UCL
Université catholique de Louvain Louvain School of Management / CESCM
SUPPLY CHAIN MANAGEMENT
Dynamic games in SCM
Per AGRELL
[email protected]
Outline Dynamic games Biform games Folk theorem Two strategies – Grim trigger – Tit for Tat
Examples
Dynamic games Multi-period games – With time-dependency Inventory games, salvage values, investments – Without time-dependency Repeated games (Nothing links specific games, used more in reputation) Debo (1999), Taylor and Plambeck (2003), Ren et al. (2003)
Repeated games 1
2
3
…
Set of NE much larger than in static games May include NE not found in static games Trigger strategies – May induce coordination if credible threat – Implicit collusion
Biform games Two periods Period 1 – Simultaneous move, non-cooperatively
Period 2 – Cooperative game
Anupindi et al. (1999, 2001)
Concepts Sequential games
Folk theorem Given an infinite sequence of payoffs r1, r2, . . . for player i, the average reward of i is Given a discounting factor 0< b <1, the future discounted reward for player i is A payoff profile r is enforceable if r >= min max u(i){s(i),s(-i)} If r is a NE payoff vector for any game G, then r is enforceable. If r is feasible and enforceable, then r is a NE for the infinitely repeated G with average rewards.
Strategies Certain strategies may be used in repeated games to induce a specific payoff r – Grim trigger – Tit for Tat
Credible strategies A threat will work only if it is credible A credible strategy is not dominated in a subgame
Grim trigger Play COOP until any deviation (cheating) – Then play NON-COOP (static NE) for rest of the game
Idea: – Players compare discounted payoff of COOP vs NON-COOP. – Cheating carries the cost of difference in profit for the rest of the game
Tit for Tat Play COOP until any deviation (cheating) – Then play NON-COOP for K≥ 1 rounds – Play COOP if counterpart played COOP last round.
Idea: – With K =1, Tit-for-Tat is just the mirror of the counterpart’s action. – Tit-for-Tat has shown to be a credible strategy in dynamic games – The strategy is better than Grim for “trembling hands”, i.e. probabilistic moves
References Cachon and Netessine (2004) – Game Theory in Supply Chain Analysis. Ch 6 in David Simchi-Levi, S. David Wu and Zuo-Jun (Max) Shen (Eds), Supply Chain Analysis in the eBusiness Era, Kluwer.
Plambeck and Taylor (2007) – Optimal relational contracts; capacity investments
Agrell and Kasperzec (2006) – Joint investment games, double moral hazard