Computational Modelling of Behavioural Mergers and Acquisitions Pricing Theory

Abstract

<p>Merger & Acquisition (M&A) pricing has traditionally been performed using financial methods. Computational modelling techniques such as agent based modelling are challenging this status quo in the area of M&A pricing. This thesis uses an agent based modelling approach in combination with non-co-operative game theory and prospect theory to develop a model that incorporates psychological biases in M&A pricing. Traditional finance models cannot include these biases, even when the biases can have a significant impact on the final outcome. Computational modelling is a vast field of study, however the use of agent based modelling allows the research in this thesis to analyse individual and overall dynamics within the M&A transaction model. The agent based model in this thesis looks at two main models, a buyer-seller game between an acquirer and target company (company being purchased by the acquirer) and multiple acquirers bidding for a single target company. Then, there are numerous variations of these two models as prospect theory and cumulative prospect theory are applied to include psychological biases. The base models consider two main psychological traits: risk aversion (a continuum of risk averse to risk taking) for the acquirer and optimism (a continuum of optimistic to pessimistic) for the target company. The model shows that the change in the level of these traits (risk averse compared to risk taking, for example) will have a different outcome based on the type of game and psychological biases associated to them. The model shows, using real world examples of the Verizon and AOL as well as the Verizon and Yahoo mergers for verification, that risk averse acquirers and a pessimistic target company will result in a lower merger price. Further, real world M&A transactions for Sanofi and Ablynx (with Nova Nordisk as the secondary acquirer) and a potential merger between Bunge Limited and Archer Daniel Midlands (with Glencore as another potential acquirer) show that multiple acquirers and an optimistic target will usually result in the merger price being much higher than expected. This primarily occurs as the acquirers bid up the price in competition to meet the price required by the target company. In conclusion, this thesis has developed a computational agent-based model that is intended to provide significant insight into M&A transaction pricing and it is one of the initial studies in this research area.</p>