Ch. 2: Self-Selection Models in Corporate Finance 71
coverage—might affect the probability of winning an underwriting mandate. A second
and perhaps more difficult issue is that of cross-sectional correlation. The 16,000+ trans-
actions in the Ljungqvist et al. sample occur over overlapping periods, which leads to
commonality across transactions and potential cross-sectional correlation in the distur-
bance terms.
- Diversification discount
The scope of the firm is an issue that has occupied economists sinceCoase (1933).One
issue in this literature has been whether firms should diversify or not. While the question
can be examined from several perspectives, a now well developed literature in finance
investigates the diversification question from a valuation perspective. Does diversifica-
tion impact firm value, and if so, in what direction, and why does diversification have
this effect? Our review of this literature focuses on self-selection explanations for diver-
sification.Chapter 8(Maksimovic and Phillips, 2007) provides a more complete review
of the now vast literature on diversification.
The recent finance literature on diversification begins with the empirical observation
that diversified firms trade below their imputed value, which is the weighted average
value of stand-alone firms in the same businesses as the divisions of the diversified firm
(see, e.g.,Lang and Stulz, 1994, Berger and Ofek, 1995, andServaes, 1996). The dif-
ference between the actual and imputed values is called the diversification discount.
The existence of a diversification discount is frequently interpreted as a value destroy-
ing consequence of diversification, although there is no consensus on the issue (e.g.,
Chevalier, 2000, andGraham, Lemmon and Wolf, 2002). We review three papers that
discuss the role of self-selection in explaining the source of diversification discount.
10.1. Unobservables and the diversification discount:Campa and Kedia (2002)
Campa and Kedia (2002)argue that firms self-select into becoming diversified and that
self-selection explains the diversification discount. They model the decision to become
diversified using a probit model
Dit=1ifZitγ+ηit> 0 , (66)
Dit=0ifZitγ+ηit 0 , (67)whereDitis a diversification dummy that takes the value of 1 if the firm operates in
more than one segment, and 0 otherwise, andZitis a set of explanatory variables. The
notations are adapted to match that in Section2. Excess valueVitis specified as
Vit=d 0 +d 1 Xit+d 2 Dit+it, (68)whereXitis a set of exogenous observable characteristics of firmiat timet. Coeffi-
cientd 2 is the key parameter of interest. If it is negative, becoming diversified causes