When I left Deutsche Bank, I claimed in a press release that I would spend my time ‘in musical and academic pursuits’. Naturally, my soon-to-be-ex-colleagues thought this was total bullshit and that I would turn up at some hedge fund within weeks of ‘retiring’. But, in fact, I was as good as my word.
If you’ve come to this blog site it will not have escaped your attention that one of the things I did since retirement was to write a book (since the site itself is a thinly-veiled series of adverts for it). Friends are aware of my ‘musical pursuits’ since I drive them crazy plugging tickets to shows I’m in. But the other academic thing I did was to study at the LSE for a Masters in economic history – a course I have just graduated from.
The first thing that the course did was to make me feel exceptionally old. I was easily capable (in terms of age) of being the father of the vast majority of my fellow students. After one seminar on banking crises that I led, a young Dutch chap congratulated me on the presentation: “It was great to hear about this stuff from someone who actually lived and traded through it. I suppose the Crash of ’29 must have been the most stressful, right?” Cheeky bastard.
The other effect was to give rigorous theoretical underpinning to my previously untrained thoughts on the impact of technology on economics and business. A theme of my book is that the onrushing development of computer and communications technology destabilised the banking markets and led to unintended – and sometimes disastrous – consequences. Under competitive pressure, firms’ use of technology took markets to the very limits of the regulatory framework that was in place. I found that this is a common theme in the economic history of business.
Thus, when it came to choosing a topic for my Masters dissertation, the history of banking was a natural place to look. I quickly homed in on one question in particular: what had happened to the English banking market in the late 19th and early 20th centuries? It was a puzzle.
In Victorian times, Great Britain was the world’s geopolitical, economic and financial hegemon. Its banking system underwrote the Gold Standard by which the financial and trade flows of the world were governed. From the mid-19th century onwards its domestic banking market was regulated under a regime that was extremely laissez faire – especially when it came to the question of mergers and takeovers. No real impediment to concentration existed.
Despite this, as late as 1890, there were over 500 banks in England. By 1920, though, there were 18. The market share of deposits of the top five banks had soared from 27% to 80%. It was gigantic process of concentration as a result of an unprecedented wave of mergers; but why had it happened?
The problem had been explored before, but to me the explanations seemed a little implausible. One was that firms had merged because to do so would allow them to create economies of scale. A sensible suggestion, but the record of firms’ accounts didn’t bear it out. Cost income ratios had actually risen in the period and profitability had dropped.
Another explanation in previous studies was that banks merged in order to gain market power. But if they did, why didn’t their profitability rise? And, what’s more, if they wanted market power, why wait until 1890? Market power could have been achieved decades before.
A third idea was that banks combined in order to keep pace with the mergers going on between their industrial customers. But the merger wave in English industry was limited to two, admittedly large, sectors (textiles and brewing) and – crucially – happened after the merger wave in banking had already started.
Given my monomaniacal obsession with banking and technology, I felt that some technological reason must be the answer. Thus I turned to economic theory.
A vital strand of the modern study of the firm is the idea of Transaction Cost Economics. First proposed in a hugely influential paper from 1937 by LSE-alumnus and future-Nobel-prize-winner Ronald Coase, Transaction Cost Economics explains why firms exist within an economy and explains the forces which determine how large they will grow – questions completely ignored by classical economics which assumes an economy of individuals all dealing with other. [i]
Firms exist, according to Coase, to reduce the transactions costs of organizing economic activity. Imagine the huge overhead involved in hundreds of self-employed individuals trying to get together to do something complex (build an oil refinery, say): each would have to have a legal agreement in place with each of the others. It would be a nightmare. Firms reduce the friction by agglomerating these agreements internally – they minimise transactions costs.
So why don’t firms grow without end? In part because of regulation, but also because, even without such limitations, there is a point at which the benefits of size (economies of scale, market power etc.) are outweighed by the costs of trying to organise the firm internally. There is a balance between the costs and benefits of scale.
And there, right in the middle of one of the most influential economics papers of all time, was the clue I had been looking for: “Changes like the telephone and the telegraph which tend to reduce the cost of organizing spatially will tend to increase the size of the firm. All changes which improve managerial technique will tend to increase the size of the firm”. Had the trigger that had enabled the explosion in the scale of banking been the rapid rise in the power of 19th century communications technology: railways, telegraphs and telephones? It seemed plausible.
But how could I test the theory? I was stumped until I had an ‘a-ha!’ moment on a long train journey from London to Newcastle. “Thank heavens”, I thought to myself as I sat staring out of the window, “I never had to take the train to Newcastle every week when I worked for Deutsche”. Then, a few seconds later, “It would have been even worse twenty years ago; and imagine the hassle for bank managers in Victorian times!” A-ha! Speed and distance! Maybe Victorian bank managers would have been more willing to buy a target further away if they could get to it, talk to it, and control it more easily than by riding to it on horseback? Could I observe the impact of technology in the progression of the distances between merging firms over the course of time?
The light bulb moment was the fun bit. Back in London I then had to sit down and catalogue (by leafing though century-old banking magazines in dusty archives or by squinting at ancient records while hunched over a computer for week after tedious week) every one of hundreds of English banking mergers over 100 years: the type of firm; the location of buyer and seller; the distance by road between them.
Happily, the result was pretty persuasive: there had been a steady rise of merger distance from 1820 onwards and an explosion of merger distance from 1890 – just as the telephone network grew. Modelling the distances using indexes of railway, telegraph and telephone coverage showed a very strong statistical impact from the advent of the telephone.
The anecdotal evidence was also strong: records showed that banks had been very early adopters of the phone and there was solid evidence that they had used phones to centralise control of their growing empires – just as theory suggested. It wasn’t proof (what can be proved in history?) but, because the statistics, the economic theory and the anecdotes all lined up, it looked extremely plausible for me to claim that technology had destabilised 19th Century English banking and had led to the rise in market concentration. The LSE examiners agreed.
So why does this all matter? I would claim that it matters because it demonstrates once again, in a historical context, the critical importance for banking of the interplay between regulation and technology – the same interplay that resulted in the huge rise in concentration in 1990s American banking (as regulation relaxed and computers and communications improved exponentially).
But even more importantly, I believe that it suggests that this interplay will continue to be crucial. Thus, as a society, we are posed with many difficult questions: what type of banking industry will we end up with if computers continue to get exponentially faster? What is the plan? Who will regulate it all? Ultimately, who is in charge?
The questions are also vital: whether we get them right may determine the success or failure of the global economy for years to come.