 Summaries of recent and interesting research papers Conglomerates and diversification
Several recent articles have studied the conglomerate
discount and have sought to test theories developed mainly in the 1970s. We
illustrate this research via four articles that deal with various facets of
the conglomerate discount.
According to classical theory, diversification has the following benefits:
- Economies of scale and a greater number of profitable
activities (Weston, 1970 and Chandler, 1977)
- Greater debt capacity, as risk is lower (Lewellen,
1971)
- An opportunity to make investments more efficient,
by setting up an internal capital market (Stein, 1997)
To these arguments in favour of diversification, there
are the following disadvantages, due mainly to higher agency costs:
- Heavier structural costs (head office costs, etc.);
- Less market understanding of the stock, as asymmetry
of information is greater (for example, which analyst should cover Bouygues:
the telecom analyst, the construction analyst, the media guy or the woman
covering utilities?);
- Inefficient investments and the phenomenon of some
divisions financing others (Berger and Ofek, 1995) and a misallocation of
resources (to low-margin subsidiaries or to subsidiaries offering limited
investment opportunities) due to internal power struggles (Rajan, Servaes
and Zingales and Stulz, 2000);
- A tendency to invest in projects offer net negative
discounted value (Jensen, 1986, and Meyer, Milgrom and Roberts, 1992); the
behaviour of division managers seeking to expand their power base tends to
make the group invest in unprofitable projects and to skew the internal capital
market, such as defined by Stein in 1997 (Scharstein and Stein, 2000).
According to recent financial literature, diversification
leads to higher agency costs than those of a single-business company, and, as
a result, a destruction of value.
Mansi and Reeb (1) have shed new light on the conglomerate
discount. They found that diversification has no impact on enterprise value
of diversified companies. However, the authors found that the decision to diversify
results in a transfer of value from shareholders to creditors.
Their reasoning is as follows: nobody contests the
fact that a company's overall risk is lessened through diversification. This
reduced risk is, by nature, "diversifiable", but the risk of default
of a diversified company is indeed lower than the risk of default of a single-business
company. So, even if shareholders place no value on a group's diversification
(as they can themselves eliminate diversifiable risk in their portfolio), diversification
will be beneficial for creditors. This empirical observation corroborates their
hypothesis: the more leveraged a diversified group is, the more that investors
will apply a conglomerate discount.
Financial wizards among our readers will be quick to
see here the real options theory applied to equity valuation.
Maksimovic and Phillips (2) tested the hypothesis of
misallocation of resources within conglomerates. Their overall findings are
that conglomerates do not misallocate their resources any more than non-diversified
groups. They found that the most productive divisions are allocated most of
the resources, thus respecting good business sense. No, the authors found that
the conglomerate discount is not due to agency issues, but to sheer size - they
found that the production units of a diversified group are less productive than
those of a non-diversified group of the same size. This lack of productivity
is due mainly to diversified companies' smallest divisions.
Graham, Lemmon and Wolf (3), meanwhile, call into question
some recent academic literature on the conglomerate discount. Some recent articles,
point out the conglomerate discount in diversified companies, are based on valuations
of individual divisions that, in turn are based on valuations of single-business
companies in the same sector. The authors demonstrate that there is a bias in
this methodology, i.e. that the very characteristics of these divisions make
them undervalued compared to comparable companies. The apparent discount of
these divisions is seen, among other times, at the moment they are acquired:
groups that are seeking to diversify through acquisitions generally make acquisitions
at a lower valuation than their sector. After the acquisition, there is no reason
for this undervaluation (which may be perfectly justified) to disappear. After
putting the lie to conventional thinking, the authors go no further, for example,
they draw no conclusions on whether diversified groups have a conglomerate discount!
Denis, Denis and Yost (4) take an original approach
to diversification, focusing on companies international diversification. They
demonstrate that US groups tend to seek to increase their international diversification.
Most importantly, the authors point out the existence of an international diversification
discount, which they found is of the same order of magnitude as the sector diversification
discount, i.e. about 20%, and the two types of discount can come on top of each
other. This would tend to show that international synergies are, on average,
exceeded by the costs of maintaining a multinational presence (complexity in
organisation, asymmetry of information between the head office and the various
subsidiaries, as well as all the causes mentioned for sectorial diversification).
With lower international transaction costs, it is now easy for an investor to
buy shares in, say, a Danish or Mexican company. Investors therefore do not
need a company to diversify their international portfolios. The same argument
applies as in sector diversification.
Clearly, opinions diverge on diversification or refocusing when it comes to
value creation. This is fortunate for investment banks, which can, with a clear
conscience continue to merge, demerge, buy up and sell off various companies,
as well as for academics, who still have lots of ground to explore!
(1) "Corporate diversification: what gets discounted?",
Journal of Finance, October 2002.
(2) "Do conglomerate firms allocate resources inefficiently across industries",
Journal of Finance, December 2001.
(3) "Does corporate diversification destroy value?" Journal of
Finance, April 2002.
(4) "Global diversification, industrial diversification, and firm value",
Journal of Finance, due out in October 2002.
|
|
