?‘genes’, in the sense that certain features of corporate culture perform the same role as genes in biology, allowing information to be stored in the ‘organizational memory’ and passed on from individual to individual or from firm to firm when a new firm is created;
?the potential for spontaneous ‘mutation’, usually referred to in the economic world as innovation and primarily, though by no means always, technological;
?competition between individuals within a species for resources, with the outcomes in terms of longevity and proliferation determining which business practices persist;
?a mechanism for natural selection through the market allocation of capital and human resources and the possibility of death in cases of underperformance – that is, ‘diff erential survival’ ;
?scope for speciation, sustaining biodiversity through the creation of wholly new ‘species’ of financial institutions;
?scope for extinction, with certain species dying out altogether.
Sometimes, as in the natural world, the financial evolutionary process has been subject to big disruptions in the form of geopolitical shocks and financial crises. The difference is, of course, that whereas giant asteroids come from outer space, financial crises originate within the system. The Great Depression of the 1930s and the Great Inflation of the 1970s stand out as times of major discontinuity, with ‘mass extinctions’such as the bank panics of the 1930s and the Savings and Loans failures of the 1980s. A comparably large disruption has clearly happened in our time. But where are the mass extinctions? The dinosaurs still roam the financial world.
The answer is that, whereas evolution in biology takes place in a pitiless natural environment, evolution in finance occurs within a regulatory framework where – to adapt a phrase from antiDarwinian creationists.–‘intelligent design’ plays a part. But just how intelligent is this design? The answer is: not intelligent enough to secondguess the evolutionary process. In fact, stupid enough to make a fragile system even more fragile.
Think of it this way. The regulatory frameworks of the post1980 period encouraged many banks to increase their balance sheets relative to their capital. This happened in all kinds of different countries, in Germany and Spain as much as in the United States. (We really cannot blame Ronald Reagan for what happened in Berlin and Madrid.) When propertybacked assets fell in price, banks were threatened with insolvency. When shortterm funding dried up, they were threatened with illiquidity. The authorities found that they had to choose between a Great Depression scenario of massive bank failures or bailing the banks out. They bailed them out. Chastened by ungrateful voters (who still do not appreciate how much worse things could have got if the ‘too big’ had actually failed), the legislators now draw up statutes designed to avoid future bailouts.
Dodd – Frank states clearly that taxpayers will not pay a penny the next time a SIFI goes bust. It is rather less clear about who will pay. Section 214 is (mercifully) unambiguous: ‘All funds expended in the liquidation of a financial company under this title shall be recovered from the disposition of assets of such financial company, or shall be the responsibility of the financial sector, through as -sessments.’ So what about secured creditors, the bank bondholders whom so much was done to protect from loss in 2008~9 ? Prudently, Dodd–Frank commissions a study on that one. After all, if the net effect of the legislation really is to rule out any public funding for a seriously bankrupt SIFI, it is hard to see how those bondholders can avoid a sizeable loss. But if that is the case, then the cost of capital for big banks must rise, even as their return on equity is going down. You wanted to reduce instability, but all you did was increase fragility.
Another and related way of thinking about the financial system is as a highly complex system, made up of a very large number of interacting components that are asymmetrically organized in a network.21 This network operates somewhere between order and disorder – on ‘the edge of chaos’. Such complex systems can appear to operate quite smoothly for some time, apparently in equilibrium, in reality constantly adapting as positive feedback loops operate. But there comes a moment when they ‘go critical’. A slight perturbation can set off a ‘phase transition’ from a benign equilibrium to a crisis. This is especially common where the network nodes are ‘tightly coupled’. When the interrelatedness of a network increases, conflicting constraints can quickly produce a ‘complexity catastrophe’.
All complex systems in the natural world – from termite hills to large forests to the human nervous system – share certain characteristics. A small input to such a system can produce huge, unanticipated changes. Causal relationships are often nonlinear. Indeed, some theorists would go so far as to say that certain complex systems are wholly nondeterministic, meaning that it is next to impossible to make predictions about their future behaviour based on past data. Will the next forest fire be tiny or huge, a bonfire or a conflagration? We can’ t be sure. The same ‘power law’ relationship seems to be apply to earthquakes and epidemics.22
It turns out that financial crises are much the same. And this shouldn’ t surprise us. As heterodox economists like W. Brian Arthur have been arguing for years, a complex economy is characterized by the interaction of dispersed agents, a lack of any central control, multiple levels of organization, continual adaptation, incessant creation of new market niches and no general equilibrium. Viewed in this light, as Andrew Haldane of the Bank of England has argued, Wall Street and the City of London are parts of one of the most complex systems that human beings have ever made (see Figure 2.1).23 And the combination of concentration, interbank lending, financial innovation and technological acceleration makes it a system especially prone to crash. Once again, however, the diff erence between the natural world and the financial world is the role of regulation. Regulation is supposed to reduce the number and size of financial forest fires. And yet, as we have seen, it can quite easily have the opposite effect. This is because the political process is itself somewhat complex. Regulatory bodies can be captured by those whom they are supposed to be regulating, not least by the prospect of wellpaid jobs should the gamekeeper turn poacher.
Figure 2.1 Network connectivity balloons for the international financial system
Source: Andrew Haldane, Bank of England (see note 23 for full reference).
They can also be captured in other ways – for example, by their reliance on the entities they regulate for the very data they need to do their work.
In his book Antifragile, the statistician and options trader turned philosopher Nassim Taleb asks a wonderful question: what is the opposite of fragile? The answer is not ‘robust’ or ‘strong’, because those words simply mean less fragile. The true opposite of fragile is ‘antifragile’. A system that becomes stronger when subjected to perturbation is anti-fragile.24 The point is that regulation should be designed to heighten antifragility. But the regulation we are contemplating today does the opposite: because of its very complexity – and often selfcontradictory objectives – it is profragile.
Lessons from Lombard Street
Overcomplicated regulation can indeed be the disease of which it purports to be the cure. Just as the planners of the old Soviet system could never hope to direct a modern economy in all its complexity, for reasons long ago explained by Friedrich Hayek and Janos Kornai,25 so the regulators of the postcrisis world are doomed to fail in their eff orts to make the global financial system crisisfree. They can never know enough to manage such a complex system.
They will only ever learn from the last crisis how to make the next one.
Is there an alternative? I believe there is. But I believe we need to go back to the time of Darwin to find it. In Lombard Street, published in 1873, Walter Bagehot described with great skill the way in which the City of London had evolved in his time. Bagehot understood that, for all its Darwinian vigour, the British financial system was complex and fragile. ‘In exact proportion to the power of this system’, he observed, ‘is its delicacy.– I should hardly say too much if I said its danger..fieven at the last instant of prosperity, the whole structure is delicate. The peculiar essence of our financial system is an unprecedented trust between man and man; and when that trust is much weakened by hidden causes, a small accident may greatly hurt it, and a great accident for a moment may almost destroy it.’ 26
No one has ever given a better description of how a bank run happens than Bagehot; those unfamiliar with Lombard Street had to find out for themselves in 2007, at the time of the runs on Northern Rock and Countrywide, and again in 2012, when it was the turn of the Spanish Bankia to lose the confidence of its depositors. One of the great beauties of Lombard Street is the way it surveys all the key institutions of the London money market – the ascendant jointstock banks, the waning private banks, the bill brokers, the new savings banks – and exposes the weakness in the position of each. In theory, Bagehot would have preferred a system in which each institution had to look to itself by maintaining a reserve against contingencies. But in practice the London market had evolved in such a way that there was only one ultimate reserve for the entire City and that was the Bank of England’s: ‘the sole considerable unoccupied mass of cash in the country’.27 As in our time, in other words, the central bank (and, behind it, the government that called it into being) constituted the last line of resistance in time of panic.
By reviewing half a century of financial crises, Bagehot brilliantly showed how the Bank of England’s role as custodian of the nation’s cash reserve was quite different from its role as defined by statute or, indeed, as understood by the men running it. In the 1825 panic the Bank had done the right thing, but much too late in the day, and without knowing quite why it was the right thing. In each of the three panics that followed the passage of the Bank Charter Act of 1844 – a piece of legislation which was largely concerned with the Bank’s noteissuing function – the Act had been suspended. There was, as in our time, uncertainty about which securities it would accept as collateral in a crisis. The Bank’s governance structure was opaque. Its governor and directors were themselves not bankers. (In those days they chose merchants; nowadays we prefer academics – which not everyone would regard as an improvement.) They barely coped when a SIFI called Overend Gurney blew up in 1866.
Bagehot’s remedies were clearcut, though I believe they are very often misinterpreted. The famous recommendation was that in a crisis the central bank should lend freely at a penalty rate: ‘Very large loans at very high rates are the best remedy ...’ 28 Nowadays we follow only the first half of his advice, in the belief that our system is so leveraged that high rates would kill it. Bagehot’s rationale was to ‘prevent the greatest number of applications by persons who do not require it’.29 Watching all banks, strong and weak alike, gorge themselves on today’s seemingly limitless supply of loans at nearzero rates, I see what he meant.
We also neglect the rest of what Bagehot said, and in particular the emphasis he laid on discretion as opposed to set rules. In the first place, he stressed the importance of having Bank directors with considerable market ex -perience. ‘Steady merchants’, he wrote, ‘always know the questionable standing of dangerous persons; they are quick to note the smallest signs of corrupt transactions.’ Executive power should be conferred on a new, fulltime deputy governor acting as a kind of permanent undersecretary. And the advisory Court should be selected so as ‘to introduce ... a wise apprehensiveness’.30
Secondly, Bagehot repeatedly stressed, as he put it, ‘the cardinal importance of [the Bank of England’s] always retaining a great banking reserve’. But he was emphatic that the size of the reserve should not be specified by some automatic rule, the way the banknote circulation was under the 1844 Bank Charter Act: ‘No certain or fixed proportion of its liabilities can in the present times be laid down as that which the Bank ought to keep in reserve.’ The ideal central bank would target nothing more precise than an ‘apprehension minimum’, which ‘no abstract argument, and no mathematical computation will teach to us’ :
And we cannot expect that they should [he went on]. Credit is an opinion generated by circumstances and varying with those circumstances. The state of credit ... can only be known by trial and inquiry. And in the same way, nothing can tell us what amount of ‘reserve’ will create a diff used confidence; on such a subject there is no way of arriving at a just conclusion except by incessantly watching the public mind, and seeing at each juncture how it is affected.31
Nor should there be predictability in the Bank’s discount rate, the rate at which it lent against goodquality commercial paper. The rule ‘that the Bank of England should look to the market rate and make its own rate conform to that ... was ... always erroneous’, according to Bagehot. The ‘first duty’ of the Bank was to use the discount rate to ‘protect the ultimate cash of the country’.32 This too of course implied a discretionary power, since the desirable size of the reserve was not specified by any rule.
There are some today, like Larry Kotlikoff and John Kay, who see the only salvation in a rootandbranch structural reform of our financial system: ‘narrow banking’ of some sort, if not the replacement of banks altogether.33 I can see the intellectual appeal of such arguments. In theory, perhaps it would be much better if big banks were chopped up, leverage ratios were drastically reduced and the interconnections between deposittakers and risktakers were reduced.34 But, like Bagehot, I take the world as I find it, and I do not expect to see in my lifetime a wholesale abandonment of the current model of ‘too big to fail’ institutions backstopped by the central bank and, if necessary, by the public purse. Our task, like Bagehot’s, is ‘to make the best of our banking system, and to work it in the best way that it is capable of. We can only use palliatives, and the point is to get the best palliative we can.’ 35
How to Encourage Bankers
‘The problem is delicate,’ as Bagehot candidly concluded his great work, and ‘the solution is varying and difficult.’ 36 It remains so today. But I believe a return to Bagehot’s first principles would not be a bad starting point. First, strengthen the central bank as the ultimate authority in both the monetary and supervisory systems. Second, ensure that those in charge at the central bank are ‘apprehensive’ as well as experienced, so that they will act when they see excessive credit growth and assetprice inflation. Third, give them considerable latitude in their use of the principal central banking tools of reserve requirements, interest rate changes and openmarket securities purchases and sales. Fourth, teach them some financial history, as Bagehot taught his readers.
