Scribd
Upload
17 views9 pages

Transversality Vs No-Ponzi Game Condition

The document discusses the Transversality Condition in the context of macroeconomic optimization, presenting a framework for determining optimal sequences under certain assumptions about bounded, continuous, and concave functions. It contrasts this with the No-Ponzi Game condition, emphasizing that the transversality condition serves as an optimality criterion rather than a constraint. The document also explores the implications of these conditions for consumer behavior in economic models.

Uploaded by

zenith6505
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
17 views9 pages

Transversality Vs No-Ponzi Game Condition

The document discusses the Transversality Condition in the context of macroeconomic optimization, presenting a framework for determining optimal sequences under certain assumptions about bounded, continuous, and concave functions. It contrasts this with the No-Ponzi Game condition, emphasizing that the transversality condition serves as an optimality criterion rather than a constraint. The document also explores the implications of these conditions for consumer behavior in economic models.

Uploaded by

zenith6505
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 9

Macroeconomics Sequence, Block I

Transversality Vs No-Ponzi Game Condition

Nicola Pavoni

October 4, 2016
The general framework (deterministic)

∞
V ∗ ( x0 ) = sup ∑ β t F ( xt , xt + 1 ) (1)
{xt +1 }t∞=0 t =0
s.t. x0 ∈ X
xt + 1 ∈ Γ ( xt ) for all t.

Time invariant function F , and correspondence Γ; β ∈ [0, 1). We


assume Γ to be non empty for all x ∈ X .

Recall the Euler Variational approach: the Transversality Condition


The Transversality Condition
Proposition Assume F is bounded, continuous, concave, and
differentiable. Moreover assume Γ has a compact and convex
graph. (i) If the (interior) sequence {xt∗ }t∞=1 with xt∗+1 ∈ intΓ(xt∗ )
for any t = 0, 1, 2, ... satisfies

F2 (xt∗ , xt∗+1 ) + βF1 (xt∗+1 , xt∗+2 ) = 0 for t = 0, 1, ... (2)

and for any other feasible sequence {xt }t∞=0 we have

lim βT F1 (xT∗ , xT∗ +1 )(xT − xT∗ ) ≥ 0, (3)


T →∞

then {xt∗ }t∞=1 is an optimal sequence.


(ii) If in addition F1 (x, x ′ ) > 0 for any x, x ′ ∈ intX and
X ⊆ IRl+ , the condition (3) can be substituted by

lim βT F1 (xT∗ , xT∗ +1 )xT∗ ≤ 0.


T →∞
General Interpretation
The transversality condition requires any alternative trajectory
{xt } satisfying

lim βt F1 (xt∗ , xt∗+1 )(xt − xt∗ ) < 0


t →∞

to be infeasible.

If given {xt∗ } it is impossible to reduce the limit value of the


optimal stock by choosing xt 6= xt∗

(except by incurring in an infinite loss because {x } is not feasible)

Then the value of the capital has been exhausted along the
trajectory, and {xt∗ } must be optimal as long there are no finite
period gains (the Euler condition).
Proof of Transversality
(i) We are done if we can show that for any feasible path we have
T T
lim ∑ βt F (xt∗ , xt∗+1 ) ≥ lim ∑ β t F ( xt , xt + 1 ) ,
T → ∞ t =0 T → ∞ t =0

From the concavity and differentiability of F we have

F (xt , xt +1 ) ≤ F (xt∗ , xt∗+1 ) + F1 (xt∗ , xt∗+1 )(xt − xt∗ ) + F2 (xt∗ , xt∗+1 )(xt +1 − xt∗+1 )

Multiplying by βt and summing up the first T terms one gets


T T
∑ βt F (xt , xt +1 ) ≤ ∑ βt F (xt∗ , xt∗+1 ) + DT , (4)
t =0 t =0

where DT  =
t F (x ∗ , x ∗ )(x − x ∗ ) + F (x ∗ , x ∗ )(x ∗
∑T

t =0 β 1 t t +1 t t 2 t t +1 t + 1 − xt + 1 ) .
Proof (Continued)
We want to show that in (4) limT →∞ DT ≤ 0.

We can rearrange the terms and obtain that


T −1
DT = ∑ βt [F2 (xt∗ , xt∗+1 ) + βF1 (xt∗+1 , xt∗+2 )] (xt +1 − xt∗+1 ) +
t =0
− βT F1 (xT∗ , xT∗ +1 )(xT − xT∗ ).

Euler conditions (2) guarantee that the fist T − 1 terms go to zero:

lim DT = − lim βT F1 (xT∗ , xT∗ +1 )(xT − xT∗ ) ≤ 0.


T →∞ T →∞

the last inequality is implied by the transversality condition.


(ii) if F1 > 0 and xT ≥ 0,

lim βT F1 (xT∗ , xT∗ +1 )(xT − xT∗ ) ≥ − lim βT F1 (xT∗ , xT∗ +1 )xT∗ ≥ 0.


T →∞ T →∞

Q.E.D.
No Ponzi Games vs Transversality
Consider a consumer facing a constant path of income and can buy
and sell in all future markets at price pt .

Problem I (Arrow-Debreu): max


{ct }t∞=0
∑ βt u (ct )
t
s.t. : ct ≥ 0, for all t, and
∑ pt ct ≤ ∑ pt y ,
t t

Suppose, we have pt = βt . Clearly, ct∗ = y for all t.


Now, denote by bt the level of debt at period t : (b0 = 0). The
agent can borrow and lend. pt = βt corresponds to 1 + r = 1β .

Problem II (Radner): max ∞


{ct ,bt +1 }t =0
∑ βt u (ct )
t
s.t. : ct ≥ 0, and
ct + bt +1 ≤ (1 + r )bt + y , for all t.
No Ponzi Games (Continued)
Recall 1 + r = 1β . Multiply by βt the BC in period t, and
rearranging the sequence of per-period budget constraints, for all T
T T
∑ βt ct ≤ ∑ βt y − βT bT +1.
t =0 t =0

A necessary condition for having the same solution in Problems I


and II is to impose limT →∞ βT bT +1 = 0 for all sequences.

Since the agent would never over-save, it suffices to require

lim pT bT +1 = lim βT bT +1 ≥ 0, (5)


T →∞ T →∞

The usual form of the NPG.

NB: The NPG is imposed as additional condition on Problem II.


Transversality
The transversality condition is an optimality condition, not a
constraint.
Since in this spefic problem, the optimal path is bt∗ = 0 for all t,
the transversality condition for this problem will be

lim βT u ′ (cT∗ ) bT +1 ≥ 0. (6)


T →∞

Note: The NPG condition does not contemplates the (subjective)


marginal utility of the agent. The transversality does. Here β is
the subjective discount factor of the agent not a price in the BC.

Sometimes one could get confused between the two conditions:


1 When u ′ is finite one can disregard it in the transversality.

2 The equilibrium price of period t consumption goods takes the


u′ c ∗
value pt = βt u ′ ((ct0 )) .
Recall the NPG condition: limT →∞ pT bT +1 ≥ 0. It clearly
resembles to condition (6) when u ′ (c0 ) = 1 (normalization).

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy