Abstract
Forward induction, as defined by Govindan and Wilson (2009. “On Forward Induction.” Econometrica 77:1–28), places a local dominance condition on off-equilibrium beliefs that restricts relevant strategy profiles for an equilibrium outcome to be infinitely more likely than profiles that include irrelevant strategies. Meanwhile, it places no global dominance restrictions and thus leaves open the possibility that a dominated strategy is deemed more likely than strategies dominating it. This paper defines strong forward induction, which improves upon forward induction. We also develop a solution concept called strong forward induction equilibrium that is obtained from iterative application of the strong forward induction criterion.
Acknowledgments
I am grateful to Hari Govindan for discussions and comments. I thank Quan Wen, Adam Wong, Wenzhang Zhang, the editor Yuk-Fai Fong and anonymous referees for comments, and the Chinese National Science Foundation for support (Project No. 71573168). I also thank Sambuddha Ghosh for careful reading and suggestions that help improve the paper. Remaining errors are my own.
Appendix A. Proof of Main Result
Proof of Proposition 1. We develop the proof in two steps. First, in Lemma Lemma 1 we show that every stable set contained in a connected component has a strong forward induction equilibrium. We then apply a result in Kohlberg and Mertens (1986) to complete the proof.
Lemma 1
Every stable set contained in a single component of equilibria of game
Proof.
The method of proof here is similar to Man (2012). Let
What this implies is that for any perturbation
Let
and
Since
For
Next, let
For any game
and
Since
We now show that first-order strong forward induction equilibrium exists. By definition of stable set, there exists
Furthermore, every
If
If
Thus, by definition, an equilibrium
Now we are in a position to show that for all
We have yet to show that
Kohlberg and Mertens (1986) show that every finite extensive form game with perfect recall has a stable set contained in a single connected component. Hence, we conclude that strong forward induction equilibrium always exists.
Appendix B. Extensive Form Forward Induction
The extensive form forward induction in GW is weaker than normal form forward induction. Man (2012) shows that it is not an invariant concept. Moreover, as it places restriction on off equilibrium path belief in weakly sequential equilibrium, it fails admissibility as well.
We first provide a brief review of the definition of extensive form forward induction. To start with, the extensive form forward induction is defined as a property of an equilibrium outcome, and built upon weakly sequential equilibrium of Reny (1992).
Definition 6.1
A consistent assessment (
Unlike the sequential equilibrium of Kreps and Wilson (1982), in which players make optimal choice at any point in the game tree, weakly sequentiality requires each player to make optimal choice only at information sets not excluded by her own strategy but puts no restrictions at information sets excluded by her own previous choice. As a result, weakly sequential equilibrium allows players to have any beliefs about an opponent’s continuation play at an information set excluded by her own strategy. As a refinement of weakly sequential equilibrium, forward induction imposes restriction on the beliefs players may hold at information sets excluded by an opponent’s own strategy. As a first step, for an equilibrium outcome, the set of relevant strategies is determined as follows:
Definition 6.2
(DEFINITION 3.3 in GW) A pure behavioral strategy
They then define relevant information set for an outcome (DEFINION 3.4 in GW) as those that can be reached with positive probability by every profile of relevant strategies for that outcome. The definition of extensive-form forward induction then places restrictions on player’s off equilibrium path belief in the following fashion.
Definition 6.3
(DEFINITION 3.5 in GW): An outcome satisfies forward induction if it results from a weakly sequential equilibrium in which at every information set that is relevant for that outcome the support of the belief of the player acting there is confined to profiles of Nature’s strategies and other players’ strategies that are relevant for that outcome.
Because of the restriction imposed by Definition 6.3, at any relevant information set
Let
To define extensive form forward induction that is equivalent to Definition 3.3, we first strengthen weak sequentiality with a form of cautious play as Reny (1992) suggests. Instead of requiring each player’s behavioral strategy
Definition 6.4
An assessment (
The requirement of cautious play suggests that the appropriate definition of relevant strategy become:
Definition 6.5
A pure behavioral strategy
Next, we derive an extensive form definition of forward induction equivalent to GW’s normal form version (Definition 3.1):
Definition 6.6
An outcome
With the modification, extensive form forward induction as defined in Definition 6.6 satisfies invariance and admissibility. However, it still leaves open the relative probabilities of dominated strategies and dominating strategies. Below we provide an equivalent definition of truly relevant strategy in the extensive form and then give the definition of extensive-form strong forward induction. First, let
Definition 6.7
A pure behavioral strategy
Finally, we define extensive form forward induction that is free of inconsistent belief:
Definition 6.8
An outcome
Banks, J., and J. Sobel. 1987. “Equilibrium Selection in Signaling Games.” Econometrica 55: 647–61.10.2307/1913604Suche in Google Scholar
Battigalli, P., and M. Siniscalchi. 2002. “Strong Belief and Forward Induction Reasoning.” Journal of Economic Theory 106: 356–91.10.1006/jeth.2001.2942Suche in Google Scholar
Battigalli, P., and M. Siniscalchi. 2003. “Rationalization and Incomplete Information.” Advances in Theoretical Economics 3.10.2202/1534-5963.1073Suche in Google Scholar
Battigalli, P., and M. Siniscalchi. 2007. “Interative Epistemology in Games with Payoff Uncertainty.” Research in Economics 61: 165–84.10.1016/j.rie.2007.10.003Suche in Google Scholar
Cho, I.-K., and D. M. Kreps. 1987. “Signaling Games and Stable Equilibria.” Quarterly Journal of Economics 102: 179–222.10.2307/1885060Suche in Google Scholar
Govindan, S., and R. Wilson. 2009. “On Forward Induction.” Econometrica 77: 1–28.10.2139/ssrn.961852Suche in Google Scholar
Hillas, J., and E. Kohlberg. 2002. “Conceptual Foundations of Strategic Equilibria.” In In Handbook of Game Theory, Vol. 3., edited by R. Aumann, and S. Hart, 1597–663. New York: Elsevier.10.1016/S1574-0005(02)03005-9Suche in Google Scholar
Kalai, E., and D. Samet. 1984. “Persistent Equilibria.” International Journal of Game Theory 13: 129–44.10.1007/BF01769811Suche in Google Scholar
Kohlberg, E., and J.-F. Mertens. 1986. “On the Strategic Stability of Equilibria.” Econometrica 54: 1003–37.10.2307/1912320Suche in Google Scholar
Kreps, D., and R. Wilson. 1982. “Sequential Equilibria.” Econometrica 50: 863–894.10.2307/1912767Suche in Google Scholar
Kuhn, H. W. 1953. “Extensive Games and the Problem of Information.” In Contributions to the Theory of Games 2: 193–216. Princeton, NJPrinceton University Press.10.2307/j.ctv173f1fh.11Suche in Google Scholar
Man, P. T. Forward Induction Equilibrium Working paper, University of Chicago, 2010.Suche in Google Scholar
Man, P. T. 2012. “Forward Induction Equilibrium.” Games and Economic Behavior 75: 265–76.10.1016/j.geb.2011.12.004Suche in Google Scholar
McLennan, A. 1985. “Justifiable Beliefs in Sequential Equilibrium.” Econometrica 53: 889–904.10.2307/1912659Suche in Google Scholar
Myerson, R. 1978. “Refinements of the Nash Equilibrium Concept.” International Journal of Game Theory 7: 73–80.10.1007/BF01753236Suche in Google Scholar
Pearce, D. G. 1984. “Rationalizable Strategic Behavior and the Problem of Perfection.” Econometrica 52: 1029–50.10.2307/1911197Suche in Google Scholar
Reny, P. J. 1992. “Backward Induction, Normal Form Perfection and Explicable Equilibria.” Econometrica 60: 627–49.10.2307/2951586Suche in Google Scholar
Stahl, D. O. 1995. “Lexicographic Rationalizability and Iterated Admissibility.” Economics Letters 47: 155–9.10.1016/0165-1765(94)00530-FSuche in Google Scholar
van Damme, E. 1989. “Stable Equilibria and Forward Induction.” Journal of Economic Theory 48: 476–96.10.1016/0022-0531(89)90038-0Suche in Google Scholar
© 2017 Walter de Gruyter GmbH, Berlin/Boston
Artikel in diesem Heft
- Research Articles
- Optimal Monetary Policy in an Overlapping Generations Model with Search Theoretic Monetary Exchange
- Getting a Job through Unemployed Friends: A Social Network Perspective
- Dynamic Stability of Post-Keynesian Pricing
- A Nonspeculation Theorem with an Application to Committee Design
- Information Acquisition in the Era of Fair Disclosure: An Application of Asymmetric Awareness
- Privatization Neutrality Theorem in Free Entry Markets
- Strong Forward Induction
- The Case of “Less is More”: Modelling Risk-Preference with Expected Downside Risk
- Stability of Equilibrium Outcomes under Deferred Acceptance: Acyclicity and Dropping Strategies
- Notes
- A Height-Based Multidimensional Extension of the Lorenz Preorder for Integer-Valued Distributions
Artikel in diesem Heft
- Research Articles
- Optimal Monetary Policy in an Overlapping Generations Model with Search Theoretic Monetary Exchange
- Getting a Job through Unemployed Friends: A Social Network Perspective
- Dynamic Stability of Post-Keynesian Pricing
- A Nonspeculation Theorem with an Application to Committee Design
- Information Acquisition in the Era of Fair Disclosure: An Application of Asymmetric Awareness
- Privatization Neutrality Theorem in Free Entry Markets
- Strong Forward Induction
- The Case of “Less is More”: Modelling Risk-Preference with Expected Downside Risk
- Stability of Equilibrium Outcomes under Deferred Acceptance: Acyclicity and Dropping Strategies
- Notes
- A Height-Based Multidimensional Extension of the Lorenz Preorder for Integer-Valued Distributions