macroeconomics: a dynamic general equilibrium...

Macroeconomics: A Dynamic General EquilibriumApproach

Mausumi Das

Lecture Notes, DSE

February 2-22, 2016

Das (Lecture Notes, DSE) DGE Approach February 2-22, 2016 1 / 105

Modern Macroeconomics: the Dynamic GeneralEquilibrium (DGE) Approach

As we have stated before, modern macroeconomics is based on adynamic general equilibrium approach which postulates that

Economic agents are continuously optimizing/re-optimizing subject totheir constraints and subject to their information set up. They optimizenot only over their current choice variables but also the choices thatwould be realized in future.All agents have rational expectations: thus their ex ante optimal futurechoices would ex post turn out to be less than optimal if and only iftheir information set was incomplete and/or there are some randomelements in the economy which cannot be anticipated perfectly.The optimal choice of all agents are then mediated through themarkets to produce an outcome for the macroeconomy.


Modern Macroeconomics: DGE Approach (Contd.)

This approach is ‘dynamic’because agents are making choices overvariables that relate to both present and future.

This approach is ‘equilibrium’because the outcome for themacro-economy is the aggregation of individuals’‘equilibrium’behaviour.

This approach is ‘general equilibrium’because it simultaneouslytakes into account the optimal behaviour of diiferent types of agentsin different markets and ensures that all markets clear.


DSG Approach vis-a-vis Traditional Macroeconomics

The need to build macro-models based on internally-consistent,dynamic optimization exercises of rational agents arose once it wasrealized that ad-hoc micro founations for the aggregative system maynot be consistent with one another.

This begs the following question: Why do we need such optimizationbased micro-founded framework at all?

Why cannot we just take the aggregative equations as arepresentation of the macro-economy and try to estimate variousparameters, using aggregative data?

After all, if we are ultimately interested in knowing how themacroeconomy would respond to various kinds of policy shocks, allthat we need to do is to econometrically estimate the parameters ofthe aggregative system.

Then from the estimated parameter values or coeffi cients, we canpredict the implcations of various policy changes.


DSG Approach vis-a-vis Traditional Macroeconomics:(Contd.)

Indeed, this is exactly how macroeconomic analysis was conductedtraditionally!As we have already seen, traditional macroeconomics was based onsome aggregative behavioural relationship (e.g., Keyensian SavingsFunction - which postulates a relationship between aggregate incomeand aggregate savings; Phillips Curve - which posits a relationshipbetween umployment rate and inflation rate).Often one would construct detailed behavioural equations for themacroeconomy and would try to estimate the parameters of theseequations using time series data.To be sure some of these equations would be dynamic in nature.But optimization over time was not considered to be important oreven relevant. Indeed, the concept of optimization itself - either byhouseholds or firms or even government - was rather alien in the fieldMacroeconomics.


Lucas Critique: Optimization Comes to the Fore

The need to build macro models based explicitly on agents’optimization exercises came from the so-called Lucas Critique.

Lucas (1976) argued that aggregative macro models which areestimated to predict outcomes of economic policy changes are uselesssimply because the estimated parameters themselves may depend onthe existing policies.

As the policy changes, these coeffi cients themsleves would change,thereby generating wrong predictions!

His solution was to build macroecnomic models with clear andspecific microeconomic foundations - models that are explicitly basedon agents’optimization exercises.


Lucas Critique: Optimization Comes to the Fore (Contd.)

Such models will enable us to differentiate between true parameters- primitives like tastes, technology etc - which are independent of thegovernment policies, and variables that treated as exogenous by theagents but are actually endogenous and are influenced by governmentpolicies.

Moreover such models would take into account agents’expectationsabout government policies.

Predictions based on such microfounded models would be moreaccurate than the aggregative models which club all the trueparameters as well as other policy-related parameters together.


How Does Micro-foundation Help? An Example

Let us see exactly what Lucas critique means in the context of asimple example.

Consider the Keynesian savings function, specified as an aggregativerelationship:

St = α1 + α2Yt + εt

An aggregative macro model would take the above behaviouralrelationship as given and would estimate the coeffi cients α1 and α2from data.

We have already provided a micro-foundation for this kind ofKeynesian Consumption/Savings function.

Let us re-visit that exercise.


Micro-foundation of Keynesian Savings Function:

We assume that the economy consists of a finite number (H) ofidentical households. We can then talk in terms of a ‘representative’household.Let us define a 2-period utility maximization problem of therepresentative household as:

Max .{ct ,ct+1}

log(ct ) + β log(ct+1)

subject to,

(i) Ptct + st = yt ;

(ii) Pet+1ct+1 = (1+ r et+1)st + yet+1.

From (i) and (ii) we can eliminate St to derive the life-time budgetconstraint of the household as:

Ptct +Pet+1ct+1(1+ r et+1)

= yt +y et+1

(1+ r et+1)


Micro-foundation of Keynesian Savings Function: (Contd.)

From the FONCs:ct+1βct

= (1+ r et+1)(PtPet+1

).

Solving we get:

Ptct =1

(1+ β)

[yt +

y et+1(1+ r et+1)

]Thus

st =β

(1+ β)yt −

1(1+ β)

[y et+1

(1+ r et+1)

]Aggregating over all households:

St =β

(1+ β)Yt −

1(1+ β)

[Y et+1

(1+ r et+1)

]Notice that an aggregative model would equate β

(1+β)to α2 and

− 1(1+β)

[Y et+1

(1+r et+1)

]to α1.


Micro-foundation of Keynesian Savings Function: (Contd.)

While the coeffi cient α2 is indeed based on true parameters(primitives) and would therefore be unaffected by policy changes,coeffi cient α1 is not.

Any policy that changes the household’s expectation about its futureincome or future rate of interest rate would affect α1.

Thus predicting outcomes of such a policy based on the estimatedvalues of the aggregative equations would be wrong.


Modern Macroeconomics: DGE Approach

The Lucas critique and the consequent logical need to develop aunified micro-founded macroeconomic framework which would allowus to accurately predict the macroeconomic outcomes in response toany external shock (policy-driven or otherwise) led to emergence ofthe modern dynamic general equilibrium approach.

As before, there are two variants of modern DGE-based approach:

One is based on the assumption of perfect markets (theNeoclassical/RBC school). As is expected, this school is critical of anypolicy intervention, in particular, monetary policy interventions.The other one allows for some market imperfections (theNew-Keynesian school). Again, true to their ideological underpinning,this school argues for active policy intervention.


Modern Macroeconomics: DGE Approach (Contd.)

However, both frameworks are similar in two fundamental aspects:

Agents optimize over infinte horizon; andAgents are forward looking, i.e., when they optimize over futurevariable they base their expectations on all available information -including information about (future) government policies. In otherwords, agents have rational expectations.

We now develop the choice-theoretic frameworks for households andfirms under the DGE approach.

As before, we shall assume that the economy is populated by Hidentical households so that we can talk in terms of a representativehousehold.


Household’s Choice Problem under Perfect Markets:Infinite Horizon

Let us examine the consumption-savings choices of the representativehousehold over infinite horizon when markets are perfect.To simplify the analysis, we shall only focus on the consumptionchoice of the household and ignore the labour-leisure choice (for thetime being).At any point of time the household is endowed with one unit oflabour - which it supplies inelastically to the market.We shall also ignore prices and the concomitant role of money andfocus only on the ‘real’variables.Let at denote the asset stock of the household at the beginning ofperiod t.Then Income of the household at time t: yt = wt + rtat .We shall assume that savings of an household in any period areinvested in various forms of assets (all assets have the same return),which augments the household’s asset stock in the next period.


Household’s Choice Problem: Infinite Horizon (Contd.)

If we do not allow intra-household borrowing, then the representativehousehold h’s problem would given by:

Max .{cht }∞

t=0,{aht+1}∞

t=0

∞

∑t=0

βtu(cht); u′ > 0; u′′ < 0

subject to

(i) cht 5 wt + rtaht for all t = 0;(ii) aht+1 = wt + rtaht + (1− δ)aht − cht ; aht = 0 for all t = 0; ah0 given.



Notice that that the household is solving this problem at time 0.Therefore, in order to solve this problem the households would haveto have some expectation about the entire time paths of wt and rtfrom t = 0 to t → ∞.We shall however assume that households’have rational expectations.In this model with complete information and no uncertainty, rationalexpectation is equivalent to perfect foresight. We shall use these twoterms here interchangeably.

By virtue of the assumption of rational expectations/perfect foresight,the agents can correctly guess all the future values of the marketwage rate and rental rate, but they still treat them as exogenous.

As atomistic agents, they belive that their action cannot influencethe values of these ‘market’variables.



Notice that once we choose our consumption time path{cht}∞t=0, the

corresponding time path of the asset level{aht+1

}∞t=0 would

automatically get determined from the constraint functions (and viceversa).So in effect in this constrained optimization problem, we only have tochoose one set of variables directly. We call them the controlvariables. Let our control variable for this problem be

{cht}∞t=0 .

We can always treat c0, c1, c2,......as independent variables and solvethe problem using the standard Lagrangean method.The only problem is that there are now infinite number of such choicevariables (c0, c1, c2,....., c∞) as well as infinite number of constraints(one for each time period from t = 0, 1, 2.....,∞) and things can getquite intractable.Instead, we shall employ a different method - called DynamicProgramming - which simplifies the solution process and reduces it toa univariate problem.


Dynamic Optimization in Discrete Time: DynamicProgramming

Consider the following canonical discrete time dynamic optimizationproblem:

Max .{xt+1}∞

t=0,{yt}∞t=0

∞

∑t=0

βt U (t, xt , yt )

subject to

(i) yt ∈ G (t, xt ) for all t = 0;(ii) xt+1 = f (t, xt , yt ); xt ∈ X for all t = 0; x0 given.

Here yt is the control variable; xt is the state variable; U representsthe instantaneous payoff function.(i) specifies what values the control variable yt is allowed to take (thefeasible set), given the value of xt at time t;(ii) specifies evolution of the state variable as a function of previousperiod’s state and control variables (state transition equation).


Dynamic Programming (Contd.)

It is often convenient to use the state transition equation given by (ii)to eliminate the control variable and write the dynamic programmingproblem in terms of the state variable alone:

Max .{xt+1}∞

t=0

∞

∑t=0

βtU (t, xt , xt+1)

subject to

(i) xt+1 ∈ G (t, xt ) for all t = 0; x0 given.We are going to focus on stationary dynamic programming problems,where time (t) does not appear as an independent argument either inthe objective function of in the constraint function:

Max .{xt+1}∞

t=0

∞

∑t=0

βtU (xt , xt+1)

subject to(i) xt+1 ∈ G (xt ) for all t = 0; x0 given.


Stationary Dynamic Programming: Value Function &Policy Function

Ideally we should be able to solve the above stationary dyanamicprogramming problem by employing the Lagrange method. Let{x∗t+1

}∞t=0 denote such a solution.

We can then write the maximised value of the objective function as afunction of the parameters alone, in particular as a function of x0 :

V (x0) ≡ Max .{xt+1}∞

t=0

∞

∑t=0

βtU (xt , xt+1) ; xt+1 ∈ G (xt ) for all t = 0;

=∞

∑t=0

βtU (x∗t , x∗t+1) .

The maximized value of the objective function is called the valuefunction.The function V (x0) represents the value function of the dynamicprogramming problem at time 0.


Value Function & Policy Function (Contd.)

Suppose we were to repeat this exercise again the next period i.,e. att = 1.Now of course the time period t = 1 will be counted as the initialpoint and the corresponding initial value of the state variable will bex∗1 .Let τ denote the new time subscript which counts time from t = 1 to∞. By construction then, τ ≡ t − 1.When we set the new optimization exercise (relevant fort = 1, 2....,∞) in terms of τ it looks exactly similar. In particular, thenew value function will be given by:

V (x∗1 ) ≡ Max .{xτ+1}∞

τ=0

∞

∑τ=0

βτU (xτ, xτ+1) ; xτ+1 ∈ G (xτ) for all τ = 0;

=∞

∑τ=0

βτU (x∗τ , x∗τ+1) .



Noting the relationship between t and τ, we can immediately see thatthe two value functions are related in the following way:

V (x0) =∞

∑t=0

βtU (x∗t , x∗t+1)

= U (x0, x∗1 ) + β∞

∑t=1

βt−1U (x∗t , x∗t+1)

= U (x0, x∗1 ) + β∞

∑τ=0

βτU (x∗τ , x∗τ+1)

= U (x0, x∗1 ) + βV (x∗1 ).

The above relationship is the basic functional equation in dynamicprogramming which relates two successive value functions recursively.It is called the Bellman Equation. It breaks down the inifinitehorizon dynamic optimization problem into a two-stage problem:

what is optimal today (x∗1 );what is the optimal continuation path (V (x∗1 )).



Since the above functional relationship holds for any two successivevalues of the state variable,we can write the Bellman Equation moregenerally as:

V (x) = Maxx∈G (x )

[U(x , x) + βV (x)] for all x ∈ X . (1)

The maximizer of the right hand side of equation (2) is called apolicy function:

x = π(x),

which solves the RHS of the Bellman Equation above.If we knew the value function V (.) and were it differentiable, wecould have easily found the policy function by solving the followingFONC (called the Euler Equation):

x :∂U(x , x)

∂x+ βV ′(x) = 0. (2)



Unfortunately, the value function is not known.In fact we do not even know whether it exists; if yes then whether itis unique, whether it is continuous, whether it is differentiable etc.A lot of theorems in Dynamic Programming go into establishingconditions under which a value exists; is unique and has all the niceproperties (continuity, differentibility and others).For now, without going into futher details, we shall simply assumethat all these conditions are satisfied for our problem.In other words, we shall assume that for our problem the valuefunction exists and is well-behaved (even though we do not knowits precise form).Once the existence of the value function is established, we can thensolve the FONC (3) (the Euler Equation) to get the policy function.But there is still one hurdle: what is the value V ′(x)?Here the Envelope Theorem comes to our rescue.



Recall that V (x) is nothing but the value function for the next periodwhere x is next period’s initial value of the state variable (which isgiven - from next period’s perspective).Since the Bellman equation is defined for all x ∈ X , we therefore geta similar relationship between x and its subsequent state value (x):

V (x) = Maxx∈G (x )

[U(x , x) + βV (x)] .

Then applying Envelope Theorem:

V ′(x) =∂U(x , x)

∂x. (3)

Combining the Euler Equation (3) and the Envelope Condition (4),we get the following equation:

∂U(x , x)∂x

+ β∂U(x , x)

∂x= 0 for all x ∈ X .



Replacing x , x , x by their suitable time subscripts:

∂U(xt , xt+1)∂xt+1

+ β∂U(xt+1, xt+2)

∂xt+1= 0; xt given. (4)

Equation (5) is a difference equation which we should be able to solveto derive the time path of the state variable xt (and consequently thatof the control variable yt).

Since it is a difference equation of order 2, apart from the initialcondition, we need another boundary condition.

Typically such a boundary condition is provided by the followingTransversality condition (TVC):

limt→∞

βt∂U(xt , xt+1)

∂xtxt = 0. (5)


Transversality Condition and its Interpretation:

The TVC is to be read as a complementary slackness condition in thefollowing way:

as t → ∞, if βt∂U (xt ,xt+1)

∂xt> 0, then xt = 0;

on the other hand, as t → ∞, if xt > 0, then βt∂U (xt ,xt+1)

∂xt= 0

In interpreting the TVC, notice that βt ∂U (xt ,xt+1)∂xt

captures themarginal increment in the pay-off function associated with an increasein the current stock, or its shadow price.

The TVC states that if the shadow price is positive then at theterminal date, agents will not leave any stock unused (or would leaveany postive stock at the end of the period); on the other hand, if anystock indeed remains unused at the terminal date, then it must be thecase that its shadow valuation is zero.


Stationary Dynamic Programming: Existence &Uniqueness of Value Function

We now provide some suffi cient conditions for the Value function ofthe above stationary dynamic programming problem to exist, to betwice continuously differentiable, to be concave etc.We just state the theorems here without proof. Proofs can be foundin Acemoglu (2009).

1 Let G (x) be non-empty-valued, compact and continuous in all x ∈ Xwhere X is a compact subset of <. Also let U : XG → < iscontinuous, where XG = {(xt , xt+1) ∈ X × X : xt+1 ∈ G (xt )} . Thenthere exits a unique and continuous function V : X → < that solvesthe stationary dynamic programming problem specified earlier.

2 Let us further assume that U : XG → < is conacave and iscontinuously differentiable on the interior of its domain XG . Then theunique value function defined above is strictly concave and isdifferentiable.


Non-Stationary Dynamic Programming: Existence &Uniqueness of Value Function

Even when the dynamic programming problem is non-stationary, wecan find analogous suffi cient conditions that will ensure the existence,uniqueness, concavity and differentiability of the corresponding valuefunction.Then we can proceed exactly as above to write down the Bellmanequation that relates the value functions of two successive timeperiods and then solve for the optimal policy function from thecorresponding Euler Equation and the Envelope condition.All the economic problems that we would be looking at in this coursewill satisfy these suffi ciency properties.So we shall stop bothering about this suffi ceny condition from now onand focus on applying the dynamic programming technique to theeconomic problems at hand.Interested students can look up Acemoglu (2009): Introduction toModern Economic Growth, Chapter 6, for the theorems and proofs.


Back to Household’s Choice Problem: Infinite Horizon

Recall that we had specified the representative household’soptimization problem under infinite horizon as:

Max .{cht }∞

t=0,{aht+1}∞

t=0

∞

∑t=0

βtu(cht); u′ > 0; u′′ < 0

subject to

(i) cht 5 wt + rtaht for all t = 0;(ii) aht+1 = wt + rtaht + (1− δ)aht − cht ; aht = 0 for all t = 0; ah0 given.

However in specifying the problem, we assumed that there is nointra-household borrowing.This assumption of no borrowing is too strong, and we do not reallyneed it for the results that follow.So let us relax that assumption to allow households to borrow fromone another if they so wish.



Allowing for intra-household borrowings means that constraint (i)would no longer hold. A household can now consume beyond itscurrent income at any point of time - by borrowing from others.

Allowing for intra-household borrowings also means that a householdnow has two forms of assets that it can invest its savings into:

1 physical capital (kht );2 financial capital, i.e., lending to other households (lht ≡ −bht ).

Let the gross interest rate on financial assets be denoted by (1+ rt ) .

Let physical capital depreciate over time at a constant rate δ. Thenthe gross interest rate on investment in physical capital is given by(rt + 1− δ) .



Arbitrage in the asset market ensures that in equlibrium two interestrates are the same :

1+ rt = 1+ rt − δ⇒ rt = rt − δ.

Thus we can define the total asset stock held by the household inperiod t as aht ≡ kht + lht .Notice that lht < 0 would imply that the household is a net borrower.

Hence the aggregate budget constraint of the household is now givenby:

cht + sht = wt + rta

ht , where s

ht ≡ aht+1 − aht .

Re-writing to eliminate sht :

aht+1 = wt + (1+ rt )aht − cht .


Household’s Choice Problem: Ponzi Game

But allowing for intra-household borrowing brings in the possibility ofhouseholds’playing a Ponzi game, as explained below.Consider the following plan by a household:

Suppose in period 0, the household borrows a huge amount b - whichwould allow him to maintain a very high level of consumption at allsubsequent points of time. Thus

b0 = b.

In the next period (period 1)he pays back his period 0 debt withinterest by borrowing again (presumably from a different lender). Thushis period 1 borrowing would be:

b1 = (1+ r0)b0.

In period 2 he again pays back his period 1 debt with interest byborrowing afresh:

b2 = (1+ r1)b1 = (1+ r1)(1+ r0)b0.

and so on.Das (Lecture Notes, DSE) DGE Approach February 2-22, 2016 33 / 105

Household’s Choice Problem: Ponzi Game (Contd.)

Notice that proceeding this way, the household effectively never paysback its initial loan b; he is simply rolling it over period after period.In the process he is able to perpetually maintain an arbitrarily highlevel of consumption (over and above his current income).His debt however grows at the rate rt :

bt+1 = (1+ rt )bt

which implies that limt→∞

aht ' − limt→∞bht → −∞.

This kind scheme is called a Ponzi finance scheme.If a household is allowed to play such a Ponzi game, then thehousehold’s budget constraint becomes meaningless. There iseffectively no budget constraint for the household any more; it canmaintain any arbitrarily high consumption path by playing a Ponzigame.To rule this out, we impose an additional constraint on thehousehold’s optimization problem - called the No-Ponzi GameCondition.


Household’s Choice Problem: No-Ponzi Game Condition

One Version of No-Ponzi Game (NPG) Condition:

limt→∞

aht(1+ r0)(1+ r1)......(1+ rt )

= 0.

This No-Ponzi Game condition states that as t → ∞, the presentdiscounted value of an household’s asset must be non-negative.Notice that the above condition rules out Ponzi finance scheme forsure.

If you play Ponzi game then limt→∞

aht ' − limt→∞

bht , when the latter term

is growing at the rate (1+ rt ).For simplicity, let us assume interest rate is constant at some r . Thenbht = (1+ r)

t b.Plugging this in the LHS of the NPG condition above:

limt→∞

aht(1+ r)t

' limt→∞

(−bht )(1+ r)t

= limt→∞

−(1+ r)t b(1+ r)t

= −b < 0.

This surely violates the NPG condition specified above.


Household’s Choice Problem: No-Ponzi Game Condition(Contd.)

At the same time the NPG condition specified above is lenient enoughto allow for some borrowing.In fact the condition even permits perpetual borrowing as long asborrowing grows at a rate less than the corresponding interest rate.To see this, suppose the household’s borrowing is growing at somerate g < r such that

bht = (1+ g)t b.

Plugging this in the LHS of the NPG condition above:

limt→∞

aht(1+ r)t

' limt→∞

(−bht )(1+ r)t

= limt→∞

−(1+ g)t b(1+ r)t

= −b limt→∞

(1+ g1+ r

)t.

Notice that g < r implies that the term(1+ g1+ r

)is a positive

fraction and as t → ∞,(1+g1+r

)t→ 0.


Household’s Choice Problem: No-Ponzi Game Condition(Contd.)

Since b is finite, this implies that in this case

limt→∞

aht(1+ r)t

→ 0.

In other words, the NPG condition is now satisfied at the margin!

Economically, this kind of borrowing behaviour implies that the debtof the agent is not exploding and the agent must have startedrepaying at least some part of it (though not all) from his own pocket!


Household’s Choice Problem - Revisited:

After imposing the No-Ponzi Game condition, the household’optimization problem now becomes:

Max .{cht }∞

t=0,{aht+1}∞

t=0

∞

∑t=0

βtu(cht)

subject to

(i) aht+1 = wt + (1+ rt )aht − cht ; aht ∈ < for all t = 0; ah0 given.(ii) The NPG condition.

Here cht is the control variable and aht is the state variable.


Household’s Choice Problem - Revisited: (Contd.)

We can now apply the dynamic programming technique to solve thehousehold’s choice problem.

First let us use constraint (i) to eliminate the control variable andwrite the above dynamic programming problem in terms of the statevariable alone:

Max .{aht+1}∞

t=0

∞

∑t=0

βtu({wt + (1+ rt )aht − aht+1

})Corresponding Bellman equation relating V (ah0) and V (a

h1) is given

by:

V (ah0) = Max{ah1}

[u({w0 + (1+ r0)ah0 − ah1

})+ βV (ah1)

].


Household’s Problem: Bellman Equation

More generally, we can write the Bellman equation for any two timeperiods t and t + 1 as:

V (aht ) = Max{aht+1}

[u({wt + (1+ rt )aht − aht+1

})+ βV (aht+1)

].

Maximising the RHS above with respect to aht+1, from the FONC:

u′({wt + (1+ rt )aht − aht+1

})= βV ′(aht+1) (6)

Notice that V (aht+1) and V (aht+2) would be related through a similar

Bellman equation:

V (aht+1) = Max{aht+2}

[u({wt+1 + (1+ rt+1)aht+1 − aht+2

})+ βV (aht+2)

].

Applying Envelope Theorem on the latter:

V ′(aht+1) = u′({wt+1 + rt+1aht+1 − aht+2

}).(1+ rt+1). (7)


Household’s Problem: Optimal Solutions

Combining (5) and (6):

u′({wt + rtaht − aht+1

})= βu′

({wt+1 + rt+1aht+1 − aht+2

})(1+ rt+1).

The above equation implicitely defines a 2nd order difference equationis aht .

However we can easily convert it into a 2× 2 system of first orderdifference equations in the following way.


Household’s Problem: Optimal Solutions (Contd.)

Noting that the terms inside the u′(.) functions are nothing but chtand cht+1 respectively, we can write the above equation as:

u′(cht)= βu′

(cht+1

)(1+ rt+1). (8)

We also have the constraint function:

aht+1 = wt + (1+ rt )aht − cht ; ah0 given. (9)

Equations (7) and (8) represents a 2× 2 system of differenceequations which implicitly defines the ‘optimal’trajectories

{cht}∞t=0

and{aht+1

}∞t=0.

The two boundary conditons are given by the initial condition ah0 , andthe NPG condition.


Optimal Solution Path to Household’s Problem: AnExample

Let us look at an explicitly characterisation of the household’s optimalpaths for a specific example.

Supposeu(c) = log c

Let us also assume that wt = w and rt = r for all t .

Then we can immediately get two difference equations characterizingthe optimal trajectories for the household as:

cht+1 = β(1+ r)cht (10)

andaht+1 = w + (1+ r)a

ht − cht ; ah0 given. (11)

The two equations along with the two boundary conditons can besolved explicitly to derive the time paths of cht and a

ht .


Household’s Problem: Optimal Solutions (Contd.)

Equation (9) is a linear autonomous difference equation, which can bedirectly solved (by iterating backwards) to get the optimalconsumption path as:

cht = βt (1+ r)tch0 . (12)

However,we still cannot completely characterise the optimal pathbecause we still do not know the optimal value of ch0 . (Recall that c

h0

is not given; it is to be chosen optimally).Here the NPG condition comes in handy in identifying the optimal ch0 .

Note that the NPG condition in this case is given by:

limt→∞

aht(1+ r)t

= 0.


Household’s Problem: Role of the NPG Condition

Now let us take the budget constraint of the household at any futuredate T > 0:

ahT+1 = w + (1+ r)ahT − chT .

Iterating backwards,

ahT+1 = w + (1+ r)ahT − chT= w + (1+ r)

[w + (1+ r)ahT−1 − chT−1

]− chT

= ....

=T

∑t=0

(w(1+ r)T−t

)−

T

∑t=0

(cht (1+ r)

T−t)+ (1+ r)T+1ah0 .

Rearranging terms:

ahT+1(1+ r)T

=T

∑t=0

(w

(1+ r)t

)+ (1+ r)ah0 −

T

∑t=0

(cht

(1+ r)t

)Das (Lecture Notes, DSE) DGE Approach February 2-22, 2016 45 / 105

Household’s Problem: Role of the NPG Condition (Contd.)

Now let T → ∞. Then applying the NPG condition to the LHS, weget:

∞

∑t=0

(w

(1+ r)t

)+ (1+ r)ah0 −

∞

∑t=0

(cht

(1+ r)t

)= 0

i.e.,∞

∑t=0

(cht

(1+ r)t

)5

∞

∑t=0

(w

(1+ r)t

)+ (1+ r)ah0 . (13)

Equation (12) represents the lifetime budget constraint of thehousehold. It states that when the NPG condition is satisfied, thenthe discounted life-time consumption stream of the household cannotexceed the sum-total of its discounted life-time wage earnings and thereturns on its initial wealth holding.

It is easy to see that even though we have specified the NPGcondition in the form of an inequality, the households would alwayssatisfy it at the margin such that it holds with strict equality.


Household’s Problem: Role of the NPG Condition (Contd.)

Given that equation (12) holds with strict equality, we can nowidentify the optimal value of ch0 .We had already derived the optimal time path of cht as:

cht = βt (1+ r)tch0 .

Using this in equation (12) above, we get:

∞

∑t=0

(βt (1+ r)tch0(1+ r)t

)=

∞

∑t=0

(w

(1+ r)t

)+ (1+ r)ah0

⇒∞

∑t=0

(βt)ch0 =

[∞

∑t=0

(w

(1+ r)t

)+ (1+ r)ah0

]

⇒ ch0 = (1− β)

[∞

∑t=0

(w

(1+ r)t

)+ (1+ r)ah0

].


Household’s Problem: NPG vis-a-vis TVC

So for this particular example, we have been able to explicitly solvefor the optimal consumption path of the households.

But there is a problem that we still need to sort out.

Recall that while discussing the dynamic programming problem wehad specified a transversality condition (TVC) as one of our boundarycondition (Refer to equation (5) specified earlier).

Then in defining the household’s problem with intra-householdborrowing, we have introduced the NPG condition as anotherboundary condition.

So we now have a problem of plenty: for a 2× 2 dynamic system, itseems that we have three boundary conditions!!!

Between the TVC and the NPG condition, which one should we useto characterise the solution?

As it turns out, along the optimal path the NPG condition and theTVC become equivalent.


Household’s Problem: NPG vis-a-vis TVC (Contd.)

To see this, let us take a closer look at the TVC as had been specifiedearlier in equation (5)In the context of the current problem, this transversality conditionwould be given as follows (verify this):

limt→∞

βtu′(cht )(1+ rt )aht = 0

For our specific example with log utility and constant factor prices,this condition reduces to

limt→∞

βt1cht(1+ r)aht = 0

Now given the solution path of cht , we can further simplify the abovecondition to:

limt→∞

βt1

βt (1+ r)tch0(1+ r)aht = 0

⇒ limt→∞

aht(1+ r)t

= 0 .


Household’s Problem: NPG vis-a-vis TVC (Contd.)

But this is nothing but our earlier NPG condition - now holding withstrict equality!

Thus when the household is on its optimal path, the NPG conditionand the Transversality condition become equivalent - except that theNPG condition now must hold with equality.

So in identifying the optimal trajectories, we could use either of themas the relevant boundary condition.


Household’s Problem: Heterogenous Agents

So far we have assumed that all households are identical so that wecould carry out the analysis in terms of a representative agent.

But if all households are identical in every respect, then allowing forintra-household borrowing and the consequent NPG condition doesnot make sense. In any case one side of the borrowing/lending marketwill always be missing and hence no borrowing or lending will evertake place.

All these conditions make sense only if households are heterogenous.

Let us now extend the framework to allow for heterogenoushouseholds.

From now on, we shall assume that households differ in terms of theirintial wealth.

However, we shall continue to assume that preference-wise they areall identical.


Household’s Problem: Heterogenous Agents (Contd.)

Let us now go back to our earlier example of log utility and constantfactor returns.We have already seen that for any household with an initial wealthlevel of ah0 will have the following optimal consumption path:

cht = βt (1+ r)tch0 .

where

ch0 = (1− β)

[∞

∑t=0

(w

(1+ r)t

)+ (1+ r)ah0

].

Notice that the rate of growth of consumption along the optimal pathis given by β(1+ r)− 1, which is independent of the initial wealthlevel (or even the accumulated wage income!).Thus along the optimal path, consumption of all households grow atthe same rate - irrespective of their initial wealth.The initial wealth only determines the level of optimal consumption:higher initial wealth means higher level of consumption.



This is a striking result because it tells us that the initial wealth hasno growth effect, only level effect.

It also tells us that when all households are following their respectiveoptimal trajectories, the initial inequality in consumption will bemaintained perpetually.

What about the asset stock?

Note that we can solve for the time path of aht (given ah0) by solving

the following dynamic equation:

aht+1 = w + (1+ r)aht − [β(1+ r)]

t ch0

This is a difference equation which linear but non-autonomous;solving this would require more elaborate technique than merebackward induction.

We shall come back to this equation once we discuss the methods ofsolving such non-autonomous difference equations.



All these results were of course derived under the assumption ofconstant factor returns. When factor returns (wt and rt) are changingover time, the consumption growth rate itself will change.We can easily generalize the results to such non-autonomous casesbut to see what is happenning to the growth rate of consumption insuch cases, we shall have to derive the precise time path of rt , whichmeans we shall have to discuss the production side story.Before we move on to the production side story, notice that inderiving all these results, we have also made use of the log utility,which we know is special.Can we generalize these results to other utility functions as well?It turns out, all the results will go through for a broad class of utilityfunctions called the CRRA variety:

u(c) =c1−σ

1− σ; σ 6= 1.



This utility function has several interesting characteristics:

1 It is associated with constant elasticity of marginal utility:−cu′′(c)u′(c)

= σ

2 It is associated with constant relative risk aversion (as defined by the

Arrow-Pratt measure of relative risk aversion):−cu′′(c)u′(c)

= σ

3 It is associated with constant elasticity of substitution between

current and future consumption:−d

(ct+1ct

)/(ct+1ct

)d(u ′(ct+1)u ′(ct )

)/(u ′(ct+1)u ′(ct )

) = 1σ

Exercise: Assume that wages and interest rates are constant and usethe dynamic programming technique to derive the dynamic equationfor the optimal consumption path of an agent with an initial assetstock of ah0 , when his utility function is of the CRRA variety, asdefined above.


Production Side Story: A Micro-founded Dynamic Theoryof Firms’Choices?

We typically do not think of the firm’s choice problem as a dynamicone. In a perfectly competitive set up, a firm is a blackbox: it doesnot own any factors of production and merely decides how muchlabour to employ and how much capital to hire in every period so asto maximise its current profit (taking all prices as given).

As long as such hiring decisions do not affect future profits, settingthe optimization problem in a dynamic framework (i.e., optimizingover multiple time periods) does not bring in any extra insight overthe static optimization problem.

The firms will have meaningful dynamic choices if and only if thefirms own the capital stock that they employ and part of the currentprofit can be invested in augmenting their capital stock which affectstheir future profitability.

We now turn to the optimal choices of firm in such a dynamic setting.


A Dynamic Theory of Firm Behaviour:

Once again we ignore the price dynamics and assume that the pricelevel remains constant at unity.Thus all variables that we consider below are expressed in terms oftheir real values.Consider a firm that which produced a final commodity in everyperiod using a production function

Yt = F (Nt ,Kt ).

The usual diminishing marginal product proprties and CRS propertyare assumed to hold.At time t, the capital stock available to the firm is given. However itcan augment its capital stock over time by investing an amount It(out of its current profit) which augments the capital stock in thenext period:

Kt+1 = It + (1− δ)Kt for all t = 0, 1, 2, ....∞. (14)


A Dynamic Theory of Firm Behaviour: (Contd.)

Assuming a competitive market structure (such that the firm is aprice taker), its instantaneous net pay-off/profit at any time period tis given by:

πt = F (Nt ,Kt )− wtNt − It .Let r be the given time-invariant net rate of interest in the economy.

The firm is believed to be infinitely-lived. So the present discountedvalue of its sum of net profit from time 0 to ∞ is given by:

V =∞

∑t=0

πt

(1+ r)t=

∞

∑t=0

F (Nt ,Kt )− wtNt − It(1+ r)t

. (15)

The firm maximizes (15) subject to its period by period capacityaugmentation equation (given by (14)), to determine its optimal levelof employment in every period (Nt ) and its optimal level ofinvestment in every period (It ).



However, since choosing current level of investment is equivalent tochoosing next period’s capital stock, we can use the capacityaugmentation equation (given by (14)) to eliminate It and write downthe optimization problem of the firm as an unconstraint problem:

Max{Nt}∞

0 ,{Kt+1}∞0

V =∞

∑t=0

F (Nt ,Kt )− wtNt −Kt+1 + (1− δ)Kt(1+ r)t

; K0 given.

Corresponding Bellman equation relating V (K0) and V (K1):

V (K0) = Max{L0,K1}

[F (K0,N0)− w0N0 − {K1 − (1− δ)K0}+

V (K1)(1+ r)

].

More generally:

V (Kt ) = Max{Lt ,Kt+1}

[F (Kt ,Nt )− wtNt − {Kt+1 − (1− δ)Kt}+

V (Kt+1)(1+ r)

].



This yields the following set of first order conditions:

(i)∂V (Kt )

∂Nt= 0⇒ FN (Nt ,Kt )− wt = 0

(ii)∂V (Kt )∂Kt+1

= 0⇒ ∂V (Kt+1)∂Kt+1

= (1+ r)

If the value function exists and is differentiable, we can once againapply the envelope theorem to the value function defined for periodt + 1, (i.e., V (Kt+1)) to get:

∂V (Kt+1)∂Kt+1

= FK (Kt+1,Nt+1) + (1− δ) (16)



This tells us that the dynamically optimizing firm would choose anemployment level and an investment level in every period such that

FN (Nt ,Kt ) = wt ;

FK (Nt+1,Kt+1)− δ = r ⇒ FK (Nt+1,Kt+1) = r .

These decision rules look exactly analogous to the decisions that willundertaken by a firm in a static optimization framework where it onlymaximises its current profit period after period.

Thus adding a dynamic framework does not add anything extrato the producer’s side of the story; their optimal decisionmaking rules remain identical to the static story.



This equivalence arises because the marginal benefits and the marginalcosts assocaited with both the exercises are precisely the same:

In the dynamic set up, an act of investment generates more profittomorrow by generating more output; hence the associated netmarginal benefit is captured by the correponding MPK − δ. Therelative cost on the other hand is measured by r , since the extra profitappearing tomorrow will be discounted at the rate r .In the static set up, an additional unit of capital currently employedgenerates more profit today by generating more output; hence theassociated net marginal benefit is once again captured by thecorreponding MPK . The relative cost on the other hand is againmeasured by r , since this is the rental price that the firm has to pay tothe capital-owners (households).

Since both the benefits and the costs appear in the same time period(tomorrow - for the dynamic problem; today - for the static problem)and since the benefits and costs associated with the two frameworksare also identical, it is not suprising that they produce identicaloptimal decision rules.



To make the dynamic story of firm behaviour look different from thecorresponding static story, we must bring in some additionalintertemporal linkages, e.g., a cost that is incurred today but thebenefit is reaped only tomorrow.

Introducing an adjustment cost of investment serves this purpose.

So let us now augment the dynamic model of the firm to take intoaccount some adjustment costs of investment.


A Dynamic Theory of Firm Behaviour: Adjustment Costs

As before, consider a representative firm which takes all its inputprices as given. It has access to a production technology that useslabour and capital as inputs:

Yt = F (Kt ,Nt )

The firm hires labour from the labour market at the market wage ratewt . But it owns the capital stock that it employs.

The stock of capital owned by the firm can be augmented over timeby investing a part of the profit.

However the investment process is now subject to adjustment costs.

Adding new machines is disruptive to the production process andleads to loss of revenue.

These adjustment costs are convex: they are low when the level ofinvestment is low; but rise steeply as the level of investment rises.


A Dynamic Theory of Firm Behaviour with AdjustmentCosts: (Contd.)

For convenience we shall assume a quadratic cost function:

C (It ) = bI 2t

At any point of time the net pay-off/profit of the firm is:

πt = F (Kt ,Nt )− wtNt − bI 2t − ItThe dynamic optimization problem of the representative firm can thenbe written as:

Max .{It}∞

t=0,{Lt}∞t=0,{Kt+1}

∞t=0

∞

∑t=0

π (Kt ,Nt , It )

(1+ r)t

subject to

(i) Kt+1 −Kt = It − δKt ; Kt ∈ < for all t = 0; K0 given.Here It and Nt are control variables and Kt is the state variable.



We can use constraint (i) to eliminate the control variable It andwrite the above dynamic programming problem in terms of the Ktand Nt and Kt+1 alone:

Max .{Lt}∞

t=0,{Kt+1}∞t=0

∞

∑t=0

1

(1+ r)t[F (Kt ,Nt )− wtNt

− b{Kt+1 − (1− δ)Kt}2 − {Kt+1 − (1− δ)Kt}]

Corresponding Bellman equation relating V (K0) and V (K1):

V (K0) = Max{N0,K1}

[F (K0,N0)− w0N0 − b{K1 − (1− δ)K0}2

− {K1 − (1− δ)K0}] +V (K1)(1+ r)

].



More generally:

V (Kt ) = Max{Nt ,Kt+1}

[F (Kt ,Nt )− wtNt − b{Kt+1 − (1− δ)Kt}2

− {Kt+1 − (1− δ)Kt}] +V (Kt+1)(1+ r)

].

From the FONCs:

∂V (Kt )∂Nt

= 0; (17)

∂V (Kt )∂Kt+1

= 0. (18)



Simplifying:FN (Kt ,Nt ) = wt ; (19)

and

1(1+ r)

∂V (Kt+1)∂Kt+1

= 2b{Kt+1 − (1− δ)Kt}+ 1

i.e., It =12b

[1

(1+ r)∂V (Kt+1)

∂Kt+1− 1]

(20)



The term ∂V (Kt+1)∂Kt+1

in the above equation, which is the derivative ofthe value function, measures the marginal valuation of an unitaddition of capital stock in terms of the entire maximised stream ofprofits.

In other words, this terms measures the shadow price of investment attime t. We shall denote this by qt .

Thus equation (20) can be written as:

It =12b

[qt

(1+ r)− 1]

(21)

Interpretation?



Assuming that the value function exists and is differentiable, we canonce again apply the envelope theorem to the value function definedfor period t + 1, (i.e., V (Kt+1)) to get:

qt ≡∂V (Kt+1)

∂Kt+1= FK (Kt+1,Nt+1) + (1− δ) [2b{Kt+2 − (1− δ)Kt+1}+ 1]

Plugging this value in equation (21), we once again get an implicit2nd order difference equation in Kt , which will determine the optimalinvestment plan and therefore the optimal capital trajectory of thefirm, subject to the given initial value of the capital stock (K0) andthe associated Transversality condition.



Notice that from equation (21):

It T 0 according as qt T (1+ r)

i.e., FK (Kt+1,Nt+1)− δ+ (1− δ) [2b{Kt+2 − (1− δ)Kt+1}] T r .Thus it is no longer optimal for firms to invest in capital stock as longthe net return (FK (Kt+1,Nt+1)− δ) is greater than the cost (r).

Indeed to induce firms to investment, the marginal benefit frominvestment now has to be suffi ciently high to cover the associatedadjustment cost. This explain the presence of an additional term inthe LHS of the above equation.



As before, it might be easier to deal with the dynamics if we convertthe 2nd order difference equation in a 2× 2 system of first orderdifference equations. This can be done in the following way.Replace ∂V (Kt+1)

∂Kt+1in equation (20) by the above expression. Further,

note that {Kt+2 − (1− δ)Kt+1} ≡ It+1, whileNt+1 : FN (Kt+1,Nt+1) = wt+1 ⇒ Nt+1 = f (Kt+1,wt+1). Using allthese in the equation (20), we get the following system of differenceequation in the control variable (It) and the state variable (Kt):

(1+ r) [2bIt + 1] = FK (Kt+1, f (Kt+1,wt+1)) + (1− δ) [2bIt+1 + 1]

Kt+1 = It + (1− δ)KtThese two equations along with the initial condition and the TVC willcompletely characterise the optimal investment path for the firm.Exercise: Use equation (15) to write down the TVC for thisparticular investment problem of the firm.



Notice that the above difference equation is still implicit. We cannotderive the explicit solution and the precise optimal investment pathunless we assume some specific production function.

Later we shall look at some specific examples and characterise theprecise optimal paths in the context of those examples.


Bringing the Households and Firms Together: The GeneralEquilibrium Set Up

So far we have looked at the households’problem and the firms’problem in isolation.

Both sets of agents were assumed to be ‘atomistic’, who take all themarket variables as exogenously given.

But in the aggregate economy, the market variables are notexogenous; they are determined precisely by the aggregate actions ofthe households and the firms.

So we now consider the general equilibrium set up where thehouseholds’and the firms’actions - mediated through the market -generates some aggregative behaviour for the entire macroeconomy.

The corresponding solution for the aggregate economy will be calledthe ‘decentralized’or ‘market’equilibrium solution (as opposed to acontrasting case where production is centralized under a socialplanner).


General Equilibrium in the Decentralized Economy:

Recall that we have solved the production side story under twoalternative of assumptions - one where investment does not entail anyadjustment costs and another one where investment is associatedwith some adjusment costs.

We have seen that in the first case, setting the firms’problem in adynamic framework (where the firms own the capital stock and carrryout the act of investment) yields results which are equivalent to theresults that we would obtain in a static framework where firms do notown the capital stock and simply rent them in from the households.

Since these two exercises are equivalent, when we consider theproblem without adjustment costs, we shall simply revert back to theassumption that all capital stocks are owned by the households; firmsmerely rent them in period by period to maximise their staticperiod-by-period profit.


General Equilibrium in the Decentralized Economy:(Contd.)

In the presence of adjustment costs, the dynamic set up generatesresults different from the static one.So in discussing the general equilibrium outcome for this case, weshall retain the assumption that capital stocks are owned by the firmsand investment are carried out by firms.However, this begs the following question: what happens to theaccumulated profits of the firms? How do we bring it back to thecircular flow of income such that it eventually goes back to thehouseholds - to be consumed or saved?Here we shall assume that even though the firms are carrying out theact of investment to maximise the profits, ultimately the householdsare the owners of the firms because they own shares of the firms.Thus the accumulated profits of the firms flow back to the householdsin the form of dividend income.Hence we’ll have to suitably modify the households’problem.


General Equilibrium in the Decentralized Economy:(Contd.)

We now look at both these general equilibrium cases one by one.


General Equilibrium: No Adjustment Costs

Let us quickly revisit the household and firm specifications for thiscase:

We have H single-membered households which are identical in terms ofpreferences but differ in terms of their initial asset holdings;Each household is endowed with one unit of labour - which it suppliesinelastically to the market in every period;Households are atomistic and take the market wage rate (wt ) andmarket the interest rate (rt ) (and the corresponding net interest rate,rt = rt − δ) as given. But they are endowed with perfect foresight - sothey can correctly guess the entire stream of current & future wagerates {wt}t=∞

t=0 , as well as the current & future interest rates {rt}t=∞t=0 .

The households own the entire labour and the capital stock in theeconomy. In addition, they also hold loans against one another.Each household maximises its lifetime utility subject to its period byperiod budget constraint.


General Equilibrium: No Adjustment Costs (Contd.)

On the production side:

There are M identical firms endowed with a technology to produce thefinal commodity.The technology uses capital and labour as inputs; it exhibitsdiminishing returns with respect to each of the inputs; it is also CRS inboth the inputs.The firms do not own any capital all labour; they hire labour andcapital from the market to carry out production in each period.The firms operate under perfect competetion; they take the marketwage rate (wt ) and market the interest rate (rt ) as given.The firms optimally decide about how much labour/capital to employin every period so as to maximise its period-by-period profit.



For expositional simplicity, we shall assume specific functional formsfor the utility function and the production function. Accordingly, let

u(c) = log c

andYt = F (Kt ,Nt ) = (Kt )α(Nt )1−α; 0 < α < 1.



The optimization problem of a household h with an initial assetholding of ah0 is given by:

Max .{cht }∞

t=0,{aht+1}∞

t=0

∞

∑t=0

βt log(cht)

subject to

aht+1 = wt + (1+ rt )aht − cht ; aht = 0 for all t = 0; ah0 given.

Characterization of the optimal paths:

cht+1 = β(1+ rt+1)cht ; (22)

aht+1 = wt + (1+ rt )aht − cht ; (23)

ah0 given; limt→∞

aht(1+ r0)(1+ r1)......(1+ rt )

= 0 (NPG/TVC).



While one can solve for the optimal paths for each household, we aremore interested in tracking the aggregate economy.

For this purpose, define per capita consumption and per capita assetholding in this economy as:

ct ≡

H

∑h=1

cht

H; at ≡

H

∑h=1

aht

H.

Recall that households hold their assets in the form of either physicalcapital or financial capital (loans) such that

at ≡

H

∑h=1

aht

H=

H

∑h=1

(kht + lht )

H=

H

∑h=1

kht

H+

H

∑h=1

lht

H.



Since one household’s lending is another household’s borrowing, on

the aggregate

H

∑h=1

lht

H= 0. Thus,

at ≡

H

∑h=1

aht

H=

H

∑h=1

kht

H≡ kt ,

where kt denotes the per capita capital stock in the economy.Notice that the individual optimal transition equations (22 & 23) canbe used to derive the transition equations for the per capitaconsumption and per capita capital stock of the economy in thefollowing way:

ct+1 ≡

H

∑h=1

cht+1

H=

H

∑h=1

β(1+ rt+1)cht

H= β(1+ rt+1)

H

∑h=1

cht+1

H= β(1+ rt+1)ct



On the other hand,

kt+1 = at+1 ≡

H

∑h=1

aht+1

H=

H

∑h=1

[wt + (1+ rt )aht − cht

]H

=

H

∑h=1

wt

H+ (1+ rt )

H

∑h=1

aht

H−

H

∑h=1

cht

H= wt + (1+ rt )kt − ct .

Finally, the individual boundary conditions can also be aggregatedover all H households to get the boundary conditions for kt as:

k0 given; limt→∞

kt(1+ r0)(1+ r1)......(1+ rt )

= 0



We have now derived the transition equations of the per capitaconsumption and per capita capital stock for the aggregativeeconomy - except that we still do not know the precise values of themarket wage rate (wt ) and the net interest rate ( rt = rt − δ).These factor prices are determined in the market by the demand andsupply of labour and capital respectively.At any time period t, total supply of capital (coming from all thehouseholds) is given by:

KSt =H

∑h=1

kht = H.kt

Likewise, total supply of labour (coming from all the households) isgiven by:

NSt = H

The demand for these factors on the other hand comes from the firms.Das (Lecture Notes, DSE) DGE Approach February 2-22, 2016 85 / 105


At any point of time t, the profit maximization problem of a firm i isgiven by:

Max .N it ,K

it

[(K it )

α(N it )1−α − wtN it − rtK it

].

Corresponding FONCs:

(1− α)(K it )α(N it )

−α = wt

⇒ (1− α)

(K itN it

)α

= wt (24)

α(K it )α−1(N it )

1−α = rt

⇒ α

(K itN it

)α−1= rt (25)



Since all firms are endowed with identical technologies and face thesame market-determined factor prices, they all employ the sameamount of capital and labour, so that the aggregate demand forlabour and capital respectively are given by:

KDt =M

∑i=1K it = M.K

it

NDt =M

∑i=1N it = M.N

it



Equilibrium in the factor market requires that:

KSt = KDt and NSt = N

Dt .

In other words, factor market clearing conditions are given by:

H.kt = M.K itH = M.N it

Writing in ratio terms, factor market clearing condition requires that:

kt =K itN it

(26)

At every point of time t, for any historically given value of per capitacapital stock (kt) owned by the households, the above equality isensured by the full flexibility of wt and rt .



Substituting K itN itby kt in equations (24) and (25), we get:

(1− α) (kt )α = wt (27)

α (kt )α−1 = rt (28)

Given kt , the wt and rt adjust in every period to maintain the abovetwo equalities.

Thus we have precisely identified the market determined values of wtand rt in every period as a function of the historically given per capitacapita stock (which is also the equilibrium capital-labour ratioemployed by each firm).

We now use these information to completely characterise the dynamicpaths of per capita consumption and per capita capital stock for theaggregative economy.



Recall the dynamic equations for ct and kt :

ct+1 = β(1+ rt+1)ct ;

kt+1 = wt + (1+ rt )kt − ct .Noting that rt = rt − δ, and replacing the market clearing values ofwt and rt derived above, we get:

ct+1 = β[1+ α (kt+1)

α−1 − δ]ct ;

kt+1 = (1− α) (kt )α +

[1+ α (kt )

α−1 − δ]kt − ct

⇒ kt+1 = (kt )α + (1− δ) kt − ct .

These two equations along with the two boundary conditionscompletely characterize the evolution of per capita consumption andper capita capital stock for this decentralized economy.We shall come back to the precise description of these time pathslater.


General Equilibrium with a Convex Adjustment Cost forInvestment:

Let us now turn to characterization of the general equilibrium wheninvestment is associated with a convex adjustment cost.

We have indicated earlier that in this case the households’problemwill have to be suitably modified.

Let us revisit the household side of the story for this case:

We have H single-membered households which are identical in terms ofpreferences but differ in terms of their initial asset holdings;Each household is endowed with one unit of labour - which it suppliesinelastically to the market in every period;The households do not own the capital stock directly; nor do theyundertake investments in physical capital.However, households hold shares of the firms which allow them to earnsome dividend income in every period. In addition, they also hold loansagainst one another.


General Equilibrium with Adjustment Costs: (Contd.)

Household Side Story:

Assume that the fims’equity prices remain constant over time -normalized to unity.Then the budhet equation of household h in any period t is given by:

cht + sht = wt + rta

ht ,

where the asset stock of the household, aht = lht + n

ht such that the

household holds its assets either in the form of intra-household loans(lht ) or in the form of equity holdings over firms (nht ).In equilibrium, the rate of return from both assets must be the same(otherwise, households will hold only one form of asset - whichevergives them higher return); the common rate of return is denoted by rthere.



Household Side Story:

The savings of the households are spent in buying new assets (newloans and/or new shares), which augments the asset stock of thehousehold over time:

sht = aht+1 − aht = 4lht +4nht .

Replacing sht by(aht+1 − aht

)in the budget equation of the household

and simplifying, we get the period by period burget constraint of thehousehold as:

aht+1 = wt + (1+ rt ) aht − cht ; ah0 given.

Each household maximises its lifetime utility subject to its period byperiod budget constraint.



On the production side:

There are M identical firms endowed with a technology to produce thefinal commodity.The technology uses capital and labour as inputs.The firms employ labour services provided by the households, but theyown the capital stock that they employ. They also carry out investmentactiviies that augment the capital stock in every period.Investment is subjet to an adjustment cost: C (It )Firms distribute part of their profits in every period as dividends (dt ) tothe existing shareholders. It finances the new investment (as well asthe adjustment cost) from the retained profits as well as by issueingnew shares.At any point of time the net pay-off of firm i is therefore given by:

πit = F (Kit ,N

it )− wtN it − dtnit + nit+1 − C (I it )− I it

The firms maximise the discounted sum of their lifetime pay-offs.



For expositional simplicity, we shall again assume specific functionalforms.

Accordingly, letu(c) = log c;

Yt = F (Kt ,Nt ) = (Kt )α(Nt )1−α; 0 < α < 1;

C (I ) = bI 2t .

We shall also assume that there is 100% depreciation of the existingcapital stock such that

δ = 1.

Notice that 100% depreciation implies,

It ≡ Kt+1.



As before, the optimization problem of a household h with an initialasset holding of ah0 is given by:

Max .{cht }∞

t=0,{aht+1}∞

t=0

∞

∑t=0

βt log(cht)

subject to

aht+1 = wt + (1+ rt )aht − cht ; aht = 0 for all t = 0; ah0 given.

Characterization of the optimal paths:

cht+1 = β(1+ rt+1)cht ; (29)

aht+1 = wt + (1+ rt )aht − cht ; (30)



Once again we are going to focus on the aggregate economy, inparticular, on the dynamics of the per capita (or average)consumption and per capita (or average) asset stock:

ct ≡

H

∑h=1

cht

H; at ≡

H

∑h=1

aht

H.

Recall that now households hold their assets in the form of eitherloans to other households or equility holdings of firms. As before theloans held by various households on the aggregate cancel each otherso that we have,

at ≡

H

∑h=1

lht

H+

H

∑h=1

nht

H=

H

∑h=1

nht

H≡ nt .



As before, aggregating over all households, the dynamics of ct and ntcan be derived as:

ct+1 = β(1+ rt )ct ;

nt+1 = wt + (1+ rt )nt − ct .These along with the corresponding boundary conditions willdetermine the time path of ct and nt .

However, what we still do not know:

1 the equilibrium values of wt and rt ;2 the evolution of the physical capital formation (kt).

For these, we will have to turn to the firms’problem.



At any point of time t, the dynamic optimization problem of a firm iis given by:

Max .{N it}∞

t=0,{K it+1}∞

t=0,{nit+1}∞

t=0

∞

∑t=0

1

(1+ r)t[F (K it ,N

it )− wtN it

− dtnit + nit+1 − b{K it+1}2 − {K it+1}]Note that now there are two state variables for the firm who valuesare historically given: the existing capital stock that the firm alreadyowns (K i0) and the existing stock of shares that the firm has alreadyissued (ni0).As before one can write down the corresponding Bellman equation interms of the value function:

V (K it , nit ) = Max

{N it ,K it+1,nit+1}[F (K it ,N

it )− wtN it − dtnit + nit+1

−b{K it+1}2 −K it+1 +V (K it+1, n

it+1)

(1+ r)].


Equilibrium with Adjustment Costs: (Contd.)

The firm now has three choice variables: N it ,Kit+1,n

it+1.

The corresponding FONCs are given by:

∂V (K it , nit )

∂N it= 0; (31)

∂V (K it , nit )

∂K it+1= 0; (32)

∂V (K it , nit )

∂nit+1= 0. (33)

Simplifying:FN (Kt ,Nt ) = wt ; (34)

1(1+ r)

∂V (K it+1, nit+1)

∂K it+1= 2b{K it+1}+ 1 (35)

1+1

(1+ r)∂V (K it+1, n

it+1)

∂nit+1= 0; (36)



Now applying Envelope Theorem to the next period’s value function:

∂V (K it+1, nit+1)

∂nit+1= −dt+1

∂V (K it+1, nit+1)

∂K it+1= FK (K

it+1,N

it+1)

Using these envelope conditions in the FONCs and noting that for thegiven production function, FK (K ,N) = αK α−1N1−α andFN (K ,N) = (1− α)K αN−α, we get the following dynamic equations for firm i :

wt = (1− α)(K it)α (

N it)−α

; (37)

dt+1 = (1+ r); (38)

I it =12b

[(α)(K it+1

)α−1 (N it+1

)1−α

(1+ r)− 1]

(39)



Aggregating over all firm, we get the dynamic equation for capitalformation in the economy as:

It =M2b

[(α) (Kt+1)

α−1 (Nt+1)1−α

(1+ r)− 1]

(40)

=M2b

[(α) (Kt+1)

α−1 (Nt+1)1−α

dt+1− 1][from (38)] (41)

Noting that all firms are identical, average capital stock in theeconomy is given by:

kt =KtM

Using the above equation and equation the demand and supply oflabour, the dynamic equation for evolution of capital stock in thiseconomy can be derived as:

it ≡ kt+1 =12b

[α (kt+1)

α−1

dt+1− 1]

(42)Das (Lecture Notes, DSE) DGE Approach February 2-22, 2016 102 / 105

General Equilibrium: What have we learnt so far?

Notice that the general equilibrium analysis for both the cases(without and with adjustment costs) specify the dynamic equationscharacterizing the evolution of the capital stock of this economy.

Given that total population/labour force is constant at H, this willalso govern the evolution of the per capita as well as aggregateoutput in this economy.

In other words, through the general equilibrium analysis, we haveactually characterized the growth path for the economy underalternative assumptions about investment costs.

This brings us directly to the realm of economic growth.

Notice however that such growth paths would be relevant only for aperfectly competetive market economy populated by rational agentswith complete information and no uncertainty.


General Equilibrium: What have we learnt so far?

Are these the only possible characterization of the growth path of aneconomy?

To put it differently, could there be alternative growth pathsassocaited with alternative specification of the macroeconomy (say,with imperfect markets or incomplete information or uncertainty)?

To answer these questions, we shall have to get into a detaileddiscussion of various theories of economic growth.

We shall stop at this point and come back to the specific issuesrelated economic growth at the third module of the course.


DGE Approach: Reference

Reference for Dynamic Programming Technique:

Daron Acemoglu (2009): Introduction to Modern EconomicGrowth; Princeton University Press, chapter 6.

Reference for DGE approach to Macroeconomics:

Michael Wickens (2008): Macroeconomic Theory: A DynamicGeneral Equilibrium Approach, Princeton University Press, chapters1& 2.

Statutory Warning: I do not follow any particular textbook word byword. The references are to be treated only as broad guidebooks,complementary to the lecture notes.


macroeconomics: a dynamic general equilibrium...

Documents