Conglomerates are dead. Long live the conglomerate

When GE announced recently that it was splitting itself into three, it triggered a rash of articles declaring the end of a defining chapter in corporate history, signalling among other things just ‘how far from favour the conglomerate form has fallen’ (FT).

Like many such obituaries (and these were far from the first), this one contains an element of truth: it’s just not the obvious one. It’s true that the capital markets don’t like industrial conglomerates such as GE. But they don’t like any companies that are badly managed, and since they think industrial conglomerates are badly managed by definition, the argument is circular. The reason they think they are badly managed (and therefore dislike them) is that sometimes and at some stage they are worth less than the sum of their parts – a crime against shareholder value – and therefore should be broken up and reallocated to managers who will put shareholders first. Of course the argument is self-reinforcing. When he retired, Jack Welch handed his successors a poisoned chalice in the shape of GE’s toxic financial services division. The longer it festered, the more GE fell out of favour, the more the vultures gathered, until the break-up became inevitable.

It may be that the break-up is in the best interests of the individual units (comprising what were GE’s aviation, healthcare and energy divisions). The best argument in its favour is the extraordinary transformation wrought in GE’s lowly appliances division since it was divested in 2016. The white goods unit, an also-ran in the US market, was ‘dying on its feet’, in the words of its CEO Kevin Nolan, when it was bought by the Chinese company Haier. Liberated (there is no other word) from its over-dominant parent, except in name today’s GE Appliances bears no resemblance to the old company, and the venerable GE brand, to which Haier cannily retained the rights, is now the fastest growing in the market. 

Driven by Haier’s ambition, that is obviously a win, one that in practice couldn’t have taken place within the old GE embrace (which is the real bad management). But for a less favourable de-conglomeration, consider the sad case of ICI. For long the bellwether of British industry, postwar ICI was was an international science-based powerhouse whose policy of channelling returns from its mature chemicals business into innovative new ones was responsible almost single-handedly for developing the skills and knowhow behind the UK’s most successful 20th century industry, big pharma. As John Kay relates, it took 20 years of losses before ICI struck gold with the commercialisation of beta-blockers, one of the industry’s first blockbuster drugs.

ICI was, in other words, a conglomerate that worked. But that didn’t prevent it from falling foul of the break-up fashion. After being put in play by Hanson, the era’s most voracious break-up exponent, in the 1990s, ICI capitulated by floating off its pharma unit, Zeneca, to focus on its ‘core’ chemicals. Ironically, ICI was less good at the short-term shareholder-value game than building new businesses for the long term, and in 2003 the remains of what had been the UK’s biggest and most advanced company was ignominiously sold off to a Dutch competitor. Any temporary benefit to shareholders from ICI’s break-up was more than nullified by the long-term institutional harm done to the UK’s industrial, skills and management base by the ideological sacrifice of one of its few world-class companies on the altar of shareholder value. 

As often happens in management, the break-up fashion emerged from a previous excess in the opposite direction: namely, the conglomeration phenomenon that swept the US in the 1960s. An early form of financialisation, it was a growth-by-acquisition strategy using clever accounting techniques to ensure a positive effect on the acquirer’s price-earnings ratio, in theory enabling it to repeat the process ad infinitum. The result was the meteoric rise of companies such as Ling-Temco-Vought, ITT Corporation, Litton Industries,Textron and Teledyne – and their equally vertiginous fall when slowing economic growth and rising interest rates revealed that in the real world indiscriminate diversification offered no protection against a sharp economic downturn, rather the reverse. 

The problem with this fake conglomeration wasn’t the impossibility of managing a multi-industry group as such, but the idea that it could be done uniquely by the numbers, no industry knowhow required. (Harold Geneen, the legendary boss of ITT, is reputed to have said that when he had finished with it, the group could be rung by a monkey.) As usual, the wrong lesson was learned. Managers were accused of self-aggrandisement and the feathering of their own nests rather than looking after the interests of shareholders, which would be better served by running simple companies that were easier to understand and allowed investors to do their own diversification. More irony, the remedy was the shareholder primacy regime which was much worse than the original ill and whose toxic legacy is disruptively playing out across the western world to this day.

History doesn’t repeat itself – it rhymes, as the saying goes. So, yes, conglomerates are out of fashion – except when they are called private equity groups, some of which, confusingly, have turned themself into public companies. But of course the point of private equity is to maximise shareholder value, so in the eye of admirers, they can’t be called conglomerates. Likewise  Berkshire Hathaway, which just happens to represent one of the greatest single investment vehicles of all time, also escapes the dreaded classification. And what about the tech behemoths? On its own account, Apple runs both hardware and software, and services ranging from music and TV to filmmaking, Amazon stretches from shops, both virtual and bricks and mortar (as does Apple), to logistics, health, cloud services and advertising. Each is at the centre of wider ecosystems (transport, health, to name but two) where traditional industry boundaries and categories become hopelessly blurred – in autonomous electric vehicles more of the value will reside in electronics and software than in conventional hardware, for instance. So what is a car company?

Conglomerates are dead. Long live the conglomerate.