Causes of the Global Financial Crisis – John Edwards

Issue 1 | 28 January 2009

On Wednesday 28 January 2009, The Sydney Institute held an early evening seminar to discuss the causes of the Global Financial Crisis then unfolding. The speakers were Mark Johnson, a former Chairman of Macquarie Bank, Dr John Edwards, Chief Economist with HSBC and Dr Timo Henckel, a Research Fellow with the Centre for Applied Macroeconomic Analysis at the Australian National University. The function attracted a capacity audience and The Sydney Institute announced further discussion seminars on the GFC would be organised throughout 2009.

CAUSES OF THE GLOBAL FINANCIAL CRISIS

JOHN EDWARDS

This evening we are talking about the causes of the global financial crisis, which blew up in the middle of 2007 with a rising rate of delinquency in US sub prime mortgage loans, dragged on for a little over a year, and then abruptly became very much worse after the bankruptcy of Lehman in September 2008. In stating the subject in this way, I suppose I have already hinted at my take on the crisis – as will become clear.

We are asked this evening specifically to talk about the causes of the crisis. We all know the most important aspect of the crisis is its consequences – the protracted slump in the US, Europe, the UK and now Japan, East Asia, Eastern Europe, and of course Australia. But we are not here tonight to talk about the consequences.

My ambition tonight is not to describe the evolution of the global economy over the last few decades, but to attempt to identify the absolutely necessary and sufficient causes of the global financial crisis itself – no less and no more than we need to explain what happened. There have been plenty of extremely useful papers published on the global financial crisis. I won’t cite them all but I should acknowledge the importance of three papers by American economists – Markus Brunnermeier, Charles Calomiris, and Gary Gorton – in influencing my thinking about the crisis.

A useful preliminary is to recognise that while all financial crises are different, they are also all the same. We have seen plenty of them in recent years. The Mexico Peso Crisis of 1994, the Asian Financial Crisis of 1997, the Russian Debt Crisis and the Long Term Capital Management Crisis of 1998, the Tech Wreck of 2001 are a few that come to mind, and there have been many, many others. They all involve what is then or later thought to be excessive risk taking. They mostly involve losses on debt (though not always – the tech wreck for example was primarily about businesses investment and share prices).The crisis usually begins with some relatively minor event. It is propagated by the sudden realisation that business and/or household debt must be reduced, and that bank lending must be reduced because losses have absorbed too much capital to support the remaining stock of assets. This reduction in leverage is sought through the sale of financial and physical assets, causing their prices to tumble. Wealth declines. Falling asset prices, a diminished appetite for risk and the restriction on lending reduce employment and output. In these respects financial crises are similar. But they are also all different – in the originating incident, in the trajectory of the crisis, and in the scale of the consequences. Accordingly, they are different in the kind of remedies they suggest to minimise a recurrence, and to minimise the consequences.

One important distinction in this particular crisis is that it arose in the United States, the world’s biggest and richest economy, rather than in the developing world. Another is that it was rapidly transmitted to other advanced economies, notably the UK and Europe, through the exposure of international banks to large losses in the US. A final distinction – and this is a controversial point, but the heart of my argument – is that it arose in the financial system itself. This is not unusual but it is important. The crisis has immense economic consequences, and one may well argue that the economic circumstances in which it occurred have amplified the economic consequences. But in my view it did not have what I would call macroeconomic causes. It had the usual causes of a strictly financial crisis – ignorance of or refusal to correctly assess risk, high leverage in financial institutions, and so forth. This is what I now wish to argue.

We are all agreed, I think, that the global financial crisis began at the end of 2006 and the first half of 2007 with rising delinquencies in housing loans to US sub prime borrowers.

There are a few key characteristics of this market which explain what happened. First of all, sub prime borrowers are those who, by definition, have experienced difficulty in the past in repaying loans. They have poor credit histories. The loan was typically at a low introductory rate, with a reset period. It was typically at a high loan to valuation ratio – after all, the borrower had little money. It would automatically reset, often to a punitive rate. The assumption was that the equity would increase with rising house prices, and the loan would be refinanced with a lower LVR at a lower long term rate – usually long before the reset period.

You also need to know that in roughly half of US states home mortgages are what we might call non-recourse or “Walkaway states”. The liability of the borrower was limited to the property – it could not or in practice did not include the other assets of the borrower, as it does in Australia.

It should be immediately obvious that given these circumstances recently written sub prime mortgages had to be in trouble once house prices stopped rising, which is what happened in 2006.

These sub prime delinquencies were large in absolute terms but relatively small in the overall US debt market. Though sub prime mortgages accounted for around half of new mortgage issuance in 2005 and 2006, the overall total was never more than 15 per cent of all US mortgages. Only a small proportion was ever delinquent. Most sub prime mortgages were being satisfactorily serviced by the borrower, and are still today. With an expected recovery rate of 50 per cent from the sale of the foreclosed property delinquent sub prime would in total cost lenders around $600 billion. This is much less than the amount of additional capital raised by the US banking system in the course of 2007, and is also small compared to the trillions of dollars of losses in financial wealth and in foregone output subsequently incurred.

So that was part one – straightforward, but not on the whole calamitous – certainly not for the global financial system or even for the US

It is immediately apparent that while the problem in the sub prime mortgage market marked the beginning of the crisis, it was not in a deeper sense the cause of what we later came to call the Global Financial Crisis. At that stage it was still The US Sub Prime Crisis, and it wasn’t nearly as serious as it became.

So how did delinquencies in a relatively small part of the US mortgage market become a global financial crisis of the kind we now witness? The answer is that sub prime problem became a general problem specifically because of the nature of new financial products. Sub prime loans had been incorporated, along with hundreds or thousands of other loans into the backing for collateralised debt obligations (CDOs). These products were issued over-the counter rather than through exchanges, so prices were not readily observable for the whole market. The impact of sub prime losses on the value of the bond was difficult to calculate as delinquencies rose, making the CDOs harder still to value. Sellers were assumed to have more information on the value of the security, which made buyers wary. As a result, trading in these securities stopped, and along with it trading in mortgage backed securities generally and a number of other asset-backed products.

Once trading ceased, and values plunged, it was only a matter of time before major financial institutions would be in big trouble. This is because so many of them held these securities. Some of them were held by banks in off balance sheet vehicles financed by issuing short term debt. Banks were obliged to pay out the short term debt as other lenders to the vehicles withdrew, and put the assets on their balance sheets. More critically, major investment banks held as assets enormous trading portfolios of CDOs and related securities, which could not be sold and which were rapidly declining in market value. These institutions typically held relatively little capital against these assets, and the capital was rapidly eroded as the values of the portfolios declined. The CDO assets, now illiquid and difficult to value, were financed with highly liquid and short term loans, often in the commercial paper market, which began to be withdrawn from the investment banks. Unable to sell the assets, faced with a growing demand for the repayment of debt, the investment banks ran into trouble one after another. This was the basic problem with Bear Sterns which went under in March 2008, and with Lehman, which went bankrupt in September. Because Lehman was bankrupted, lenders in the short term money market incurred losses, and that market for company paper temporarily closed. So too, lending between financial institutions temporarily halted, because all major institutions had exposure of one kind or another to Lehman, and they could not be sure which of their fellows would be the next to go down.

There were lots of other things happening of course, but that was the basic story of this financial crisis. It began with delinquencies in sub prime lending, it affected major financial institutions because they owned financial assets which included sub prime loans, and it caused a wider crisis when lenders would no long support some of these institutions. It is now coursing through the real economy. Bank lending is restricted in the US the UK and Europe as banks work through their losses, and run their assets down to match the depleted level of their capital. Consumer and business confidence has plummeted, and along with them investment and consumer spending. Asset prices have tumbled as institutions have sold off what they can to reduce assets, which can no longer be financed. In the US and the UK and some other economies house prices are falling, diminishing household wealth and building activity. The fall in share prices has also reduced wealth and increased the cost of equity capital. The economic impact of the financial crisis is all the greater because in both the US and the UK it coincided with the beginning of the end of housing booms, and of course because the economies most affected by financial losses – the US, the UK and Europe – account for over half of global GDP.

I have deliberately confined the explanation of the causes of the crisis to two very straightforward mechanisms. It seems to me the best and most convincing explanation is the simplest and most obvious, so long as it is both necessary and sufficient to account for what happened. As I pointed out earlier, both of these mechanisms originated within the financial system rather than in the wider economy. Neither of them was inevitable or the necessary consequence of wider economic trends. Neither of them could have operated in a financial system which was well regulated.

I will say a bit more in defence of this proposition a bit later but first I want to turn to some of the wider and I think less persuasive accounts of the causes of the crisis. In accounting for the crisis, we need to be wary of the temptation to list all the major trends in the global economy in the last 20 or 30 years, and blame them for inducing what is likely to be the biggest downturn since the Great Depression.

This analytic approach involves an assumption that these trends could have no outcome other than this slump, which I think is untrue. It also neglects all the considerable benefits of most of these trends. For example, it is unquestionably the case that financial globalisation facilitated the propagation of a rising delinquency rate in a sector of the US home loan market to financial institutions in the UK and Europe, but it would seem to me silly to blame globalisation for the crisis. It is unquestionably the case that securitisation of financial assets permitted mortgage originators to distance themselves from later delinquencies. But it does not necessarily follow that securitisation in all circumstances increases financial system risk. On the contrary, there are plenty of ways of controlling the separation of origination and risk. Indeed, one of the odd aspects of the current crisis is that the originators did hold on to the risky tranches of the debt they originated or arranged.

We also need to be wary of confusing an analysis of what got us into a crisis with an analysis of its consequences, or a view about what is necessary to get us out of the economic slump. Among the wider and grander explanations of the crisis that I want to mention are a few I am less sure about or perhaps I just don’t understand.

One suggestion I don’t understand very well is that the crisis was caused by China’s current account surplus and the US current account deficit. The link is said to be that China’s capital flows to the United States permitted interest rates there to be lower than they otherwise would be, and loans to be more freely available than they otherwise would be. It seems to me there are plenty of holes in that one. This is apparent from the Australian perspective. For most of the period Australia had a bigger current account surplus compared to GDP than the US and had no difficulty financing it. But it did not have particularly low interest rates, its banks did not lend large amounts to people with poor credit histories, and it did not have investment banks which bought illiquid assets with short term and very liquid debt (while preserving very little capital against risk).

China did not beg the US institutions to ignore prudential considerations. Nor did it urge the Federal Reserve or the SEC or any of the many other regulators in the US to ignore their responsibilities.

It is true that if the rest of the world did not have surpluses the US could not have deficits, but it does not follow that in these circumstances interest rates would have been lower, or loans less freely available. It depends on the counter factual we have in mind. If China over the period had consumed more and invested more and saved less, for example, and if the US has saved more and invested less and consumed less – all for the same level of global GDP, and the same level of global saving – one would expect global interest rates to be much as they were. In a world where capital can freely flow from surplus countries to deficit countries it does not matter if the saving occurs in the US or in China. At all events it is not obvious to me why global interest rates would be lower or higher in that particular counter-factual. It is interesting that the US is still running a current account deficit and it is still in part matched by China’s current account surplus, but no one these days says there is a savings glut or too much “liquidity”. I am often troubled by the many ways in which this term “liquidity” is used, but it is evident that the avidity of lending is something quite independent of the availability of funds.

It’s also often said that the US got into trouble because it has a lot of debt compared to GDP. These are explanations which show how much more debt there is compared to the post Depression years or even compared to a couple of decades ago. But it would be very puzzling indeed if debt had not increased as a share of GDP. The Depression after all was nearly 80 years ago and in a very different world. Since the end of World War Two we have had more than 60 years of reasonable average economic growth. In other words it is 60 years since a major wealth destruction event such as war between the industrialised nations, and 80 years since the wealth destruction event of the Depression.

In that time wealth has accumulated on a far greater scale than ever before in human history. This wealth is in the form of factories, mines, roads, ports, houses, capital equipment, and more highly educated workers. Wealth may also be held in the form of financial claims – as currency, as shares in an enterprise, or as debt contracts of one kind or another which are provided in exchange for money. The accumulation of physical wealth is driven by increasing output per hour worked, or technological advancement and capital deepening. Capital is a stock and GDP is a flow, and there is no reason to expect the growth rate of capital and of physical wealth to match the growth rate of GDP. In fact you cannot have capital deepening unless the capital stock does grow faster than GDP. The growth of the stock of financial claims should be related to the growth of wealth and to the extent of financial leverage – that is, to the assets or wealth a borrower must have to satisfy a lender there is a reasonable chance of getting the money back as agreed.

So we should not be surprised that capital stock has increased compared to GDP, and wealth has increased compared to GDP. In a modern financial system one would expect to see debt increasing compared to GDP because debt must be some variable function of wealth, and wealth is increasing compared to GDP. I am not saying that the level of debt in the US was sensible – only that we should expect it to grow faster than GDP, and that to show it has doesn’t get us far in accounting for the global financial crisis. The real issue is the degree of leverage, and there is no doubt that in financial businesses in the US leverage had increased quite dramatically. But if that is all we are saying in this debt story, we are not saying anything which is not widely agreed, frequently repeated, and well understood. Nor are we pointing away from the basic mechanism I have described or the remedies they imply.

I suppose there is more force in the argument that the crisis was caused by Alan Greenspan lowering the US cash rate to 1 per cent in 1993 and keeping it there until mid 1994. This is part of the story which says interest rates were too low, and this induced reckless lending. But I think we also need to recognise that the cash rate was increased at each successive FOMC meeting from the middle of 2004, and that the sub prime loans which went bad were those issued from 2005. It is also relevant that the ten year government bond rate – was increasing all through that period. It is true that credit spreads were narrowing, but that was not because of the Federal Reserve. It was highly likely influenced by the financial technology of securitisation and off balance sheet debt. There is a good argument to say that US Federal Reserve interest rate increases were an important factor in ending the house price boom, which was in turn the cause of the rapid increase in sub prime mortgage delinquencies. But as I have argued, the sub prime delinquencies themselves do not really explain the severity of the financial crisis which followed.

Again, it is said that the sub prime crisis is an aspect of a wider unsustainable boom on house prices, which lead inevitably to the global financial crisis. The Australian perspective is also useful here. According to OECD numbers house prices rose faster in Australia, the UK, Spain, Ireland, NZ, Canada and Denmark in the decade to 2007, than they did in the US. Household debt to disposable income is much the same in Australia in the US. Yet Australia – like Canada – has only a small proportion of delinquent debt in housing mortgages. You need something more to explain why the US got into trouble when Australia and Canada did not. The difference is simply the quality of regulation in banking and finance, and the culture of the banking community.

Finally it is widely and insistently said that the global financial crisis was caused by Americans consuming too much and saving too little. This is so universally said and apparently believed it is almost impossible to contest. But let us remind ourselves of some key facts. The loan delinquencies which originated the crisis began occurring among people struggling to buy a house. Not consumer loans, but mortgage loans – and mortgage loans which by and large were not refinancing an existing loan, but were the original loan used to actually buy the property. Sub prime delinquencies became a problem because of the overleveraged circumstances of major financial institutions – not households, not non financial businesses, but major financial businesses. The origins of this thing had in my view virtually nothing to do with the savings habits of the American consumer. It is ironic indeed that many of the economists and analysts who complained of the savings habits of consumers and warned of dire consequences worked for the very investment banks whose trading activities were soon to bring the global financial system to its knees. It would be a good and useful thing if Americans saved more and consumed less, and it would have been good and useful if the Bush Administration had brought US federal deficits down faster than it did. But I do not think these issues had much to do with the causes of the global financial crisis.

Finally, it is highly unlikely that the Community Reinvestment Act of 1977 had anything much to do with a global financial crisis which emerged 30 years later. This act did make it more difficult for regulated institutions to redline districts by requiring that lending in a district bear some relation to the bank deposits made within it. But it did not force banks to make loans to people unable to pay it back, and the notion that banks lent in the sub prime market through any charitable impulse or because they were cajoled by government is an unusually silly delusion. It can be shown that most sub prime lending was not in fact originated by lenders covered by the CRA. To blame a crisis on a piece of Democrat legislation from 30 years before, especially when there had been 20 years of Republican administrations since, is I think a particularly slender explanation.

It is the failure of these wider explanations which leads me back to the two basic mechanisms I mentioned. There was nothing about the lethality of the basic mechanisms which was in any way novel, or sophisticated. Sub prime borrowers are by definition people with poor credit histories, and the success of the anyway dubious transactions depended on a continuing increase in house prices. This form of financial stupidity is as old as the hills. So too the vulnerability of financial businesses which accumulate huge quantities of illiquid term assets, financed in the most liquid of overnight markets and supported by a razor thin margin of capital.

In summary, there are numerous contributing causes but the indispensable causes for this particular crisis were the peculiar nature of the US mortgage market, the changed nature of investment banking (especially in the US), and the refusal of successive US administrations, of congress and US Federal Reserve chairmen to property regulate risk in the US market.

The lesson I draw is that the US regulatory framework is appallingly bad and that we will have recurrent financial crises until it is fixed. It would be good to have global regulation but it is imperative to have US regulation. Whatever political difficulty there is in achieving US regulation it will be easier than achieving global regulation. There are plenty of good models – Canada’s regulatory framework is one, Australia’s is another. It is not complicated, but it is essential. It is appalling that the US has at least five separate federal regulators for its banking system. It is appalling that three different regulators claim responsibility for the credit default swap market – and that none of them actually regulated it. It is appalling that investment banks were for all practical purposes not regulated by any agency, unless you count the blessing of the SEC as regulation. The world needs US regulation and it also needs to correct the procyclicality of the mark to market rules supported by Basle 2, it needs to encourage its central to lean against house price booms and financial asset booms in general, it needs global oversight of permitted leverage, transparency and disclosure – especially of cross border businesses, as most serious players are.

References

Markus Brunnermeier, Deciphering the 2007-2008 Liquidity and Credit Crunch, Journal of Economic Perspective (forthcoming)

Charles Calomiris, The Sub prime Turmoil: What’s Old, What’s New and What’s Next, paper delivered at the Federal Reserve Bank of Kansas City’s Annual Symposium, Jackson Hole , August 2008

Gary Gorton, The Panic of 2007, paper delivered at the Federal Reserve Bank of Kansas City’s Annual Symposium, Jackson Hole, August 2008