Turnpike properties of capital accumulation games (Q1082277)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Turnpike properties of capital accumulation games
scientific article

    Statements

    Turnpike properties of capital accumulation games (English)
    0 references
    0 references
    0 references
    1986
    0 references
    The paper deals with a two-person nonzero-sum differential game in which each player i accumulates capital \(K_ i\) according to \(\dot K_ i=I_ i-\delta_ iK_ i\). The quantity \(I_ i\) denotes investment by player i and \(\delta_ i\) is the depreciation constant. The payoff for each player is its total discounted profits, where the instantaneous profit is given by \(\pi_ i(K_ 1,K_ 2)-C_ i(I)\); the function \(C_ i\) denotes the costs of investing. The interaction of the two players is through the functions \(\pi_ 1\) and \(\pi_ 2\). The equilibrium concept used is that of Nash. The information structure is open-loop (i.e. the investments \(I_ i\) are time-functions; in the closed-loop sense \(I_ i\) would depend on the current state \(K_ 1\), \(K_ 2).\) Three turnpike theorems are formulated and proved. Some of the results are generalizations and extensions of other already known results (cited). The first turnpike theorem specifies conditions under which every equilibrium path of the infinite horizon game converges to the unique stationary equilibrium. The second turnpike property describes a relation between the equilibrium paths of the finite and infinite horizon games. The third turnpike theorem deals with conditions under which the finite horizon equilibrium path stays within an \(\epsilon\) neighbourhood of the stationary equilibrium, except for some initial time required to accumulate capital and some final time in which ''end game'' considerations take over. Assumptions under which the turnpike theorems hold, mainly deal with properties of the functions \(\pi_ 1\) and \(\pi_ 2\) and their derivatives. The paper is well-written. The reviewer wonders whether in an extension to the n-person game \((n>2)\) similar theorems can be formulated (with similar proofs).
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    Nash equilibrium
    0 references
    open loop information structure
    0 references
    two-person nonzero-sum differential game
    0 references
    investments
    0 references
    turnpike theorems
    0 references
    infinite horizon game
    0 references
    stationary equilibrium
    0 references
    0 references