View original document

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

MILTON FR IED M A N
. . . . . .

C A P I T A ? -------------

E LY , V E R M O N T 0 S 0 4 4

Dear Arthur:

August 11, 1970
.

In accordancex w ith our co n v ersa tio n , l e t me " lis t the changes
in reserve requirem ents th a t i t seems to me would f i c i l i t a t e
the conduct o f monetary p o lic y .

2 . S h iftin g r ese rv e computations from a per calendar day b a sis
to a per working day b a s is . As m atters now stan d , Friday d e p o sits
and r ese rv e s aaoonbc have a w eight o f 3/7 in computing weekly
a v era g es. This makes what happend on Friday o f s p e c ia l importance
and in tro d u ces an unnecessary elem ent o f asymmetry in to d ay-today tr a n s a c tio n s . (Therw w i l l s t i l l remain an elem ent o f asymmetry
because 6 f in t e r e s t c a lc u la tio n s on fe d e r a l funds borrowed over
the w eek-end). In a d d itio n , i t i s a n ecessa ry p r e -r e q u is ite fo r
th e n ext item .

3. Stagger reserve settlement periods. If a weekly reserve period
is used, as now, divide all banks into five categories on some
random basis (final digifl of bank identifying number, or some
similar divice). Have one-fifth of banks use settlement per iod±
ending on Monday, one-fifth on Tuedday, etc. This ■JbttpHKwill greatly reduce need for Fed defensive operations within the
week. These are very large and contribute a great deal to the
aggregate operations of the Desk (see Cox and Leach on this).
1|. Re le n g th o f r ese rv e p erio d s, I see no o b je c tio n to keeping a
week or even len gth en in g to two weeks provided sta g g erin g i s
adopted. However, an a lte r n a tiv e to both item s 2 and 3 i s to
reduce r ese r v e p eriod to one day; i . e . , req u ire banks to meet
reserve requirem ents on a d a ily b a s is . For g iv en reserve
requirem ents, t h i s would req u ire banks to h old la r g e r average
r e s e r v e s . Hence, to be n e u tr a l, t h is change should in p r in c ip le
be accompanied by a d e c lin e in reserv e requirem ents. I have no
b a sis fo r judging the q u a n tita tiv e importance o f th is p o in t,
i . e . , whether the required d e c lin e i s so sm all as to be n e g lig ib le .
Some evid en ce could be obtained on t h is from th e e f f e c t o f the
s h i f t from a two-week reserv e p erio d fo r c o u tir y bantti to a
one-week p e r io d , though t h is change w eit along w ith the in t r o ­
duction tff a two-week la g . However, the la g a ff e c te d a l l banks
so in p r in c ip le th ere i s sep arate evidence on the e f f e c t o f a
shortened reserve p e r io d .

None of these items is of decisive importance. Yet I believe
feasible improvement in the technical arsangements can make an
appreciable contribution toward more precise control of monetary
aggregates.


B est


reg a rd s.

Yours,

„

---.------------

•./ t t

Photocopy from Gerald R. Ford Library

1 . E lim in a tio n o f f the two week la g between d e p o s its requiring
r ese rv e s and d e p o sits a t Federal Reserve Banks regarded as s a t i s ­
fy in g th ose requirem ents. T his* was introd uced about a year and
a h a lf ago and by now we have some evidence on i t s e f f e c t s . A
stu d en t o f mine has been working on t h is and fin d s th a t almost
a l l r e le v a n t v a r ia b le s are su b je c t* to g r ea ter week-to-week f l u c ­
tu a tio n s sin c e the change than b efore — which i s what might be
exp ected from gen eral c o n sid e r a tio n s.