The causes and consequences of the current financial turbulence

By Raghuram G Rajan

(pages 3-9 of printed journal)

The proximate causes of the crisis

The global financial crisis has two basic elements. One is that bad loans and bad investments were made, especially by the banks, resulting in excessive credit. The second element is that these were financed with a substantial amount of leverage based on short-term borrowing. I will argue that these two elements led to a sequence of events. But first, a number of questions need to be considered: Why were these bad loans and investments made? And why were they financed on such a short term?

To answer these questions, it is necessary to go back in time. In the late 1990s, the emerging markets faced yet another financial crisis. This occurred in Asia, especially East Asia; in Latin America, especially in Brazil; and Russia. The response this time was a little different from earlier responses. A number of countries, in East Asia in particular, decided to protect their economies effectively by managing their exchange rates to keep them as competitive as possible. This would increase exports and build up reserves as a buffer against any possible future turmoil.

So rather than absorbing savings from the rest of the world, many of these countries began to export savings, becoming net exporters of capital. The Latin American countries eventually also followed this course, both Brazil and Argentina running large surpluses. The upsurge of commodities prices generated surpluses in a number of commodity exporters. Where were those savings being absorbed? A good use of such financial savings in one part of the world would be if investment somewhere else increases to absorb those savings.

For a time, that happened in the industrial world. In the case of what became known as 'the information technology bubble', a lot of investment took place in industrial countries, perhaps excessively so. When that bubble burst, corporations in industrial countries became very cautious about investing. Nonetheless, stimulated by expansionary monetary policy, demand increased in two areas. One was household consumption, especially in the US, and the other was residential investments. A lot of money went into expanding housing, increasing asset prices and construction. All this was to the good, at least for a while.

Prices of houses rose in a number of countries - Ireland, Spain, the UK and Australia. The crisis first emerged in the US. Why the US? It had an extremely innovative financial sector. It saw a mismatch between the investment opportunities that were emerging in the real estate sector, and the financial savings that the rest of the world wanted to supply.

The alchemy of turning lead into gold

However, a central bank in Asia or an insurance company in Germany, for example, would be extremely reluctant to buy a local mortgage from a US bank because they would know nothing about the borrower and would be concerned about the liquidity of such an asset. To get over this problem, the local mortgage needs to be converted into a financial asset that the rest of the world is willing to buy. The US financial sector was very good indeed at this conversion process. It converted local mortgages into AAA-rated financial assets that the rest of the world would be willing to buy - a process of alchemy by which the lead of local mortgages was turned into gold that the rest of the world wanted to buy. These 'sub-prime' mortgages, made to borrowers who had little credit history, no employment security and no asset backing, were packaged together and then sold to international markets!

The driving elements

How was this possible? One element driving this process was that house prices were rising. Initially, those who got loans were credit-worthy; but as house prices kept rising, the risk of lending even to borrowers with no credit standing disappeared. Houses were effectively liquid assets and if borrowers defaulted, their houses could be sold without loss to the lender because house prices would have appreciated further. Alternatively, borrowers for housing could be attracted by low interest rates because they could sit comfortably at the end of the year on a ten per cent equity appreciation that would allow them to refinance at the lower rates.

Thus, on the one hand, rising property prices would cover all potential credit risks; and, on the other hand, banks had the ability to package and sell these loans in the international market.

The process works something like this. A local bank would put together in a package, say, a variety of 2,000 loans it had made to households. It would issue securities against that package. Based on past experience, some two per cent of these packaged loans could be expected to default. So if two per cent of the loans defaulted, then the first lot of defaults could be absorbed by buffers, called the equity tranche. Where there is very little risk of further default, higher rated securities could be issued.

For example, assume that $100 of mortgages are taken out. Against these mortgages, the following are issued: $10 of equity tranche, $10 of BBB-rated securities, $5 of A-rated securities, $5 of AA-rated, and $70 of AAA-rated securities. Why seventy dollars? Because the risk of defaulting for the AAA-rated securities requires 30 per cent default on the mortgages. This is unlikely, and was hence the logic behind the mortgage-backed securities.

However, the innovators were not happy to stop here. Having squeezed $70 of AAA-rated securities out of the mortgages, they could go further. The BBB-rated securities issued by the packages were then put together in a fresh package called collaterised debt obligation (CDO), against which 70 per cent of AAA-rated securities could be issued. The process of converting into AAA-rated securities could go on - CDO squared, CDO cubed - so that $100 of mortgages could be converted into $85 of AAA-rated securities and sold to the rest of the world. This was the process of securitisation and it went on for some time.

Consequential problems

There are, however, at least two problems. One is that these are very complicated securities with varying degrees of risk. They trade easily only if the chance the borrower will default is small. Once substantial defaults occur, these securities are in serious trouble because they are relatively new to the market. Moreover, when house prices fall across the US, the diversification that investors felt they had by buying packages of mortgages from all parts of the country becomes irrelevant.

The complexity is magnified by the fact that there is very little information about the borrowers and their credit worthiness. As was noted earlier, that did not matter when house prices were rising and mortgages continued to be serviced. However, once house prices stopped rising, such information mattered a great deal - but was not readily available.

The other problem, which the originators of these securities did not understand fully, is that as the securitisation process goes on, the value of information about the credit worthiness of packages also starts to drop off. In the late 1990s when a loan officer made a loan to somebody with a certain credit quality, they did not stop at asking what the credit quality was. They went further to assess more subjectively the reliability and honesty of the borrower in a process known as seeking 'soft information'.

In the circumstances, the result of this form of information-gathering is that there was very little correspondence between the public credit rating of the individual and the interest rates that were charged for different individuals. The credit rating was only the first step in the process of gathering information to be supplemented by 'soft information'. However, by 2005 there was a very close correspondence between credit ratings and interest charged. Why? Because banks made loans knowing that they were not going to hold on to them but were going to sell them to the securitised market. So they had no incentive to qualify the credit rating by soft information. Thus, the credit quality of the loans looked better on paper than they actually were.

In the US, there is a credit rating called the FICO Score (Fair Isaac Corporation) as a measure of the borrower's credit quality. This is the guideline used by Freddy Mack and Fannie Mae.¹ A credit quality rating at and above 620 would allow the loan to be securitised and sold in the markets. Two things happened when this was introduced. One, not surprisingly, was that the number of loans just above 620 took a quantum leap, suggesting that securitisation made more loans available. The other, and unexpected, outcome was that those with a score above 620 (with securitised loans) were more likely to default than those below 620 (not securitised). This suggests that securitising reduces the incentive for lenders to be concerned about the quality of the loans. Thus, there was a steady deterioration in the quality of loans - even though, on paper, their credit score was higher.

Who is to blame?

When did the system break down and who is to blame?

One factor is a lack of knowledge about the quality of the securities. Buyers of securities in Asia and Europe were willing to accept securities so long as they were highly rated. However, it was difficult, even for the financial experts, to distinguish between triple-rated securities that were CDO, CDO squared or CDO cubed. The degree of risk varies greatly between them. An AAA-rated mortgage-backed security would probably be trading at 60 or 70 cents in the dollar while an AAA-rated CDO cubed would be trading for 20 cents in the dollar.

The second factor is the work of the rating agencies. Product rating became a big part of business for agencies like Moody and Standard & Poor, and they had very little incentive to drive away business by being tough on ratings. It is arguable that they should have recognised that this was a new kind of financial instrument which they had little experience with, and that it was different from corporate debt securities. Therefore, they should have been more cautious about their ratings. Alarm bells should have rung as these securities were paying 50 and 100 basis points above AAA-rated corporate securities.

Investment banks were also to blame. They should have known that they were packaging securities with varying risks, which the market was prepared to buy. Their confidence was reinforced by the fact that the investment banks that put together and sold these securities held on to the riskiest portions themselves as a sign of good faith on the quality of the packages, although many sold these holdings later.

Finally, many of the homeowners cannot be absolved from blame. They would have known that they could not afford the loans they incurred but were perhaps deluded by the prospect of a 'free lunch'.

A puzzle - the case of Citibank

One of the intriguing questions is why the banks held on to these securities? They had huge amounts of them, not just for inventory purposes but also as investments. One possibility is that they really believed in their product. Any risk would be in the category of a 'one in a thousand year flood'. But this is surely an illusion because it seems to happen every 10 to 15 years with regularity. Citibank has been in trouble three times in the last three decades - in the mid-1980s because of loans to developing countries; in the 1990s because of real estate loans, until it was rescued by a Saudi prince, who injected a substantial amount of equity into the bank. And now because of loans made in the mortgage-backed market. It is a worry when the largest bank - indeed perhaps the icon of American banking - is in trouble every 10 to 15 years.

One possible explanation is that there was a great deal of competitive pressure. In the words of the Chairman of Citibank, 'When the music stops, in terms of liquidity, things will be complicated. But so long as the music is playing, you've got to get up and dance.' Merrill Lynch went into mortgage-backed securities in a big way because it saw Goldman Sachs making a great deal of money from it. The 'herd mentality' seems to plague financial markets.

It also appears that the left hand of the bank did not understand what the right hand was doing. Those who were packaging the loans did not fully indicate their quality to those who were making the loans. For example, the investment banking unit of UBS borrowed at UBS's cost of capital, investing in AAA-rated mortgage-backed securities, thereby making a spread of 15 to 20 basis points. This does not look like a lot of money, but when multiplied by a trillion it amounts to a great deal.

There is a saying in the financial markets: 'There is no return without risk.' Why were they making 20 basis points extra if there was no risk? Top management did question the wisdom of such a large volume of investment. But their voices were drowned by the argument that the investment bank was making two to three hundred million dollars in profits every quarter.

The regulatory standpoint

From a regulatory standpoint, as the cycle progresses, risk management becomes less and less able to exercise control on risky activities. As long as the cycle progresses, the risks do not show up. At the point of maximum danger to the bank, risk management may be at its weakest. Risk control mechanisms seem to break down all the time. It is not that boards are stupid. They are made up of clever people. Robert Rubin is one of the smartest treasury secretaries. At Citigroup, though, he looked on while it suffered more than $72 billion in losses.

The debate will go on for some time, but it does seem there was a total breakdown in governance and in the compensation structures of the banks. To draw a simple analogy: the banking system was writing earthquake insurance - it was taking long-term risks, but the premiums collected were not set aside as reserves for the day the earthquake actually happens. Instead, the premiums were being paid out as bonuses and dividends to shareholders. When the earthquake actually happened, the banking systems were grossly under-capitalised and there were no reserves to draw from. The compensation structures were such that payments were made for short-term performance.

Leverage

As noted, a critical issue in the crisis is the extent of leverage. Short-term leverage is perhaps the cheapest form of financing when risky activities are undertaken. Short-term borrowing was rolled over relatively risk-free because there was an ample supply of finance. Moreover, borrowing short-term was cheaper than long-term, especially where financial institutions were taking substantial risks, as it allowed the lender to get out more quickly.

Short-term debt was also preferred because of the market's reliance on the 'Greenspan Put' policy, which stated that, if a serious downturn eventuated, the Federal Reserve would facilitate liquidity promptly and cut interest rates sharply. This factor may have added to acceptance of the large volume of illiquid assets financed with short-term debt.²

The elements of the crisis as it unfolded

Eventually house prices stopped rising and mortgage defaults began to increase. The equity cushion provided by rising house prices in refinancing loans disappeared. As the mortgage defaults mounted, the securities issued to back these mortgages became difficult to price. Suddenly, the market stopped accepting the securities that the investment banks had financed by borrowing short in the market. Banks became illiquid. For example, Bear Sterns was unable to find $41 billion of securities and had to go to the Federal Reserve for financial support.

The second stage of the illiquidity problem emerged as these mortgage-backed securities continued to fall in value because the banks were essentially insolvent, the size of the losses having eliminated their capital backing. The maturing of their short-term debts compounded their insolvency. Long-term debts are sustainable because the day of repayment is some distance away and the insolvency problem can be deferred. Not so short-term debts.

The myth that bank deposits or debts can be repaid only holds so long as most creditors of the banks do not call for their money on the same day - because banks hold most of their assets in illiquid securities. Mechanisms have been developed - such as deposit insurance and borrowing facilities from the Federal Reserve - to deal with any potential run on the banks. However, while these mechanisms were available to commercial banks, they were not available to investment banks. When a run started on Lehman, the Federal Reserve and the Treasury were not willing to support them financially. The result was a run on Goldman Sachs and JP Morgan; and on a number of banks in Europe in panic mode.

With the benefit of hindsight, it appears that the response of policy has always been one step behind the event. Policymakers thought the problem was inflation when it was illiquidity. They started tackling liquidity when the problem had already developed into insolvency; and they started tackling insolvency when the problem had turned into full-scale panic. Eventually they found themselves having to deal with full-scale panic.

In the US - and this may change with the new administration - the remaining problems in the system are not being fixed in time. Ideological issues have intervened. Right-wing ideology is resistant to large scale public intervention, despite the fact that the private sector is in great trouble and needs intervention. To intervene marginally is the worst form of intervention. While it is desirable to avoid intervention, when public money is at stake, everything necessary should be done to ensure the safety of such money and to put the system back on track. At present, this is not being done.

There is a mistaken belief that money pumped into the large banks will somehow find its way into the rest of the undercapitalised banking system. Such a course could lead to the banking system being cartelised by a few leading banks. To some extent, this was happening with the Bank of America, Wells Fargo and JP Morgan taking over large banking assets. However, it is unlikely to go much further. The remaining banks that the Treasury is not willing to capitalise will remain in difficulty. To deal with them effectively will probably require major intervention in order to decide whether they should be closed down or kept alive with financial infusion. Yet it is unlikely that the US authorities would be willing to undertake such a course, which means the problems could continue to fester. Recovery may come in time but it is possible that things could get worse, and intervention on the scale needed will be undertaken.

Some lessons

First, it should be understood that the entire financial system is integrated not just within a country, but across the world. Problems can emerge from anywhere and infect the rest of the system. We have had runs on banks in India and Hong Kong, based on some notion that they may be exposed to crashing prices, when in fact the exposure was not that great.

Second, in monitoring risks, regulators are often focused on the wrong places. Before this crisis, much attention was on hedge funds in the belief that hedge funds were going to be the problem. In fact hedge funds, especially the larger ones, are generally well managed. They have fairly good incentive systems in place and fairly good risk management structures.

Third, there is an increasing tendency to regulate the financial system more strictly by imposing much higher capital requirements on the system. However, this becomes part of the problem. By imposing large capital requirements, a great deal of banking activity was driven from the regulated part of the system to the unregulated part. The Structured Investment Vehicle³ (SIV) was in some sense a way for commercial banks to arbitrage - as in, to avoid - capital requirements by creating entities that were not subject to capital requirements. These eventually turned out to be banking risks because they came back on their balance sheets.

A great deal of our regulation assumes that management has control and cares about the long run. However, the problem has been that management did not have control and did not care about the long run. This is a governance problem, and unless it is fixed, regulation has no hope in doing anything useful. Having a variety of institutions helps. Things could have been much worse but for the deep-pocketed investors - the sovereign wealth funds and the Warren Buffets of the world - who have helped to stabilise the system.

A final thought

We keep hoping to create a financial sector that is stable. However, we should accept that competitive financial systems carry the small risk that they will collapse periodically. Rather than writing more and more 'fire codes' to prevent the fires, we should recognise that fires will happen. This is not to say that we should not keep trying to get things right; but it should be understood that fires will break out. Accordingly, more time should be spent in making sure there are sprinklers to put out the fires, rather than simply writing fire codes.

In other words, let us make sure that when a crisis occurs, the private sector does not just dump it into the lap of the public sector to deal with it. Instead, the private sector should be required to devise a mechanism by which it, rather than the taxpayer, bears the cost of any breakdown in the financial system. This calls for further thought. In this connection, it has been suggested that a form of insurance, called capital insurance, should be available to enable firms and banks to buy a form of insurance that may help to reduce serious consequences of a financial crisis.⁴

¹The Federal Home Loan Mortgage Corporation (Freddie Mack) and the Federal National Mortgage Association (Fannie Mae) are government sponsored enterprises buying mortgages in the secondary market, and selling them to investors as mortgage-backed securities. Both are directed to assist home ownership of low and middle income families.

²Data six months ago show that the banks with the greatest short-term leverage and the least amount of capital were the ones that suffered most in terms of stock prices when the crisis hit.

³An institutional entity based on funds borrowed on short- term securities at low interest rates (close to inter-bank interest rates) and buying long-term securities at higher interest rates.

⁴See Anil K Kashyap, Raghuram G Rajan and Jeremy C Stein, 'Rethinking Capital Regulation', Federal Reserve Bank of Kansas City symposium on 'Maintaining Stability in a Changing Financial System', Jackson Hole, Wyoming, August 21-23, 2008.

Top ^