Win-Stay, Lose-Shift : A Strategy of Serial Acquirers

Research output: Conference PapersRGC 32 - Refereed conference paper (without host publication)peer-review

View graph of relations

Author(s)

Related Research Unit(s)

Detail(s)

Original languageEnglish
Publication statusPublished - 19 Dec 2017

Conference

Title2017 Auckland Finance Meeting
PlaceNew Zealand
CityQueenstown
Period18 - 20 December 2017

Abstract

We show that serial acquirers over-extrapolate from their own past experiences while making future acquisition decisions: firms likely repeat (avoid) choices that have led to good (bad) outcomes from the past, even after controlling for aggregate time-series shocks, economic factors, rational learning about acquisition skill, and firm fixed effects. We also find that a firm experiencing high announcement returns in early acquisitions has a higher chance of becoming a serial acquirer. Moreover, serial acquirers with greater positive (negative) return experiences are more likely to initiate value-destroying (value-enhancing) mergers in terms of both market reaction and operating performance. This behavior is consistent with a reinforcement learning heuristic. We also discover that higher institutional ownership mitigates serial acquirers’ excessive acquisitiveness following good experiences, whereas financial expertise on corporate boards helps identify value-enhancing deals after bad outcomes. Finally, CEO overconfidence increases after past firm successes, but remains immune to failures. Hence, past successes provoke future mergers by making managers more overconfident whereas negative experiences directly curb serial acquirers’ acquisitiveness.

Research Area(s)

  • Serial Acquirers, Mergers and Acquisitions, Corporate Governance, Reinforcement Learning, Overconfidence

Citation Format(s)

Win-Stay, Lose-Shift: A Strategy of Serial Acquirers. / Bharath, Sreedhar T.; Cho, Duckki; Choi, Lyungmae.
2017. Paper presented at 2017 Auckland Finance Meeting, Queenstown, New Zealand.

Research output: Conference PapersRGC 32 - Refereed conference paper (without host publication)peer-review