The Global Financial Crisis and Behavioural Economics*
The author thanks John Freebairn, Joe Isaac, Elias Khalil, Carsten Murawski, Richard Pomfret and Ross Williams for comments on an earlier draft.
Ian M. McDonald, University of Melbourne, Victoria 3010, Australia. Email: I.McDonald@Unimelb.edu.au
Conventional economics, which is based on Homo economicus, cannot provide a satisfactory explanation for the global financial crisis. However, behavioural economics, and the concept of present bias, self-serving bias, ‘new era’ stories, money illusion, comparisons with reference levels and herding, can provide an explanation.
The biggest surprise to economists from the global financial crisis is that many financial firms took actions that led them to insolvency. As Alan Greenspan expressed it, “Those of us who have looked to the self-interest of lending institutions to protect shareholders' equity (myself especially) are in a state of shocked disbelief.” The reason that this is such a surprise is the inconsistency of these actions with the actions that Homo economicus, that cool calculator of self-interest, would have taken. Homo economicus would not have been so incautious in protecting his wealth. This implies that the economic model based on Homo economicus needs serious reformulation.
Economics has been dominated in the past two decades by a free-market ideology, which includes a virtual opposition to market regulation. The free-market ideology has downplayed, almost to the point of rejection, the concerns of economists about the traditional market failures, that is, externalities, public goods, imperfect competition, asymmetric information with adverse selection and moral hazard. For example, economists have not led the environmental movement, even though economic analysis based on Homo economicus reveals that excessive pollution is a consequence of free markets. Consider the contrast with the movement to free trade, where economists have led. In environmental issues, economists have been followers rather than leaders, notwithstanding their important contribution when they did choose to follow. In macroeconomics, especially in academic circles, the free-market ideology has been even more dominant. A beneficial role for governments to use aggregate demand policy to influence the level of activity has been dismissed by the dominant school of thought, the rational expectations school.
The free market focus has led to considerable deregulation in financial markets since 1980. The move in Australia in 1983 to a flexible exchange rate was a socially beneficial result of this focus. The abolition of ceilings on interest rates offered by banks to their depositors was also in all probability socially beneficial. However, the removal of the distinction between banks and non-banks caused by the repeal in the United States of the Glass-Steagall Act is now questioned as a socially beneficial measure. Certainly some Australian commentators on financial markets feel that the Wallis Inquiry was wise in maintaining the bank/non-bank distinction, for example, Ian Harper in his lecture to the Economic Society of Australia, Victorian Branch, 31 March 2009.
The self-interest approach based on rational behaviour led to the new financial economics, from which developed the efficient market’s view – that share prices are unbiased estimates of true values. The new financial economics also led to a large burst of innovation in financial products. These new products have the socially beneficial characteristic of spreading risk and of dealing with asymmetric information in an efficient manner. They also made home ownership possible to a large number of people. However, it now seems that the efficient markets’ view is in tatters and that the new financial products brought huge social costs.
2. What Happened?
Jay Light of the Harvard Business School, see the Harvard Financial Markets Panel Discussion on 26 September 2008 at: http://www.advisorperspectives.com/pdfs/Differing_Perspectives_on_the_Credit_Crisis.pdf, described the cause of the global financial crisis as a collision between a collapsing housing bubble and a new but untested financial system. This collision had its origin in a huge increase in the issue of subprime mortgages, one of the financial innovations that evolved from the new financial economics. Subprime mortgages were introduced to improve the access by poor people, with minimal or non-existent credit profiles, to the housing market. The achievement of this desirable objective was enhanced by new quantitative methods for judging credit worthiness and by the technique of securitization, through which mortgages would be bundled together and sold to other financial institutions. In this process, described as the originate-and-distribute model, the originator of the mortgage is different from the final holder of the mortgage.
Whilst there was some “government push” behind the introduction of subprime mortgages, for the rapid expansion in the number of subprime mortgages to have occurred relied on the huge appetite of the financial sector to buy the securitized product. This rapid expansion occurred in parallel with a rapid rise in house prices. There was of course a two-way interaction between these phenomena. When house prices turned down, a number of mortgagees defaulted on their mortgages, returning the keys and titles of their houses to the originator of the mortgage. The widespread use of non-recourse loans, for which the borrower does not have personal liability, encouraged these defaults. Financial institutions now found a substantial decrease in the value of their assets. This decrease in value threatened their financial viability. This threat was increased by the very low capital-to-asset ratios that financial institutions had chosen. The precarious nature of many financial institutions led to a drying up of credit, including credit to non-financial firms. This led to decreases in the activity and employment of the credit-starved firms, which through multiplier effects spread the decrease in employment across the economy.2
The globalised nature of the world economy, with substantial interconnections between economies through trade and capital flows, caused the US subprime crisis to have global effects, dragging down the global economy. Because of a somewhat similar problem in the UK housing market, a UK financial crisis had similar effects on the UK economy, which then spread in a similar fashion to the global economy. The toll on the global economy is immense. For example, from September 2007 to October 2008 world stock market capitalisation decreased by about USD26,400 billion; see Blanchard (2009).
3. Conventional Economics and the Subprime Crisis
Conventional economics, that is, the model based on Homo economicus, can only offer an incomplete explanation of the subprime crisis.
The decisions by many to default on mortgages were rational and thus can be explained by Homo economicus. For those mortgagees for whom the value of the outstanding debt on their mortgage exceeds the value of their house, there is a clear incentive to default. Given that in the United States the penalty for many subprime mortgagees of defaulting is slight, because they are non-recourse loans, defaulting becomes a rational choice. However, this does not explain why originators did not take this into account when making the mortgage.
The originators of the subprime mortgage had less incentive to be cautious in making mortgages because they planned to sell off the mortgage as part of a securitized asset. A default would fall mainly on the holder of the securitized asset. However, this does not explain why the purchaser of the securitized asset failed to take into account the high default risk of the subprime mortgages.
The purchases of the securitized assets were typically made by employees of financial institutions such as merchant banks. These employees, it appears, were on contracts that encouraged volume and a short-term outlook. Thus, it appears rational for these employees to take little account of the riskiness of the assets they were responsible for purchasing. However, this does not explain why the managers of the financial institutions did not write contracts for their employees that encouraged actions that would protect the firm’s value and why the owners of the firms did not insist that better contracts be written.
At this point explanations based on Homo economicus are hard to come by. Robert Merton, see the Harvard Financial Markets Panel cited before, has argued that the management and boards of directors of financial institutions were out of their depth, unable to understand the new products created by the new financial economics. But they did not recognise this deficiency or if they did chose to shroud it in a cloud of optimism. Coolness in calculation should have led them to face up to their ignorance. Failing that, coolness in calculation by the owners, those whose wealth was at risk, should have forced management to correct its optimistic view and face up to its deficiencies.
Widespread financial failures are suggestive of more than random errors in decision making and appear inconsistent with the actions of Homo economicus. It was the tendency of owners of financial institutions to show such little care over their wealth that put Alan Greenspan into “a state of shocked disbelief.”
4. Behavioural Economics and the Subprime Crisis
Behavioural economics, drawing on a wide range of evidence that includes psychology and results from experiments, finds significant deviations from Homo economicus in the behaviour of human beings. It appears that humans are plagued by present bias, self-serving bias, seeing illusory patterns, reference standards, money illusion and herding. These weaknesses can be seen in the subprime crisis and offer an explanation for that crisis which Homo economicus lacks.
At each of the steps in the road to financial failure described before, the deviations in human behaviour from the behaviour of Homo economicus revealed by behavioural economics hastened the journey to catastrophe. Applying behavioural economics to the subprime crisis reveals the following insights.
1. Subprime mortgages exploit present bias. According to present bias, also called hyperbolic discounting, people tend to focus on the present, that is on “now,” and undervalue the future. The negative effect of this “now-focus” on the present is that people will make decisions that they subsequently regret. This is because “now” moves with time. For example, people may put off an unpleasant decision, planning to do it in the future, but, when the future comes, they change their minds and put off the unpleasant decision again. The change of mind will cause a regret of the previous decision. Thus, an individual may put off reading the small print in the subprime mortgage contract. Furthermore, at the time of taking out a subprime mortgage an individual may feel prepared to accept cutbacks in consumption in the future, when the period with the low teaser rate ends, but find later, when the time comes, that this is too difficult. This behaviour resulting from now-focussed decisions is described as time-inconsistent. For the subprime mortgage industry it makes lending easier and is a force tending to increase the volume of risky assets.
2. Experimental work by behavioural economists has revealed a strong tendency for experimental subjects in laboratory asset-trading games to bid the price of an asset above its fundamental price. Even though experimental subjects know the fundamental price of the asset, that is, the price the asset will be worth to the holder at the end of the experiment, in the interim they are prepared to bid up the price of the asset above the fundamental price. However, eventually the price crashes. This is the pattern of a bubble. That intelligent subjects will create an asset price bubble in a controlled environment, where they know with certainty the fundamental price of the asset, and thus the potential losses are clear, makes the occurrence of actual bubbles in the real world stock and property markets less surprising. This bubble behaviour is repeated when the same experimental subjects start again in a second, third, etc. asset-trading game.
It seems probable that people in bidding up the price of an asset above its fundamental value believe that they can get out before the crash. Such a belief is consistent with the idea of self-serving bias, that is, a tendency to overrate one’s abilities. Self-serving bias seems to be a fact of human nature. For example, 90 per cent of drivers think they have above-average driving skills. In the context of a bubble, if people believe that they have the ability to get out before the crash this means that they rate their ability above that of the average asset holder.
3. Based on the new financial economics, mathematical models have been developed to calculate the risk of assets. In applying these models, it appears that the risk of a fall in house prices was ignored; see Eichengreen (2008). Given the enormity of the consequences of falling house prices, and given that not too long ago, in the 1980s, house prices fell, it seems extraordinary that financial experts ignored this risk. An explanation from behavioural economics, put forward by Akerlof and Shiller (2009), is a tendency for people in a booming market to believe in “new era” stories. Thus, in the 2000s according to the story repeated widely in the media, a new era had arrived as a result of advances in information technology and to the experience of a long period of sustained economic growth and sustained increases in asset prices. In this new era, recessions and asset price falls were a thing of the past. To disagree with the new era story would attract the pejorative label “dinosaur.” Akerlof and Shiller find that new era stories are a general feature of booms.
4. Akerlof and Shiller (2009) also argue that people suffer from money illusion, unlike Homo economicus. An implication of this, they argue, is an overestimation of the real value of an increase in house prices. To calculate the increase in one’s house price, the natural comparison for people to make is to compare the current value with that when they purchased it. For people who have owned the house for a number of years, the increase in value would appear larger. Furthermore, in a rising market, this effect would be increased. Overestimation of assessments of increases in wealth would lead to people borrowing and consuming excessively. This phenomenon, Akerlof and Shiller argue, can explain to some extent the large decline in household saving in the United States in recent history.
5. When people base performance on comparisons with reference standards, such as comparing the rate of return from an asset with a benchmark reference rate then, as a result of loss aversion, people may take excessive risks. This tendency is supported by a well-documented phenomenon in betting is for punters to bet on outsiders, which are very risky, in the last race of the day. The behavioural explanation for this is that punters are trying to avoid a net loss on their days betting, that is, to avoid a net outcome for the day that is below the reference level of zero. The reference level of zero reflects their concern with how well they have done on the day’s racing. Because of loss aversion, falling below the reference level of zero generates an extra dollop of disutility, large enough for the punter to incur extra risk to try to avoid this loss. Turning back to behaviour in financial markets, it seems that in recent years market returns have been low and that this has encouraged a search for yield; see Corden (2008). A search for yield implies a reference rate for yield above average market rates for low-risk assets, thereby introducing a tendency to take on an increased risk to avoid shortfalls of the actual rate received from the reference rate.
6. Herding probably magnifies the effects described before. A tendency for people to go with the crowd, well documented in psychological experiments, will increase the tendency to sign up to complicated subprime mortgages which are not understood because others are signing up, to stay in a rising market because other people are staying in the market, to ignore the risk of falls in house prices because others are ignoring this risk, to overvalue the real value of the rise in the price of one’s house because others are doing this, to borrow and consume by large amounts because others are also borrowing and consuming by large amounts and to take on excessive risk in a search for yield stimulated by fear of the loss aversion from falling below an arbitrary reference standard because others are taking on increased risk.
5. Policy Implications
In as far as people’s behaviour deviates from the behaviour of Homo economicus and is influenced by the behavioural factors discussed before, there is a case for protecting people against themselves. This argument for regulation differs from the arguments of conventional economics based on Homo economicus. In conventional economics people, behaving as Homo economicus, are assumed to act in their best self-interest and thus their valuations are reasonably assumed to be the best for themselves. However, the behavioural factors imply that people are not making the valuations that are in their best interests. People need to be protected from their own decisions. They need to be induced to take a longer view, a more detached view of their own abilities, to be more suspicious of new era stories, to convert nominal changes into real changes and to avoid reference standards.
But, as an important strand in conventional economics argues, one may be reasonably suspicious of the ability of government to make better judgements on behalf of people. Governments are unaware of many details of people’s lives, details that whilst important to a particular individual are idiosyncratic. Can the bureaucrat know better? Furthermore, governments are subject to biases that may cause them to underweight the individual’s true interests. For example, as Stigler (1975) argued, there are economic forces that cause regulatory agencies to get captured by the industry they are supposed to regulate. And probably governments and their bureaucrats lack the incentive to place a high enough weight on the interests of individuals. As a solution to the problems of government ignorance, bias and lack of feeling, behavioural economists have suggested libertarian paternalism as a somewhat non-intrusive way through which government may improve individual’s decision-making and offset some of the behavioural biases from which individuals suffer.
The idea of libertarian paternalism is to set a default which would protect people subject to behavioural bias. For example, Shiller (2009) has proposed continuous work out mortgages as the default. In these mortgages, the outstanding balance is adjusted in line with an index of house prices in the neighbourhood. This contract would share the risks of changes in house prices between the borrower and lender, an improvement on the traditional arrangement where the house owner bore all this risk and also an improvement on subprime mortgages, where the risk-sharing details were too complex to be understood by many borrowers. Evidence shows that there is a tendency for people to choose the default. Thus, if the default is tailored for those with little ability or little motivation to understand a more complicated contract, then such people are more likely to be protected.
Just how irrational exuberance, Shiller’s term for the optimistic sentiment in a boom market, see Shiller (2005), should be controlled is hard to say, but one approach to dealing with this behavioural phenomenon is for regulation to respond automatically to rising markets. For example, increasing capital adequacy requirements and increasing liquidity requirements could be linked to statistical measures of rising markets. Michael Brennan has suggested the introduction of “S&P 500 Strips” into financial markets; see Shiller (2005, pp. 227–8). These assets would establish markets for the dividends of S&P 500 companies for particular years in the future, thereby focussing attention on future prospects. This would increase the salience of future outcomes, and push people to take a longer-term view. For example, excessive optimism in the short-term would be revealed by considering the market’s judgement of longer-term prospects.
6. The Impending Danger – Throwing out the Baby with the Bathwater
The global financial crisis has revealed the inability of unregulated markets to correct themselves. The remedy for this is regulation of markets. However, that regulation has to be framed with care. Excessive regulation can stifle market behaviour. History abounds with the failures of economic planning, such as the Soviet Union and Mao’s China.
Keeping a balance in the regulation of markets is a big ask for democracies. One problem is that extreme views, such as no government regulation, or the other extreme, that all important market outcomes should be determined by the government, are easier for people to understand than are balanced views. The appropriate degree of regulation weighs up the costs and benefits of regulation, seeking to maximise consumer and producer surplus. It uses instruments such as a carbon tax, rather than quantitative restrictions on carbon emissions.
Simple demands, such as “trees are sacred,” are more effective in motivating widespread support for environmental protection than are the complicated measures suggested by balancing costs and benefits. It is hard to imagine demonstrators chanting “what do we want?”–“consumer surplus;”“when do we want it?”–“optimally spread throughout our lifetimes.”
Various weaknesses in human behaviour revealed by behavioural economics raise concerns about the ability of the public to reach sensible decisions. Of particular concern is the propensity to see patterns where patterns do not exist. This weakness makes humans susceptible to the propaganda of interest groups. But even the experts are not immune. The global financial crisis shows that economists were susceptible to the idea of a self-correcting free market.
For managing one’s financial assets, behavioural economics provides some valuable guidelines. First, beware of present bias. Do not let the immediacy of the current situation cause you to undervalue the future. Consider the future consequences carefully. Take the long view in determining your level of saving. Second, beware of self-serving bias. Do not ascribe financial gains to your own skill when they are simply the result of a rising market. Beware of thinking you are smart enough to get out of a bubble before other people. Third, be sceptical of new era stories, especially when they are peddled by people wishing to sell you something. Fourth, do not allow money illusion to bolster your view of how smart an investor you are. Fifth, do not allow reference standards to push you into buying risky assets. Sixth, beware of following the crowd.
There are also valuable lessons from economics based on Homo economicus, such as do not think you can beat the market and do not think that higher return does not come without higher risk. If people had followed these maxims, then perhaps the global financial crisis would not have happened. However, people’s judgements were clouded, clouded it seems because they did not follow the lessons from behavioural economics.
Akerlof and Shiller (2009) have argued for adding a financial multiplier to the Keynesian aggregate demand multiplier to better capture the behaviour of aggregates in a fluctuating economy.