Introducing money.

Olivier Blanchard
April 2002

14.452. Spring 2002. Topic 6.

14.452. Spring, 2002

No role for money in the models we have looked at. Implicitly, centralized
markets, with an auctioneer:
Possibly open once, with full set of contingent markets. (Remember,
no heterogeneity, no idiosyncratic shocks. (Arrow Debreu)

More appealing. Markets open every period.

Spot markets, based on expectations of the future. For example, market for goods, labor, and one{period bonds. A sequence of temporary
equilibria (Hicks).
Still no need for money. An auctioneer. Some clearing house.
So need to move to an economy where money plays a useful role.
The ingredients.
No auctioneer. Geographically decentralized trades.
Then, problem of double coincidence of wants. Barter is not convenient. Money, accepted on one side of each transaction, is much more
so.
Two types of questions
Foundations
Why money? What kind of money will emerge?
Can there be competing monies?
Fiat versus commodity money?
Numeraire versus medium of exchange? Should they be the same, or
not?

14.452. Spring, 2002

Not just abstract, or history. The rise of barter in Russia in the 1990s.
\Natural" dollarization in some Latin American countries. \Units of account" in Latin America.
But most of the time, we can take it as given that money will be used
in transactions, that it will be at money, and that the numeraire and the
medium of exchange will be the same.
If we take these as given, then we can ask another set of questions:
How dierent does a decentralized economy with money look like?
What determines the demand for money, the equilibrium price level,
nominal interest rates?

How does the presence of money aect the consumption/saving choice?

Steady state and dynamic eects of changes in the rate of money
growth.

Start by looking at a benchmark model. Cash in advance.

Then, look at variations on the model; money in the utility function.
Then focus on price and ination dynamics, especially hyperination.

A cash in advance model

Think in terms of a decentralized economy (although we shall see that there

is an optimization problem which replicates the outcome).

1.1 The optimization problem of consumers/workers

Consumers/workers maximize:
i=1
X

E[

i=0

i U(Ct+i j t ]

14.452. Spring, 2002

subject to:
Pt Ct + Mt+1 + Bt+1 = Wt + t + Mt + (1 + it )Bt + Xt
and
Mt Pt Ct
Note that I ignore:
Uncertainty Because it is not central to the points I want to make.
But there is no problem in introducing it in the usual way.

The labor/leisure choice. It would be aected. But I leave it out

for simplicity. People supply one unit of labor inelastically.

The notation:
Pt is the price of goods in terms of the numeraire (the price level).
Mt and Bt are holdings of money and bonds at the start of period t.
Wt and t are the nominal wage and nominal prot received by each
consumer respectively.
it is the nominal interest rate (the interest rate stated in dollars, not
goods) paid by the bonds.
Xt is a nominal transfer from the government (which has to be there if
and when we think of changes in money as being implemented by distribution of new money to consumers).
Now turn to the assumptions underlying the specication:
Consumers care only about consumption. They do not derive utility
from money.

The rst constraint is the budget constraint. It says that nominal consumption plus new asset holdings must be equal to nominal income|

wage income (the labor supply is inelastic and equal to one) and prot

14.452. Spring, 2002

income|plus initial asset holdings, including interest on the bonds,

plus nominal government transfers.
If the only constraint was the rst constraint, then people would hold
no money: Bonds pay interest, money does not.

The second constraint explains why people hold money. It is known

as the cash in advance (CIA) constraint. People must enter the
period with enough nominal money balances to pay for consumption.
One story here. People are composed of a worker and a consumer.

The worker goes to work. The consumer goes to buy goods, and must
do this before the worker has been paid. So he must have sucient
money balances to nance consumption.

One can think of more sophisticated, smoother, formulations. For example: The cost of buying consumption goods is decreasing in money
balances. I shall return to this below.
Let t+i i be associated with the budget constraint, t+i i be associated
with the CIA constraint. Set up the Lagrangian and derive the FOC.
Ct :

U 0 (Ct ) = (t + t )Pt

Mt+1 :

t = (t+1 + t+1 )

Bt+1 :

t = (1 + it+1 )t+1

Interpretation of each.
Can combine them to get:
U 0 (Ct )
Pt
U 0 (Ct+1 )
=[
(1 + it+1 )]
1 + it
Pt+1
1 + it+1

14.452. Spring, 2002

Note that Pt =Pt+1 = 1 + t+1 . If we dene the real interest rate as:
(1 + rt+1 )

Pt
(1 + it+1 )
Pt+1

We can rewrite the rst order condition as:

U 0 (Ct )
U 0 (Ct+1 )
= (1 + rt+1 )
1 + it
1 + it+1
Interpretation.
Because people have to hold money one period in advance, the eective
price of consumption is not 1 but 1 + i.

Once we adjust for this price eect, then we get the same old relation,

between marginal utility this period, marginal utility next period, and
the real interest rate.

Note the role of both the nominal and the real interest rates. Note that
the nominal interest rate is constant, the equation reduces to the standard
Euler equation:
U 0 (Ct ) = (1 + rt+1 )U 0 (Ct+1 )
This characterizes consumption. Consumption behavior is very similar to that in the non monetary economy. Two dierences:
The relative price eect, if it is dierent from it+1 .
The fact that the rate of return on total wealth is lower (as some of
wealth does not yield interest), so the feasible level of consumption is
lower.
Given consumption, the characterization of the demand for money is
straightforward. The CIA holds as an equality:

14.452. Spring, 2002

Mt
= Ct
Pt
Pure quantity theory. No interest rate elasticity. Simple, but possibly
too simple. Will look at extensions below.

1.2 The optimization problem of rms

Firms produce goods using labor and capital. They pay labor a wage Wt .
They buy capital for use in the next period, and they nance these purchases
of capital by issuing nominal bonds.
Their nominal cash ow is thus given by:

t = Pt F (Kt ; Nt ) Wt Nt (1 + it )Bt + Pt (1 )Kt Pt Kt+1 + Bt+1

Cash ow is equal to production minus the wage bill, minus payment
of interest and principal on bonds issued last period, plus the value of the
remaining capital stock minus the value of the capital purchased, plus bond
issues.
The value of a rm is given by the present value of nominal cash ow,
discounted by the relevant nominal interest rate.
Vt = t + (1 + it+1 )1 Vt+1
The three FOC for rms are given by:
Nt :

Pt FN (Kt ; Nt ) = Wt

Bt+1 :

1=1

14.452. Spring, 2002

Kt+1 :

Pt = (1 + it+1 )1 [Pt+1 (1 + FK (Kt+1 ; Nt+1 ))]

Note the second FOC: It says that the amount of bonds issued by rms
is irrelevant. They could nance purchases of capital from current prot,
or partly through bond issues, or fully through bond issues. Their decisions
would be the same. (But, under our assumption, there are nominal bonds
in the economy, which makes it easier to think about the nominal interest
rate).
The third FOC can be rewritten as:
(1 + FK (Kt+1 ; Nt+1 )) = (1 + it+1 )

Pt
Pt+1

Or:
(1 + FK (Kt+1 ; Nt+1 )) = (1 + rt+1 )
Firms purchase capital to the point where the marginal product of capital
is equal to the real interest rate.

1.3 The equilibrium and the steady state

To close the model, we have that:
Nt = 1
Turn to the government budget constraint. Assume that the stock of
money is changed through transfers to people:
Xt = Mt+1 Mt
Putting things together, the dynamics of the economy are characterized
by the following equations:

14.452. Spring, 2002

U 0 (Ct )
U 0 (Ct+1 )
= (1 + rt+1 )
1 + it
1 + it+1
(1 + it ) = (1 + rt )(1 + t )

(1 + rt ) = 1 + FK (Kt ; 1)
Mt
= Ct
Pt
Kt+1 = F (Kt ; 1) + (1 )Kt Ct
I shall not attempt to look at dynamics, but just focus on steady state:
Suppose that the rate of growth of nominal money is equal to x, so
Xt
Mt
Mt
= (1 + x 1)
=x
Pt
Pt
Pt
.
In steady state, Ct ; Kt ; rt ; it ; t are constant, so:
From the FOC of the consumer, and the demand for capital by rms:
(1 + r) = 1 + FK (K; 1) = 1=
This is the same rule as without money: The modied golden rule.
In steady state, real money balances must be constant, so:
=x
Ination is equal to money growth. And so, i = + r = x + r. This one
for one eect of money growth on the nominal interest rate is known as the
Fisher eect.

14.452. Spring, 2002

10

Using these relations in the budget constraint of the consumer gives:

C = F (K; 1) K
So, on the real side, the economy looks the same as before. In addition people hold money. And ination proceeds at the same rate as money
growth. The fact that, in steady state, money growth has no eect on the
real allocation is refered to as the superneutrality of money.
Is this superneutrality a general result? I now explore an alternative
formalization.

Money in the utility function

The CIA constraint is too tight. One can clearly maintain a lower level of
real money balances is one is willing to go to the ATM machine more often.
More reasonable to assume that
The higher the level of real money balances one holds, the lower the
transaction costs, so the higher the level of output net of transaction
costs,
Or the higher the level of utility, again net of transaction costs.
One can formalize this explicitly, A dynamic Baumol Tobin model. This
is what is done by Romer (see original article or BF). Very useful, but a bit
heavy for here.
One can take short cuts. Real money balances in the production function, or in the utility function.
See eects of putting money in the utility function. (Sidrauski model).
So the optimization problem of consumers/workers is:
X

E[

i U(Ct+i ;

Mt+i
) j t ]
Pt+i

14.452. Spring, 2002

11

subject to:
Pt Ct + Mt+1 + Bt+1 = Wt + t + Mt + (1 + it )Bt + Xt
where, plausibly Um > 0 and Umc 0 (why?).
Let t+i i be the lagrange multiplier associated with the constraint.
Then the FOC are given by:
Ct :

Bt+1 :

Mt+1 :

Uc (Ct ;

Mt
) = t Pt
Pt

t = t+1 (1 + it+1 )

t = t+1 +

1
Mt+1
Um (Ct+1 ;
)
Pt+1
Pt+1

Interpretation. Can rewrite as:

An intertemporal condition:
Uc (Ct ;

Mt
Mt+1
) = (1 + rt+1 )Uc (Ct+1 ;
)
Pt
Pt+1

An intratemporal condition
Um (Ct ;

Mt
Mt
)=Uc (Ct ;
) = it
Pt
Pt

Interpretation. Note that the second says that the ratio of marginal
utilities has to be equal to the opportunity cost of holding money, so i, the
nominal interest rate.
If for example,
U(C; M=P ) = log(C) + a log(M=P )

14.452. Spring, 2002

12

Then,
Mt
Ct
= (a=)
Pt
it
This gives us an LM relation. (Indeed you can think of the rst condition
as giving us a simple IS relation, this giving us an LM relation. More on
this in the next lectures).
The demand for money is a function of the level of transactions, here
measured by consumption, and the opportunity cost of holding money, i.
Turn to steady state implications. (rms' side is the same as before).
1 + r = 1=

C = F (K; 1) K
Um (C;

M
M
)=Uc (C; ) = (x + r)
P
P

So, same real allocation again. And a level of real money balances inversely proportional to the rate of ination, itself equal to the rate of money
growth.
Dynamic eects? Yes. But nothing very exciting. Can make it more
exciting by modelling trips to the bank and having people come at dierent
times. Then, distribution eects. But does not seem to capture much of
what we actually observe.
So, bottom line: Money as a medium of exchange, without nominal
rigidities gives us a way of thinking about the economy, the price level, the
nominal interest rate, but not much in the way of explaining uctuations.
Very useful however when money growth and ination become high and
variable. Turn to this.

14.452. Spring, 2002

13

Money growth, ination, seignorage

Start with the money demand we just derived:

Mt
= Ct L(rt + te )
Pt
If money growth and ination are high and variable, M, P and e will
move a lot relative to C and r. So assume, for simplicity, that Ct = C, and
rt = r, so:
Mt
= C L(r + te )
Pt
This gives a relation between the price level and the expected rate of
ination. The higher expected ination, the lower real money balances, the
higher the price level.
This relation, together with an assumption about money growth, and the
formation of expectations, allows us to think about the behavior of ination.
This is what Cagan did. Looking at hyperinations, he asked;
Was hyperination the result of money growth, and only money
growth?

Why was money growth so high? Did it maximize seignorage. And if

not, then why?

Now have a quick look at his model (Read the paper, written in 1956. It
is a great read, even today). Also, read BF4-7, and BF10-2. What follows
is just a sketch.
Continuous time, more convenient here. Assume a particular form for
the demand for money:
M=P = exp( e )
So, in logs:

14.452. Spring, 2002

14

m p = e
Log real money balances are a decreasing function of expected ination.
Or dierentiating with respect to time:
x = de =dt
Assume that people have adaptive expectations about expected ination.
(In an environment such as hyperination, this assumption makes a lot of
sense. More on rational expectations below).
de =dt = ( e )
Money growth and ination
Suppose money growth is constant, at x. Will ination converge to
= x? To answer, combine the two equations above and eliminate de =dt
between the two, to get:
x = ( e )
This is a line in the (; e ) space. For a given x, de =d = (1)=,

so if < 1 the line is downward sloping. If > 1 upward sloping.

If < 1, then the equilibrium is stable. Start with x > 0, and = 0.

Then converge to = e = x.

If > 1, then not. Why?

Cagan estimated and , found < 1. Hyperination was the result
of money growth, not a bubble.
Seignorage

14.452. Spring, 2002

15

What is the maximum revenue the government can get from money
creation (called seignorage:
S

dM=dt M
dM=dt
=
= x exp(e )
P
M P

So, in steady state:

S = x exp(x)
So x = 1=
Much lower than the growth rates of money observed during hyperination.
But just a steady state result. Can clearly get more in the short run,
when e has not adjusted yet. This suggests looking at dierent dynamics:
Given seignorage, dynamics of money growth and ination.
Seignorage, money growth and ination
Start from:
S = x exp( e )
For a given S, draw the relation between e and x in e ; x space. Concave. Can cross the 45 degree line twice, once if tangent, not at all if no way
to generate the required seignorage in steady state.
Which equilibrium is stable? Using the equation for adaptive expectations and the money demand relation in derivative form:
de =dt = ( e ) = (x + de =dt e )
Or:
de =dt = 1=(1 ) (x e )

14.452. Spring, 2002

16

If two equilibria, lower one is stable. Start from it, and suppose S increases so no equilibrium.
Then, money growth and ination will keep increasing. This appears to
capture what happens during hyperinations.
Some other issues
Adaptive or rational expectations? (see BF 5-1)
Fiscal policy, and the eects of ination on the need for seignorage.
(See Dornbusch et al)

Unpleasant monetarist arithmetic? (see BF 10-2)

From Cagan:
Seven Hyperinations of the 1920s and 1940s

Country

Beginning

End

PT =P0 Average Monthly Average Monthly

Ination rate (%) Money Growth (%)

Austria

47

31

Germany Aug. 1922 Nov. 1923 1.0x1010

322

314

Nov. 1943 Nov. 1944 4.7x106

365

220

46

33

19,800

12,200

82

72

57

49

Greece

Oct. 1921 Aug. 1922

Hungary 1 Mar. 1923 Feb. 1924

70

44

Hungary 2 Aug. 1945 Jul. 1946 3.8x1027

Poland

Jan. 1923 Jan. 1924

699

Russia

Dec. 1921 Jan. 1924 1.2x105

PT =P0 : Price level in the last month of hyperination divided by the

price level in the rst month.