Not all diversified groups deserve the “conglomerate discount”. Some types are more likely to add value than others.
Conglomerates were in fashion until the 1970s, when Michael Jensen, finance professor and Nobel Prize winner, observed that they suffered from “billions in unproductive capital expenditures and organisational inefficiencies.” This led to the belief that companies or groups with unrelated multi-business portfolios do a lot of things badly rather than a few things well.
As a result, the ‘conglomerate discount’ — the 10-15 percent that penalty markets are thought to impose on diversified multi-division enterprises — entered business parlance. Yet, Berkshire Hathaway and General Electric (GE) have been able to avoid the conglomerate discount, and business groups are rather common in emerging markets.
After many detailed discussions with INSEAD professor Phanish Puranam, I have distinguished between four different types of multi-business portfolios and would argue that one must not necessarily be pessimistic about them all.
- Type A are holding groups like Berkshire Hathaway that ‘hold’ a portfolio of usually listed companies for the long term. The group centre leaves the individual companies to manage on their own without actively searching for synergies between them. HQ views itself as an investment company and, consequently, has few employees (fewer than 30 in Berkshire Hathaway’s case). While they can easily exit a business by selling their holding in it, they do so infrequently.
- Type B are classic conglomerates like GE, where the holding company is listed while individual businesses are unlisted. As strategy evolves, businesses are acquired or divested according to need, but in general they are in businesses they wish to “run” over the long term. The headquarters is substantial in size and actively pursues synergies, deploying common practices and policies across all units. Since the subsidiaries are wholly owned, synergies that are win-win (both cooperating units see gains) as well as win-lose (one unit gains while the other loses, albeit less than the gains of the former) are sought.
- Type C are private equity firms that usually acquire companies and take them out of the market’s view to ‘repair’ them. Both the holding company and the individual business units are unlisted. The reason to acquire is to exit at a much higher valuation within a tightly defined time horizon. The relatively lean headquarters is populated with a few experienced executives with turnaround capabilities. These specialists are deployed to individual units and usually revert back to the private equity firm after the exit. No synergies between individual businesses are sought since they have to be ultimately shed as standalone enterprises.
- Type D are business groups like Aditya Birla Group or Tata, which comprise an unlisted holding company and individual businesses that are usually listed. Business groups are essentially in ‘build’ mode since most of their businesses are incubated in-house. As is to be expected, when sequentially launching new businesses there will be a substantial failure rate, especially compared with the ‘repair’ business of private equity firms. But, overall, successes will compensate for failures in well-managed business groups. In relative terms, the headquarters will be smaller than at Type B conglomerates, but larger than at holding companies and private equity firms. Only win-win synergies will be pursued by the group centre as the rights of minority shareholders in the individually listed companies (as well as joint ventures that may be unlisted) must be protected.
Professors Tarun Khanna and Krishna Palepu of Harvard Business School argued that the popularity and superior performance of business groups in emerging economies stemmed from the poor quality of institutions (e.g., capital markets, talent markets) there. This came to be known as the ‘institutional voids’ theory. Proponents of the theory hypothesised that as emerging markets matured, business groups would go out of fashion.
Since then, various studies have demonstrated that business groups continue to thrive in some well-developed markets such as Singapore and Sweden, whilst there are some emerging markets such as Pakistan and Peru where business groups perform poorly. As a result, the validity of the institutional voids theory was questioned.
A subsequent study examined the performance (on ‘return on assets’) of 10,500 Indian companies between 1994 and 2009, and found:
- Overall in India, companies affiliated with business groups did not outperform companies that were not so affiliated.
- Listed companies affiliated with business groups outperformed both listed companies that were unaffiliated with business groups as well as unlisted companies affiliated with business groups.
- Over time, despite the Indian markets maturing, the performance of listed firms affiliated with business groups improved, contrary to some interpretations of the institutional voids theory.
Listed business group companies obtain the benefits of being affiliated with the business group, combined with the scrutiny that comes from markets that helps to reduce the many disadvantages of belonging to a business group.
The benefits include access to internal capital, talent, technology and product markets at lower transaction costs. In addition, there’s the weight the business group carries with various stakeholders.
Market scrutiny protects business group companies’ performance from the potential downsides of affiliation, such as the security given to managers who do not perform, leaders owing their positions to nepotism more than ability, and costly group functions that do nothing for individual enterprises.
The most convincing evidence of business groups’ inefficiency used to be the cross-subsidies among affiliated companies. Fortunately, increased regulation in India has led to greater minority shareholder protection, and related-party transactions now have to be above board. This has dramatically reduced the practice of forcing companies to buy uncompetitive inputs from other group companies, of having poorly performing companies subsidised by better ones, and of “tunneling”, or moving profits from companies where the business group had a lesser holding to those where its shareholding was greater.
In conclusion, listed companies affiliated with business groups — as at least Indian data indicates — do rather well when compared with other types of companies. And their superior performance is increasing rather than decreasing as markets become more sophisticated.