1

Dynamic Moral Hazard

Intertemporal risk-sharing Better information (output, actions, consumption) Larger games (action spaces) Generic complexity (?spot contracting)

No savings or borrowing. Principal chooses: wi, wij ; Agent: , ai. maxwi,wij

h i P P 1 2 pi() V (qi wi) + j pj (ai)V (qi wij ) , i

s.t. , ai arg max AG(, ai, wi, wij ), and IR. Euler equation:
2 1 V 0(qi wij ) V 0(qi wi) X = pj (ai) u0(wi) u0(wij ) j

Simple M (separable): t = 1, 2.
t a A, #Q = n, Pr(qi = qi|a = at) = pi(at) > 0.

"

When V 0 = const, we have smoothing

"

Agent: u(c) (a) (in each t), limc u(c) = . c Principal: V (q w).
1 2 1 2 Contracting: t = 1: {a1, w1(qi ), a2(qi ), w2(qi , qj )}. (RP)

X 1 1 = pj (ai) 0 . 0(w ) u i u (wij ) j

Two observations: (1) Optimal contract has memory, No memory would imply RHS is constant for all i, perfect insurance in period 1, wrong incentives.

(2) Agent wants to save (and so the contract is frontloaded).

EU = P p (a )u0(w ) u0(w ) 0 (Jensens inij i j j i s

Free savings. Example: Eort in t = 2, consumption in both periods (borrowing in the rst period) a {H, L}, (H) = 1, (L) = 0. q {0, 1}, pH = p1(H) > pL > 0. Suppose a = H. Contract (w0, w1).

equality).

Monitored savings Add ti, si (principal, agent)s savings. The above contract can be achieved without historydependent wages, and, so, is spot-implementable. Set: cij = wij = wj + si, wi = ci si. Problem separates to: incentive provision and consumption smoothing.

Let cj be consumption with planned j = H, L. cj arg maxc u(c) + pj u(w1 c) + (1 pj )u(w0 c). We have u(cH ) + pH u(w1 cH ) + (1 pH )u(w0 cH ) 1 = = u(cL) + pLu(w1 cL) + (1 pL)u(w0 cL) > u(cH ) + pLu(w1 cH ) + (1 pL)u(w0 cH ) Thus ICH2 is slack. Room for renegotiation (unless CARA)

1.1

T-period Problem

Subcases:

Repeated Output (better statistical inference) Repeated Actions (multitask in time) Repeated Consumption (consumption smoothing) Repeated Actions and Output (consumption at the end)

Innitely repeated Actions, Output, and Consumption.