Goldman Sachs and the lessons for co-CEOs

Stay informed with free updates

Simply sign up to the Inside Business myFT Digest — delivered directly to your inbox.

As pressure continues to pile on Goldman Sachs boss (and part-time DJ) David Solomon, the Wall Street rumour mill has been throwing out new theories about his potential replacement.

According to one report in June that has not been corroborated elsewhere, a co-chief executive plan is being discussed “at the highest levels”, with Jim Esposito, who co-heads banking and markets, and Marc Nachmann, the asset and wealth management chief, the favourites to secure the twin prize.

The idea may come to nothing, but it resonated in Europe last week, as a smaller bank-cum-asset manager announced just such a split of the top job. Vontobel, the venerable Swiss group, has appointed Christel Rendu de Lint and Georg Schubiger as co-CEOs. The pair hail from the group’s two core businesses of investment and wealth management and will make a balanced team to smoothly implement a previously agreed long-term strategy. That, at least, is the theory.

The precedents are hardly stellar. First Republic, one of a trio of US regional banks that failed earlier this year, experimented with the co-CEO structure. The first incarnation of that had broken down a year earlier after Hafize Gaye Erkan (now governor of Turkey’s central bank) had resigned after just six months in the job.

The co-CEO idea in that case was a succession-planning tactic to pass the baton gradually from bank founder James Herbert to the next generation. Erkan’s replacement Mike Roffler was made acting co-CEO before being appointed sole CEO of the bank — with Herbert moving on to the position of executive chairman. The continued overlap of executive power, though, might have contributed to the lack of assured leadership when First Republic was caught off guard by a frenzy of withdrawals by panicked depositors, triggering an eventual rescue by JPMorgan Chase.

When Anshu Jain and Jürgen Fitschen were appointed co-CEOs of Deutsche Bank in 2011 it was for a different logical reason: a desire to reconcile the two engines of the group — a vast Anglo-American investment bank that Jain had helped build, and a domestic German operation of which the Anglo-Indian Jain had scant understanding. Infighting and a resistance to the changes wrought by the 2008 financial crisis contributed to poor performance. By 2015 both men had resigned, leaving Deutsche in a weak state that persisted for years.

Alongside finance, it has been the tech industry that has most enthusiastically embraced the co-CEO model. For a period it was the height of tech fashion. Netflix, Salesforce, Oracle, SAP and BlackBerry (then called Research In Motion) all tried it. With the exception of Netflix, they abandoned the experiment — though in Blackberry’s case only after the co-CEOs had all but killed the company by failing to respond to a fast-evolving smartphone market.

And yet, for all the power of such bad-news evidence, the average company run by co-CEOs has significantly outperformed, according to a study published by the Harvard Business Review. The authors found that there were nearly 100 US-listed companies (out of 2,200 in total) that were run jointly at some point over a 25-year period up to 2020. They had generated an average annual shareholder return of 9.5 per cent, compared with a 6.9 per cent baseline.

The structure works best in the tech sector, the study suggests, or when there is a clear division of labour. Co-CEOs “can form a left-brain/right-brain partnership”. Complementary skills may also come as a result of gender balance and some companies have used a co-CEO arrangement to advance management diversity. That is the case at Vontobel, where Rendu de Lint and Schubiger also offset each other in their career backgrounds and styles.

When Goldman was run as an unlisted partnership, it was often led — successfully — by co-heads (a practice that is still a divisional norm). Addressing Harvard Business School two decades ago, one such alumnus was clear about the merits of that tradition. “Two people making a decision will come up with a better decision than one person leading alone,” said John Whitehead, who co-ran Goldman up to 1984. In the decades since, the competing interests of the group’s divisions, particularly in investment banking and trading, have been hard to reconcile via one leader. Whenever David Solomon hangs up his banker’s suit and embraces DJing full-time, a co-CEO structure, done right, might indeed be worth considering again.

patrick.jenkins@ft.com