"Inductive reasoning about novelties, and reflexive expectation formation" by Timo Ehrig

Past Event

29 January 2013, 1:30pm - 3:00pm

Said Business School
Park End Street, Oxford OX1 1HP

This seminar is organised by the Oxford Martin Programme on Complexity and the CABDyN
Complexity Centre

Speaker: Timo Ehrig, Research Scientist, Max Planck Institute for Mathematics in the Sciences

Abstract: In this talk, Timo Ehrig will explore two topics:

  1. decision-making when we face novelties
  2. how beliefs about other agents' beliefs matter for price and interest rate formation.

The first topic is about how we (should) make decisions when probabilities are unavailable - how we
(should) make use of induction. In particular, I will explore how expectations for novel opportunities -like
Amazon in the perspective of 1998- (should) come about.

To form such expectations deliberately, decision-makers need to make use of induction: In their reasoning, they derive plausible conclusions that go beyond the information in the premises, for instance, by using analogies. Taking a normative perspective, I explain why asymmetric expectations among individuals can exist, even when information is symmetric: Differences in the epistemic entrenchment of (defined as ’a preference ordering over') elementary inductive explanations of the novelty bring about heterogeneity in final expectations. I identify the skills of decision-makers to generalize, to detect and resolve inconsistencies, and to distinguish inductive explanations that have an effect on an important consequence from others as psychological sources of advantages in forming expectations.

For the second topic, I will present a theory model (which is joint work with Juergen Jost) to explain how uncertainty about the beliefs of other market participants flows into interest rates. If there are coordination uncertainties, prices or interest rates may not only reflect agents' beliefs regarding fundamentals, but also their beliefs about each other. For instance, if short term debtors decide whether or not to roll over the debt of a firm that may default, they also need to anticipate the beliefs of other short term debtors, as their behavior co-determines the default risk of the firm. We present a model in which we study how uncertainty about other agents' beliefs flows into interest rates, and how agents read higher order uncertainty from interest rates. That coordination risks and corresponding dynamics of higher order beliefs play an important role in driving expectations during times of crisis has already been articulated, for instance by Morris and Shin in context of the global games literature. However, the analysis in global games is not complete. In particular, prices or interest rates do not serve as signals in global games. In our model, we can study systematically how interest rates reflect and mediate higher order beliefs, and how interest rates drive coordination dynamics when there are default risks. Our model consists of a strategic market game with default risks, in which agents reflexively reason how the beliefs of other agents may evolve over time.

For further information please contact the Cabdyn Administrator: info.cabdyn@sbs.ox.ac.uk, 01865 288785

Sandwiches and drinks will be provided