Thats up by nine percentage points on what the top 10 took as recently as the second half of 2011 and a return to the depths of the financial crisis at the start of 2009, when with many banks reeling, JPMorgan alone had a 10.2% share, compared with its 8.2% today.
Back at the start of 2009, the markets had largely shut down and, even after they reopened to a wave of crucial capital raisings and refinancings, first-half revenues were a moderate $26 billion. This more recent withering of competition in much healthier, more robust and larger financial markets does not look a particularly welcome development, increasing, as it does, the dependence of corporations, governments, large investors and other banks on just a few providers and along with it the exposure of the global financial system to risk of failure at any of these top 10.
Even as regulators claim to be tackling the risks from so-called too-big-to-fail banks, they appear to have created a system where the biggest banks have such an advantage of scale that they cannot but win market share.
Its easy to find cyclical forces at work. The US market was the biggest share of the investment banking fee pool in the first half of this year and it has been notoriously difficult for non-US banks to penetrate. Debt capital markets also accounted for an unusually high share of the fee pie. No wonder, then, that the top five in order are JPMorgan, Bank of America Merrill Lynch, Goldman Sachs, Morgan Stanley and Citi.
But this trend is structural, not cyclical. Euromoney speaks to the head of investment banking at one of those top US firms that has been picking up market share at a faster pace than the markets themselves have been growing. He suggests this zero-sum game is likely to continue. And once he has finished beating his chest, he admits that regulation has become so burdensome that it is forcing capacity out of the system, leaving a greater share in the hands of fewer large banks.
Barriers to entry have risen. In areas such as global transaction banking, increasingly the key underpinning for wholesale banking relationships, the costs of building from scratch are simply prohibitive. Restrictions on mergers and acquisitions close off entry via that route. There are no rivals on the horizon. The top banks are being protected from competition.
This is why the intense debate about ending the too-big-to-fail problem sounds so fraudulent. By seeking new ways to resolve large banks if they get into trouble, regulators are simply striving to narrow the group potentially exposed down from all taxpayers to just investors in banks equity, debt and deposits. The fact that in many cases these are domestic savers doesnt seem to constitute much progress, nor is it easy to see how this has in any way broken the negative feedback loop between sovereigns and banks. One way or another, these banks would have to be propped up.
Denmark earned fame for being the first European country to bail-in bank creditors of two small mortgage banks in 2011. But resolving Amagerbanken would not be a model for resolving Danske Bank, let alone JPMorgan. And even the Danish National Bank admitted that some banks are so important to the overall economy that neither the resolution scheme nor compulsory liquidation would be considered desirable.
Far more important than trying to pretend that banks are by some process of magic no longer contingent liabilities of their sovereigns because they quite clearly still are is redoubling efforts to ensure that banks are managed by executives with a keen risk sensitivity and good risk management information systems that can articulate a risk appetite properly, adjust risk assumed deftly and, where thats not possible, diversify it prudently.
The unfolding rise in interest rates will be a revealing test of this.
Meanwhile banks customers, investors in their debt and equity, and depositors should all be asking regulators and legislators in what ways the much more concentrated banking systems they have engineered are more resilient and what their plans are for encouraging competition.