Firm size, market conditions and takeover likelihood

Abstract

Purpose The firm size hypothesis – takeover likelihood (TALI) decreases with target firm size (SIZE) – has enjoyed little traction in theTALImodelling literature; hence, this paper aims to redevelop this hypothesis while taking account of prevailing market conditions – capital liquidity and market performance. Design/methodology/approach The study uses a logit modelling framework to modelTALI. Model performance is assessed using receiver operating characteristic (ROC) curve analysis. The empirical analysis is based on a UK sample of 34,661 firm-year observations drawn from 3,105 firms and 1,396 M&A deals over a 30-year period (1987-2016). Findings While acquirers generally seek smaller targets because of transaction cost constraints, the paper shows that the documented negative relation betweenSIZEandTALIarises from sampling bias. Over a full sample, mid-sized firms are most at risk of takeovers. Additionally, market conditions moderate theSIZE–TALIrelationship, with acquirers more inclined to pursue comparatively larger targets when financing costs are low and market growth or sentiment is high. The results are generally robust to endogeneity. Research limitations/implications Sample truncation on the basis ofSIZEleads to empirical misspecification of theTALI–SIZErelation. In an unbiased sample, an inverse U-shaped specification betweenTALIandSIZEsufficiently models the underlying relation and leads to improvements in the predictive ability ofTALImodels. Originality/value This study advances a new firm size hypothesis which is consistent with classic M&A theories. The study also evidences market conditions as a moderator of the acquirer’s choice of targetSIZE. A new model specification which recognises the non-linear relation betweenTALIandSIZEand accounts for the moderating effect of market conditions on theSIZE-TALIrelationship leads to improvements in the performance ofTALIprediction models.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 26 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.