining the higher

order aggregates according to the type of deposit rather than accord­
ing to type of institution would help solve this problem.




- 6 -

Money and Income: Answers from a Reduced Form Approach
The San F r a n c i s c o
of

two r e d u c e d

fo rm e q u a t io n s

s u p e r io r it y

of ^

e v id e n c e

p re s e n te d

e q u a t io n s

is
in

s tu d y r e l i e s

o v e r M^ i n
to

T a b le 2 o f

to j u s t i f y

bn a c o m p a ris o n o f f o r e c a s t s
t h e i r c la i m

p r e d i c t i n g GNP.

th e e f f e c t

R a th e r c le a r - c u t

t h a t o f th e

th e A p p e n d ix ,

two f o r e c a s t i n g

th e seco n d e q u a t io n ,

in g M2

as an e x p la n a t o r y v a r i a b l e ,

d oes a b e t t e r

GNP i n

1975-76 th a n d o e s

e q u a t io n ,

a g g re g a te
"M 2

is

has s t e a d ily

th e r e c e s s io n
of

M^.

th e s e

th e

To i n f e r

f ir s t

fro m t h i s ,

o f re ce n t

as

jo b

c o n t a in ­

o f fo r e c a s t in g

i n w h ic h

th e e x p la n a t o r y

th e memorandum d o e s ,

th a t

o u tp e rfo r m e d M^ as a p r e d i c t o r o f n o m in a l GNP s in c e

tro u g h i n

1 9 7 5 .1 " r e q u i r e s

th a t

th e

s p e c if ic a t io n s

two e q u a t io n s be a d e q u a te , r e g a r d in g b o th

th e

o f a l l r e le v a n t v a r i a b l e s

and th e a p p r o p r ia t e

t h a t do a p p e a r .

t h e i r r e s u l t s w o u ld s im p ly be s p e c i f i c

That is ,

t o th e m odel s e le c t e d
f u n c t i o n a l fo rm o f

if

t h e ir

fo rm f o r

in c lu s io n

a d d in g a n o th e r v a r i a b l e
e q u a t io n s

a lt e r e d

th e v a r i a b l e s

o r c h a n g in g

th e

th e o b s e rv e d p e rfo rm a n c e

o f M^ and M2 .
M a i n t a in i n g

f id e lit y

to

th e s p i r i t

o f t h e ir

a p p ro a c h ,

Lindsey reports estimated reduced form equations which considerably
mute the force of their results.— ^
a f i s c a l and s t r i k e

v a r ia b le

in

p a s s a Chow t e s t o f c o e f f i c i e n t
is
1J

b o t h in c lu d e d o r e x c lu d e d

Both of his equations, containing

a d d it io n
s t a b ilit y

fro m th e

See D a v id L in d s e y , " R e l a t i o n o f M,
M r. K a lc h b r e n n e r , J u n e 1, 1977.

to M^ and M2

g ro w th ,

e a s ily

when th e p ost-1 9 7 3 Q 2 p e r io d

sam ple p e r io d .

M o re o v e r,

and M^ t o G N P ," memo t o




- 7 -

b o t h h i s M^ and M2
p r o p e r t ie s
r e s u lt s ,

to

th e

e q u a t io n s d i s p l a y

two r e p o r t e d b y th e San F r a n c is c o

th e M^ and M2

e q u a t io n s p r o v id e

o n e - q u a r t e r ahead f o r e c a s t s

in

to w a rd o v e r p r e d i c t i n g

and

r e s u lt s

o f f o r e c a s t o f GNP i n

s ta n d a r d e r r o r s

a b ilit y

of

o f e s t im a t e o f a b o u t

or

la t o r y

to a cco u n t f o r

s h ift

th e d i s t o r t i n g

in

e ffe c ts

p o s t- s a m p le f o r e c a s t i n g
In p a r t ic u la r ,

o f GNP f o r M^ and r i s e s

to

i n s t i t u t i o n a l in f lu e n c e s
in

advance,

t h is

e v id e n c e

a p p ro a c h f o r
1/

.5

w e l l b e lo w

th e
These

th e r e l a t i v e

fro m 1975Q4 on
o f re ce n t re g u ­

o u tp e r fo r m s h i s M2

as m easured by a l l

th e mean e r r o r i n
.22 f o r M2 .

th e summary

1976 f a l l s

to

.12 p e r c e n t

To th e e x t e n t t h a t th e s e

on money demand c o u ld h ave b een p r e d ic t e d
s u g g e s ts

t h a t an em p h asis upon M^ o v e r

w o u ld h ave b e e n w a r r a n te d o v e r th e p a s t
M o reo v er,

about

e s t im a t e d b y P a u lu s

and A x i l r o d —^ - - L in d s e y 1s M^ e q u a t io n a c t u a l l y

s t a t is t ic s .

and th e o n e-

has o c c u r r e d .

c h a n g e s and f i n a n c i a l i n n o v a t io n s - - a s

e q u a t io n i n

a c cu ra te

.8 p e r c e n t .

M o r e o v e r , when th e m o n e ta ry g ro w th r a t e s
a re a d ju s te d

. In h is

1976 d o es a b i a s

1976 i s

e q u a t io n s ,

do n o t s u g g e s t t h a t a s i g n i f i c a n t

fo r e c a s t in g

O n ly i n

e i t h e r e q u a t io n ,

.2 p e r c e n t o f GNP f o r h i s M^ and M2

e q u a t io n s '

s ta ff.

in d is t in g u is h a b ly

e rro r.

GNP show up i n

ahead mean e r r o r

s im u la t io n

1975 and 1976 as m easured by r o o t

mean s q u a re e r r o r o r mean a b s o lu t e

q u a rte r

s u p e r io r p o s t - s a m p le

two y e a r s .

s e v e r a l d ra w b a ck s in h e r e n t i n

e v a lu a t in g m o n e ta ry a g g re g a te s

See J o h n P a u lu s and S te v e n A x i l r o d ,
and F i n a n c i a l I n n o v a t io n s A f f e c t i n g
A g g r e g a t e s , " November 2 , 1976.

th e re d u c e d

form

s u g g e s t th e m s e lv e s .

'T le c e n t R e g u la t o r y Changes
th e G row th o f th e M o n e ta ry




- 8 -

The first applies particularly to equations involving M 2 :

the stable

relationship picked up by the regression equation may be in large
part illusory, in the sense that it reflects the effects of regulatory
actions, rather than underlying private economic behavior.

Since

the early 1960's major regulatory changes--such as the authorization
of new types of deposits and increases in interest rate ceilings-would have had the effect of raising M 2 demand and lowering M 2
velocity, other things equal.

However, these depressing effects

have been by and large just offset by the effects of the "other
things" working in the opposite direction.

As a result,

has

shown a remarkable secular stability since the early I960's.
It is difficult to estimate precisely the effects of these
regulatory actions.— ^
seen.

But a sense of their importance can be readily

Since the early 1960's,

has displayed a sawtooth pattern.

Relative peaks in the V2 sawtooth are typically associated with high
interest rates and with increases in Regulation Q interest rate ceil­
ings. Prior to the peak as other time and savings deposits grow more slowly
1/

There are a number of technical reasons for this statement,
but the principal one is statistical and has to do with the
relatively small number of observations representing RegulationQ
changes. An increase in the rate ceilings at commercial banks
has two immediate effects on the demand for M£. It lowers the
demand for demand deposits and increases the demand for time
and savings deposits. Presumably, the second effect dominates
and M 2 demands increase with an increase in the ceiling. It
has been argued that conventionally estimated demand deposit
demand functions overstate the effect on demand deposits as
a consequence of aggregation bias; see Farr, Porter, and Pruit,
"The Demand Deposit Ownership Survey," pp. 34-35, in Staff
Studies for the Committee on Monetary Statistics, forthcoming.




- 9 or even begin to decline, V2 increases sharply.

An increase in

Regulation Q ceiling rates then typically occurs, which permits
banks to increase offering rates on these deposits.

Over time funds

tend to be drawn away from competing market instruments and into
M2 .

Ultimately, the spurt in ^

*s reversed.

The attached memo

by Dave Wyss provides an estimate of such an effect for the January,
1970 increase in Regulation Q using the Board's quarterly econometric
m o d e lr^ The s im u la t io n s

in d ic a t e

th a t

w o u ld h ave re m a in e d s i g n i -

2/

ficantly higher if the Q ceilings rates had not been raised.—

The secular stability of V2 since the early 1960's is
replicated on average in the San Francisco Federal Reserve Bank
staff's M 2 equation because the sum of money and income coefficients
equalling less than unity just counteracts the positive time trend.
But because the equation does not capture the depressing impact
of the Regulation Q changes on V2 > it must necessarily underestimate
the effects working in the other direction of the "other things" mentioned
above.

It basically averages out the two effects.

Hence, if

Regulation Q were administered differently in the future than it
has been in the past, their M 2 equation should be expected to system­
atically misforecast GNP.
The second drawback of the reduced form approach is that
it can not solve the problem of simultaneous interaction between
\_f

2/

David Wyss, "Effect of Regulation Q Changes in the MPS Quarterly
Model," memo to Mr. Porter, June 1, 1977.
The effect on V2 would have been larger if the disaggregation
argument of the previous page's footnote is valid.




- 10 -

the monetary aggregate and GNP.

The Federal Reserve's control over

; quarterly monetary growth is not exact.

Nor, throughout most of

the sample period, has the Federal Reserve attempted to closely
control the monetary aggregates, and even less so for M 2 than for
M^.

As a result, quarterly money growth would, through the demand

side, generally have an endogenous component as it is affected by
current GNP.

Both the San Francisco Federal Reserve Bank's equations

and the Lindsey equations contain a current quarter's money term
to help explain current GNP.

Despite their protestations to the

contrary, the San Francisco Federal Reserve bank staff's causality
tests do not rule out a causal influence going from current GNP
to current M^ and

The existence of simultaneous feedback

from income to money, via a money demand function would imply an
artificially good fit of their reduced forms and a bias in at least
the coefficient of the current quarter's money aggregate in their
equations and in all the monetary aggregate coefficients in the
Lindsey equations fit by Almon lags.

The effects on GNP of deliberate

policy manipulation of the aggregate could prove to be at odds with
2/

Their memorandum asserts that "the causal tests are important
because the lack of feedback allows us to assign the relatively
strong contemporary correlation between income and money to
the relations explaining income with money" (p. 27). This
statement is in error. Tests of causality in the Granger sense
are silent on the question of the direction of contemporaneous
causality. The San Francisco Federal Reserve Bank staff's tests
suggest only that current money is independent of lagged GNP,
not current GNP. The statement that "R^ will be zero if and
only if P^ is zero" (p. 25) is not true because only a contem­
poraneous relationship may exist.




- l i ­

the forecasts of either sets of equations.
Federal1Reserve switched to an ^

In particular, if the

targeting procedure and, thus,

altered the behavior of this aggregate over time, there is every
reason to expect that the coefficients of the "reduced form" equa­
tions would be systematically altered and that the predictive ability
of the estimated equations would break down.— ^
The third limitation of the reduced form approach for
determining which aggregate provides closer projections of GNP is
procedural in nature.

The Board staff and the FOMC do not rely

solely upon formal econometric model simulations in making projections
or determining policy actions.

A great deal of other information

is brought to bear in an attempt to compensate for emerging errors
in econometric relationships.

A wide variety of techniques, some

of them purely judgmental, are employed in determining projections
of nominal GNP.

It is highly uncertain whether such "pooled" fore­

casts would have been better if the San Francisco Federal Reserve
Bank's reduced form with M 2 had been included as one of the inputs
and the M^ equation had been excluded.
A related point is that the theory of optimal forecasting
suggests that both the M 2 and M^ equation's predictions should be

2/

considered in arriving at a pooled GNP forecast.—
if

2_f

The optimal

See Robert Lucus, "Econometric Policy Evaluation: A Critique,"
in Carnegie-Rochester Conferences Series on Public Policies, Vol. 2,
See-John Kalchbrenner and Peter Tinsley, "On Filtering Auxiliary
Information in Short-Run Monetary Policy," Board of Governors
of the Federal Reserve System mimeograph.




- 12 -

weights to apply to each equation's predictions depends on the vari­
ance and covari«nce properties of the equation's5errors.

The point

to be made here is that only in pathological cases would a weight
of 100 percent be assigned to one equation's predictions and a weight
of zero be assigned to another equation's predictions.

Errors in Estimates of the Demand for Money
It is unquestionably true that the standard specification
of the demand for demand deposits as in the Board staff FMP quarterly
model as well as alternative specifications— ^

fail to explain the

slow demand deposit growth relative to changes in GNP and interest
rates since mid-1974.

The Paulus and Axilrod adjustments are by

no means sufficient to account for the errors.

Unfortunately, to

our knowledge the explanation for the entire shortfall in M^ has
not been provided.

i

However, Goldfeld has fit demand equations for M 2 and
other time and savings deposits, which predict recent levels quite
closely when adjusted by the Paulus and Axilrod figures.

But

the implicit M^ equation embedded in Goldfeld's M 2 and time

\l

See, for example, Jared Enzler, Lewis Johnson, and John Paulus,
"Some Problems in Money Demand," BPEA, 1:1976, pp. 261-80 and
Stephen M. Goldfeld, BPEA, 3, 1976, pp. 683-739.

\




- 13 and savings functions is rather odd, to say the least, and we are
inclined to doubt that Goldfeld's M 2 equation is much more than a
statistical fluke.— ^
Moreover, regardless of one's view of Goldfeld's M 2 demand
equation, its relationship to the San Francisco Federal Reserve
Bank's M 2 reduced form equation remains totally unspecified.

The

properties of a structural equation do not necessarily apply to

a reduced form and vice versa.
Changing Response of the Time and Savings Component of M 2 to Interest
Rate Movements
As noted above, the components of M 2 are not homogeneous
b u t a re

im p e r f e c t s u b s t i t u t e s

F e d e r a l R e s e r v e Bank s t a f f
h e t e r o g e n e it y b y a r g u in g

f o r each o th e r .

a p p e a rs

The San F r a n c i s c o

t o make a v i r t u e

t h a t th e i n c l u s i o n

of

ou t o f t h is

some l a r g e C D 's i n

Mj has recently been beneficial over the last 1% years.
1 /

Using

Shown below are the elasticities of M^ with respect to real
income and the Treasury bill rate implied by the Goldfeld
equations in Table 16 (p. 723);
Real Per Capita GNP

Bill Rate

Short Run

Long Run

Short Run

Long Run

.12

1.23

.003

-.027

These are based on weights of 3/7 for M]^ and 4/7 for other
time and savings. Note that the bill rate has a positive
coefficient in the short run and a very small elasticity in
the long run. Also, the elasticity with respect to real
income is greater than one suggesting M]_ is a luxury good.
I n m o s t periods, this equation would give rather poor predic­
tions .




- 14 this reasoning, however, the impact on M 2 growth of the runoff of
large CD's included in M 2 has roughly been offset by the increase
in interest-sensitive business savings deposits at commercial banks.
As Brayton indicates in the attached memo,— ^

there is no reason

to expect that this offsetting tendency will continue to hold in
the future.

The large CD component of M 2 is roughly ten times

greater than business passbook accounts.

Moreover, a large fraction

of the growth in these business accounts was a one-time occurrence
reflecting portfolio adjustments resulting from the new opportunity
to hold these accounts.

Thus, the future behavior of M 2 may be

influenced more than in the recent past by its large CD component.
Thus, M 2 may come to behave more like the past behavior of M^ than
that of M 2 .
The upshot of this line of reasoning is that one should
look at the relationship between M^ and GNP as a guide to the future
strength of relationship between M 2 and GNP.

When this is done,

Brayton finds that of the four aggregates, (M^, M 2 , M^, and M^),
M^ performs worst both within and outside the sample period.

1J

See Flint Brayton, "Comments on the San Francisco Memo," memo
to Mr. Kalchbrenner, May 25, 1977.


Federal Reserve Bank of St. Louis, One Federal Reserve Bank Plaza, St. Louis, MO 63102