In Pursuit of the Unknown (59 page)

Poovey's fifth example was derivatives, and it was the most important of them all because the sums of money involved were so gigantic. Her analysis largely reinforces what I've already said. Her main conclusion was: ‘Futures and derivatives trading depends upon the belief that the stock market behaves in a statistically predictable way, in other words, that mathematical equations accurately describe the market.' But she noted that the evidence points in a totally different direction: somewhere between 75% and 90% of all futures traders lose money in any year.

Two types of derivative were particularly implicated in creating the toxic financial markets of the early twenty-first century: credit default
swaps and collateralised debt obligations. A credit default swap is a form of insurance: pay your premium and you collect from an insurance company if someone defaults on a debt. But anyone could take out such insurance on anything. They didn't have to be the company that owed, or was owed, the debt. So a hedge fund could, in effect, bet that a bank's customers were going to default on their mortgage payments – and if they did, the hedge fund would make a bundle, even though it was not a party to the mortgage agreements. This provided an incentive for speculators to influence market conditions to make defaults more likely. A collateralised debt obligation is based on a collection (portfolio) of assets. These might be tangible, such as mortgages secured against real property, or they might be derivatives, or they might be a mixture of both. The owner of the assets sells investors the right to a share of the profits from those assets. The investor can play it safe, and get first call on the profits, but this costs them more. Or they can take a risk, pay less, and be lower down the pecking order for payment.

Both types of derivative were traded by banks, hedge funds, and other speculators. They were priced using descendants of the Black–Scholes equation, so they were considered to be assets in their own right. Banks borrowed money from other banks, so that they could lend it to people who wanted mortgages; they secured these loans with real property and fancy derivatives. Soon everyone was lending huge sums of money to everyone else, much of it secured on financial derivatives. Hedge funds and other speculators were trying to make money by spotting potential disasters and betting that they would happen. The value of the derivatives concerned, and of real assets such as property, was often calculated on a mark to market basis, which is open to abuse because it uses artificial accounting procedures and risky subsidiary companies to represent estimated future profit as actual present-day profit. Nearly everyone in the business assessed how risky the derivatives were using the same method, known as ‘value at risk'. This calculates the probability that the investment might make a loss that exceeds some specified threshold. For example, investors might be willing to accept a loss of a million dollars if its probability were less than 5%, but not if it were more likely. Like Black–Scholes, value at risk assumes that there are no fat tails. Perhaps the worst feature was that the entire financial sector was estimating its risks using exactly the same method. If the method were at fault, this would create a shared delusion that the risk was low when in reality it was much higher.

It was a train crash waiting to happen, a cartoon character who had walked a mile off the edge of the cliff and remained suspended in mid-air
only because he flatly refused to take a look at what was under his feet. As Poovey and others like her had repeatedly warned, the models used to value the financial products and estimate their risks incorporated simplifying assumptions that did not accurately represent real markets and the dangers inherent in them. Players in the financial markets ignored these warnings. Six years later, we all found out why this was a mistake.

Perhaps there is a better way.

The Black–Scholes equation changed the world by creating a booming quadrillion-dollar industry; its generalisations, used unintelligently by a small coterie of bankers, changed the world again by contributing to a multitrillion-dollar financial crash whose ever more malign effects, now extending to entire national economics, are still being felt worldwide. The equation belongs to the realm of classical continuum mathematics, having its roots in the partial differential equations of mathematical physics. This is a realm in which quantities are infinitely divisible, time flows continuously, and variables change smoothly. The technique works for mathematical physics, but it seems less appropriate to the world of finance, in which money comes in discrete packets, trades occur one at a time (albeit very fast), and many variables can jump erratically.

The Black–Scholes equation is also based on the traditional assumptions of classical mathematical economics: perfect information, perfect rationality, market equilibrium, the law of supply and demand. The subject has been taught for decades as if these things are axiomatic, and many trained economists have never questioned them. Yet they lack convincing empirical support. On the few occasions when anyone does experiments to observe how people make financial decisions, the classical scenarios usually fail. It's as though astronomers had spent the last hundred years calculating how planets move, based on what they thought was reasonable, without actually taking a look to see what they really did.

It's not that classical economics is completely wrong. But it's wrong more often that its proponents claim, and when it does go wrong, it goes very wrong indeed. So physicists, mathematicians, and economists are looking for better models. At the forefront of these efforts are models based on complexity science, a new branch of mathematics that replaces classical continuum thinking by an explicit collection of individual agents, interacting according to specified rules.

A classical model of the movement of the price of some commodity, for example, assumes that at any instant there is a single ‘fair' price, which in
principle is known to everyone, and that prospective purchasers compare this price with a utility function (how useful the commodity is to them) and buy it if its utility outweighs its cost. A complex system model is very different. It might involve, say, ten thousand agents, each with its own view of what the commodity is worth and how desirable it is. Some agents would know more than others, some would have more accurate information than others; many would belong to small networks that traded information (accurate or not) as well as money and goods.

A number of interesting features have emerged from such models. One is the role of the herd instinct. Market traders tend to copy other market traders. If they don't, and it turns out that the others are on to a good thing, their bosses will be unhappy. On the other hand, if they follow the herd and everyone's got it wrong, they have a good excuse: it's what everyone else was doing. Black–Scholes was perfect for the herd instinct. In fact, virtually every financial crisis in the last century has been pushed over the edge by the herd instinct. Instead of some banks investing in property and others in manufacturing, say, they
all
rush into property. This overloads the market, with too much money seeking too little property, and the whole thing comes to bits. So now they all rush into loans to Brazil, or to Russia, or back into a newly revived property market, or lose their collective marbles over dotcom companies – three kids in a room with a computer and a modem being valued at ten times the worth of a major manufacturer with a real product, real customers, and real factories and offices. When that goes belly-up, they
all
rush into the subprime mortgage market. . .

That's not hypothetical. Even as the repercussions of the global banking crisis reverberate through ordinary people's lives, and national economies flounder, there are signs that no lessons have been learned. A rerun of the dotcom fad is in progress, now aimed at social networking websites: Facebook has been valued at $100 billion, and Twitter (the website where celebrities send 140-character ‘tweets' to their devoted followers) has been valued at $8 billion despite never having made a profit. The International Monetary Fund has also issued a strong warning about exchange traded funds (ETFs), a very successful way to invest in commodities like oil, gold, or wheat without actually buying any. All of these have gone up in price very rapidly, providing big profits for pension funds and other large investors, but the IMF has warned that these investment vehicles have ‘all the hallmarks of a bubble waiting to burst. . . reminiscent of what happened in the securitisation market before the crisis'. ETFs are very like the derivatives that triggered the credit crunch,
but secured in commodities rather than property. The stampede into ETFs has driven commodity prices through the roof, inflating them out of all proportion to the real demand. Many people in the third world are now unable to afford staple foodstuffs because speculators in developed countries are taking big gambles on wheat. The ousting of Hosni Mubarak in Egypt was to some extent triggered by huge increases in the price of bread.

The main danger is that ETFs are starting to be repackaged into further derivatives, like the collateralised debt obligations and credit default swaps that burst the subprime mortgage bubble. If the commodities bubble bursts, we could see a rerun of the collapse: just delete ‘property' and insert ‘commodities'. Commodity prices are very volatile, so ETFs are high-risk investments – not a great choice for a pension fund. So once again investors are being encouraged to take ever more complex, and ever more risky, bets, using money they don't have to buy stakes in things they don't want and can't use, in pursuit of speculative profits – while the people who do want those things can no longer afford them.

Remember the Dojima rice exchange?

Economics is not the only area to discover that its prized traditional theories no longer work in an increasingly complex world, where the old rules no longer apply. Another is ecology, the study of natural systems such as forests or coral reefs. In fact, economics and ecology are uncannily similar in many respects. Some of the resemblance is illusory: historically each has often used the other to justify its models, instead of comparing the models with the real world. But some is real: the interactions between large numbers of organisms are very like those between large numbers of stock market traders.

This resemblance can be used as an analogy, in which case it is dangerous because analogies often break down. Or it can be used as a source of inspiration, borrowing modelling techniques from ecology and applying them in suitably modified form to economics. In January 2011, in the journal
Nature
, Andrew Haldane and Robert May outlined some possibilities.
⁴
Their arguments reinforce several of the messages earlier in this chapter, and suggest ways of improving the stability of financial systems.

Haldane and May looked at an aspect of the financial crisis that I've not yet mentioned: how derivatives affect the stability of the financial system. They compare the prevailing view of orthodox economists, which
maintain that the market automatically seeks a stable equilibrium, with a similar view in 1960s ecology, that the ‘balance of nature' tends to keep ecosystems stable. Indeed, at that time many ecologists thought that any sufficiently complex ecosystem would be stable in this way, and that unstable behaviour, such as sustained oscillations, implied that the system was insufficiently complex. We saw in
Chapter 16
that this is wrong. In fact, current understanding indicates exactly the opposite. Suppose that a large number of species interact in an ecosystem. As the network of ecological interactions becomes more complex through the addition of new links between species, or the interactions become stronger, there is a sharp threshold beyond which the ecosystem ceases to be stable. (Here chaos counts as stability; fluctuations can occur provided they remain within specific limits.) This discovery led ecologists to look for special types of interaction network, unusually conducive to stability.

Might it be possible to transfer these ecological discoveries to global finance? There are close analogies, with food or energy in an ecology corresponding to money in a financial system. Haldane and May were aware that this analogy should not be used directly, remarking: ‘In financial ecosystems, evolutionary forces have often been survival of the fattest rather than the fittest.' They decided to construct financial models not by mimicking ecological models, but by exploiting the general modelling principles that had led to a better understanding of ecosystems.

They developed several economic models, showing in each case that under suitable circumstances, the economic system would become unstable. Ecologists deal with an unstable ecosystem by managing it in a way that creates stability. Epidemiologists do the same with a disease epidemic; this is why, for example, the British government developed a policy of controlling the 2001 foot-and-mouth epidemic by rapidly slaughtering cattle on farms near any that proved positive for the disease, and stopping all movement of cattle around the country. So government regulators' answer to an unstable financial system should be to take action to stabilise it. To some extent they are now doing this, after an initial panic in which they threw huge amounts of taxpayers' money at the banks but omitted to impose any conditions beyond vague promises, which have not been kept.

However, the new regulations largely fail to address the real problem, which is the poor design of the financial system itself. The facility to transfer billions at the click of a mouse may allow ever-quicker profits, but it also lets shocks propagate faster, and encourages increasing complexity. Both of these are destabilising. The failure to tax financial transactions
allows traders to exploit this increased speed by making bigger bets on the market, at a faster rate. This also tends to create instability. Engineers know that the way to get a rapid response is to use an unstable system: stability by definition indicates an innate resistance to change, whereas a quick response requires the opposite. So the quest for ever greater profits has caused an ever more unstable financial system to evolve.