When it comes to leaving a city, companies and markets are very different creatures. Companies will shift from one global city to another in search of a more favourable tax regime or more beneficial employment law. For markets — particularly complex ones with strategic functions — it is much harder to move wholesale. The City of London is such a market.

Given London’s uncertain status following the Brexit vote, the predictions about the effect on Hong Kong’s financial market when China was about to take over sovereignty from the British in 1997 are useful. Beijing aimed to make Shanghai China’s main financial centre. The idea was that important Hong Kong markets would migrate to the mainland. That did not happen. Hong Kong is still China’s most international and strongest market. Shanghai, meanwhile, did expand as a financial centre but one geared more towards domestic investors.

One possible outcome for London of leaving the EU is that more routine functions, such as euro clearing, move elsewhere while the more complex markets remain. This possibility is supported by the fact that vital to the more complex London markets are strategic intermediate functions that depend on a mix of very specific types of companies and skills. This is a deeply networked function — that is, companies need each other because each individual company has only partial knowledge or highly specialist capabilities. No single business can give the leading financial centres everything it needs, as the corporate behemoths than once ruled these economies did in the 1960s and 1970s.

It is wrong to think, as many do, that our global cities are about headquarters. One feature of the globalisation that took off in the 1980s is the rise of strategic intermediate sectors, with a vital role for finance and innovative legal, accounting and digital techniques. If a business operates in 20, 30 or more countries it can no longer hire full-time staff for all its high-level know-ledge needs. If you want to do deals in Mongolia and you need 30 hours of Mongolian legal, accounting and financial advice, you buy a service, you do not hire staff. What you need is access to important markets that enable this. Our globalised economy has raised the importance of intermediation to un-precedented levels, and financial centres are the Silicon Valleys of this domain.

The rise of intermediation as an essential part of our increasingly globalised economy also explains the rapid increase in the number of global cities from the 1990s onwards. Today, we can identify 100 or more such centres, albeit with very diverse powers to shape broad global trends and very diverse capacities to develop and invent new instruments. One, mostly overlooked, fact is that even minor global cities have invented new instruments and built new markets, often based on a single commodity — as with Buenos Aires and sunflowers, for instance.

The global city function is made over time, and the process of making it is complex and multi-faceted. And, even though it is highly digitised, it is situated in concrete places.

This also explains the proliferation of global financial centres. We could have moved towards one global superfinancial centre, as many in the late 1980s thought we would. I did not. At the time, I saw three rising global financial centres — New York, London and Tokyo — that were, crucially, quite different from one another. Eventually we saw a proliferation of global financial centres, each with its own specific specialised strengths. In Europe alone, London, Frankfurt and Paris are the leading centres, but they are very different in terms of specialist capacities.

It makes one think that it is to Frankfurt’s advantage to keep London’s most complex financial functions in London.