7

In God We Trust: banks and their errors

The bill so far

The International Monetary Fund (IMF) estimates¹ the total losses to banks on loans and securities will exceed $4 trillion, of an estimated total of $54 trillion outstanding – a loss rate of 7.4 per cent. Past financial crisis have been thought to be bad when losses exceeded 5 per cent. This episode will be extremely costly for society, with by far the largest bill in the short term presented by the banks themselves, but with a much larger longer-term bill potentially presented to taxpayers. There is simply not enough private capital available at present to make good these losses; government assistance was inevitable. In spite of the large amounts of government capital committed, there may be more needed, even now.

The IMF also pointed out that although the crisis began in America, and concentrated on its assets, the effect will be felt by banks elsewhere, particularly in Europe. Losses on loans and securities among European banks – not including those from the UK – are expected by the IMF to amount to $1.1 trillion, about 10 per cent higher than the losses expected by US banks. Partly, the added losses are due to investments made by European banks into America, but they also reflect their higher exposure to emerging European economies, an exposure US banks largely avoided. UK banks alone are expected to accumulate losses of $316 billion, a third of the losses of America, for an economy that is approximately one tenth of the size.

The bill from the banks may continue to be paid by citizens for another generation. The discussions about how to reform the system to avoid a similar calamity in future are already well advanced. How did banks get here? What do the losses represent in geo-political terms? These are the questions we’ll look at in this chapter.

Tensions arise

American capitalism focuses on market-delivered lending and consequently emphasises a form of money based on the ‘medium of exchange’ as much as debt. But as we saw in Chapter 2, money grew from the transfer of debts and the settlement of these debts. The tension between the two has become unbearable. These tensions have always existed, even at a local level, but they have become much greater with the worldwide financial marketplace. The resolution of this tension, or amelioration of its worst effects, is essential if the benefits of globalisation are to continue.

It is also of more immediate concern. A failure of a banking system is a potentially a failure of an entire society. This is what almost occurred in late 2008, in the aftermath of Lehman Brothers’ bankruptcy. For two weeks in late September and early October there was no banking system capable of more than the most trivial transactions, and sometimes even these transactions failed. Transfers mysteriously arrived late, or did not arrive at all. Banks refused to deal with other banks. Several European banks (Dexia, Hypo Real Estate, Fortis, Royal Bank of Scotland, Barclays and HBOS) almost followed Lehman into administration. Governments across the developed world were forced to provide aid and the European Central Bank and Bank of England provided huge loans to prevent another major lender from collapse. Lehman’s collapse triggered a systemic financial crisis of unprecedented proportions. No other crisis had spread across so many countries and continents so fast before. For a few weeks it seemed better to hold gold coins than a debit card in case the system collapsed. Some people did so.

The crisis had been burning for over a year before Lehman crystallised the frightening prospect of a collapse in the payments system. The surprise, in retrospect, was not that the collapse of an American investment bank would cause such devastation. The surprise, to those who work in finance, is that those who allowed it to happen did not realise the likely consequences of insolvency, especially when there had been such ample warning of the fragility of the system. In addition, US Treasury secretary Hank Paulson, the man who made the bankruptcy inevitable, had worked at an investment bank, Goldman Sachs, for thirty years before joining the Bush administration. During that time it is surprising he did not learn how serious the implications of a bankruptcy would be, even for a relatively small bank such as Lehman. How did the Western banking system get to this state?

Originate and distribute becomes ‘originate and hoard’

Banks perform three roles: they create credit through a balance sheet entry; they take savings from one set of customers and lend to another set of customers; they are commercial custodians of a system of payments based on trust. When banks create credit they expect that credit to be balanced by a lender; if not immediately, then sometime in the future. A bank’s role is to move money to where it will be used; they are intermediaries for the transfer of money in the financial system. For that to continue they must ensure that their own privileges in creating money do not overshadow their accepted role in payments. The utility role of banking, the payments system, has traditionally been separated from the investment banking role, which sees payments merely as the end result of transactions. This is a good example of the clash between the role of money as debt and as medium of exchange.

In the past twenty years the role of banks in the payments system and the increase in transactions caused a tension between the two purposes of money. ‘Originate and distribute’ is a model of banking that shifts the creation of credit into a creation of securities sold to investors; the debt is originated and distributed to investors in packages as securities. This was an important change in the way credit had been created in the past. Freed from the necessity to find a lender to create credit, banks saw credit creation as a series of transactions, rather than as debt. It was a system that had a great deal going for it; more efficient use of capital, a way to gain market share without a large balance sheet, a way to spread risk to those best able to look after it.

Unfortunately, while most of the expectations were met, ‘originate and distribute’ failed to transfer risk to those best able to manage it; it seemed rather to transfer risk to those least able to understand it, or not to move it at all. Many banks faced with competition for investment returns chose to keep the securities themselves and receive a better-than-average yield. As a consequence, they were left with the problems once the assets were rejected by other investors.

Would it have made a difference if the banks had sold the securities to other investors? The damage to the payments system might have been avoided. If the securities had been passed on, as intended, the losses would probably have been treated under different accounting rules. This might have avoided the catastrophic collapse that occurred in October 2008, even if eventual losses were very large. Why did banks not pass on the securities?

One reason why banks held on to securities was that they needed to have their cake and eat it to survive, or at least to be regarded as competitive by their shareholders. Competition had grown to such an extent that extra earnings from securities were needed. Banks were also persuaded of the benefits of securitisation because they were involved in constructing the vehicles. Low real rates encouraged banks to keep larger amounts of securities on their balance sheets and fund them in the wholesale market. The Federal Reserve (and other central banks) had shown they could be relied upon to deliver liquidity if there were temporary difficulties and the cost of funding was favourable compared with the higher rate of return obtained from the securities. The objective of banks turned from providing credit into managing assets backed by credit, or gathering fees from originating these assets. For a while this worked splendidly, as the annual returns and the share performance of banks from 2002 until 2007 testified.

Ways were found for banks to benefit directly from the securities they issued, rather than pass them on to investors. Structured investment vehicles – which extended the balance sheets of banks outside international banking regulations – were extended. More importantly, fund management arms of banks were extended.

The most attractive form of fund management was that practised by hedge funds. Hedge funds can make the highest returns, and generate the highest fees, using the lowest overheads of any fund management technique. Banks saw the hedge fund industry was becoming increasingly mature, and concentrated. If banks were to continue to benefit from an industry that had provided an increasing stream of income since 2001, they would have to emulate it. From 2002 onwards, hedge funds were either established by the banks themselves or bought from outside banks. JPMorgan took a majority stake in Highbridge Capital, which became the largest hedge fund in the world. My own employer, UBS, established a hedge fund of its own, Dillon Read Capital Management, to gain from the fees and high returns of the sector.

Acquisition of fund management arms was, in retrospect, a dangerous overextension within banking. But it was no more than a continuation of a cycle of competition that has always dogged banking. The most successful hedge fund managers tended to come from bank trading rooms. Banks had been losing their best brains for years before they responded by either taking stakes in the funds their traders were establishing, or establishing funds of their own to keep traders while still offering them similar incentives.

The ease with which banks sourced low cost of funding, through deposits or through wholesale markets, also supported the move to holding securities that gave a better return.

An era of free money

Much has been made of the effect of ultra-low levels of interest rates offered by the Fed between 2002 and 2005, during which rates were below 3 per cent. I have commented on this as one of the causes of the crisis elsewhere. The Fed certainly over-stimulated the US economy. The European Central Bank has received less focus but it too provided easy money. In addition, the wide variety of collateral accepted by the ECB when it lent money to banks enhanced the policy of low rates further. The acceptance of asset-backed bonds as collateral by the ECB encouraged banks to issue more of these securities and to hold them on their balance sheets for presentation to the central bank. Through accepting lower quality assets as collateral in its finance operations, the ECB widened the definition of ‘central bank money’ to include the types of debt that would later cause problems.

Cheap funding became the norm on both sides of the Atlantic. Some banks were able expand their cheap funding through their customers. Private banking and derivative management systems often required large amounts of low-cost cash and this was delivered to banks, either as collateral or as funds on deposit awaiting investment. Many of the structured investments created during the boom years were prompted by an aim to fund banks cheaply. Investors would receive a coupon above the market interest rate as long as a particular event occurred, perhaps a steepening of the yield curve. The yield curve exposure was hedged, leaving solely a low-cost source of funding for the bank.

Funding costs also seemed guaranteed to remain low for a long time. Inflation rates were low and central banks felt they could afford to relax their vigilance over prices. Rate increases after 2004 were slow despite a strong economic recovery in the US. They were also flagged well ahead of time, allowing banks the opportunity to plan ahead. Central banks had also demonstrated their willingness to cut rates if there was a financial problem – as in the Long Term Capital Managaement episode and after the dot-com bust. This encouraged risk-taking because any difficulty was expected to be met with a swift reduction in rates. With many funding products tied to Libor rates – the benchmark rates set in London every day that reflect the cost of bank funding – and central banks apparently intent on keeping rates predictably within a defined, and low, range, the funding of assets became much more attractive. Not surprisingly, leverage ballooned.

The growth in repurchase agreements (repos) reflects the confidence in the source of funding. A repo is the lending of an asset in return for a loan of cash – or lending secured on a bond, or share. The huge increase in repo transactions and outstanding amounts of repo from 2003 onwards shows the extent to which the banking systems of America and Europe were creating their own medium of exchange through the transference of commercial debts. In June 2001, outstanding European repo positions stood at €1.8 trillion. By June 2007, their peak, the outstanding repo positions stood at €6.8 trillion, a growth rate of 25 per cent a year, much, much faster than the rate of economic growth. A similar strong rate of growth was also recorded in the US.

Why mortgages?

But why was there such an explosion of mortgage lending? One reason is that following the dot-com bust, the finance industry had run out of investable opportunities, yet the dollars kept coming into the country in recycled deficits. The ‘sunny uplands’ of technology innovation never captured the financial imagination again after 2001. Ironically, the profitability of web-based companies improved and the web went on to produce the most spectacularly successful service firm in history with Google. However, banks chose not to pursue the equity route after 2001. Equities appeared inherently less stable than fixed-income products, and more difficult to price reliably.

Fixed-income products are usually more amenable to analysis than equities because their cashflows are more predictable. After the financial and regulatory problems of the dot-com bust, investment banks chose to turn to the more certain asset class of fixed income. Of course, what you need to price fixed income reliably is consistent funding and predictable default rates. The lack of both would prove that fixed income can be as unreliable as equities. But in the dark days of equity markets in 2002, fixed income was a ray of hope.

The dominance of real estate lending among American investment banks was not just due to financial innovation. The dot-com bust of 2001 removed equities and venture capital as a source of funds and the investment banks that did worst in 2001–2003 were those that had concentrated on equities. The banks that had a market lead in mortgages reported consistently record results. Leading the industry were Lehman Brothers and Bear Sterns, both of whom would later collapse.

Regulators also encouraged the mortgage bonds from ‘private label’ issues, those outside the agency model, which would later prove the most toxic. In 2003, the US regulator of the agencies (OFHEO) placed limits on the balance sheets of Freddie Mac and Fannie Mae following revelations of manipulation of their earnings over a number of years. We now believe this was a political gain for the Bush administration, and allowed banks to provide the securities that investors had previously sought from agencies alone.

Investors had been keen to buy mortgage-backed bonds throughout the 2001 recession; now they switched to non-agency bonds in increasing amounts. The agencies never recovered their dominant position until it was forced on them by the government in the aftermath of the credit crunch.

With the agencies hobbled, investment banks were keen to package and sell the bonds, acting with mortgage originators across the country. Lehman Brothers and Bear Stearns had always been dominant in the sector. After 2003, they were joined by Merrill Lynch, Goldman Sachs and Morgan Stanley. From 2003 till 2006 the fee income from mortgage issuance for these banks – and related securitisations such as CDOs – outgrew all other major streams of revenue.

Morgan Stanley’s annual report of 2002, at the depths of the equity decline, was depressing:

‘2002, it was a most difficult year. We began the year with the memory of 9/11 still fresh in our minds. In the months that followed, we had to contend with the continued decline in equity markets and a highly publicised investigation of Wall Street research and investment banking practices.’

In 2003, net income declined from $3.5 billion the year before to less than $3 billion, but signs of recovery were building in the fixed income market. The company shed 14 per cent of its workforce during the year but reported ‘fixed income had an excellent year’. Fixed income contributed to the improved results again in 2004. ‘2004 was a very good year for your firm. I am pleased by the strategic initiatives we put in place. Record revenues from our fixed-income division.’

By 2005, Morgan Stanley confidently announced:

‘We are masters of complexity and engineers of creative solutions… We boast a superb investment banking franchise, pre-eminence in key businesses such as prime brokerage and commodities, and leading positions in high-growth emerging markets… In 2005, we achieved record fixed income sales and trading revenues, record prime brokerage revenue… We will expand our capabilities across fixed income, equities and investment banking by creating structured products and more innovative solutions for clients.’

Even in 2005, there was more than a whiff of hubris. The following year, the annual report was even more celebratory.

‘Net revenues rose 38 per cent to $21,562 million driven by record results in fixed income and equity sales and trading and fixed income underwriting, along with strong results in advisory revenues… Fixed income sales and trading revenues were a record $9,577 million, up 41 per cent from a year ago.’

Since 2004, Morgan Stanley had been retaining securitised assets on its balance sheet, presumably in the form of highly-rated tranches that were efficient in terms of meeting the regulatory capital rules of Basel I set in 1988. The Annual Report of 2006 reported: ‘increase in other revenue was attributable to an increase in the fair value of the Company’s retained interests in securitised receivables.’

The bank did not take on the risks with its eyes shut. In the ‘risks’ section of the annual filing with the Securities and Exchange Commission, Morgan Stanley listed ‘liquidity risk’ as the number one risk to its business.² In the event, it was exactly this liquidity risk that brought down many in the credit crunch – though not Morgan Stanley.

Merrill Lynch and Goldman Sachs followed a similar business development path to Morgan Stanley. As with Morgan Stanley, they worried about ‘liquidity risk’. Goldman Sachs indicated its concern as early as 2003:³

‘We also perform various scenario analyses, asking “What if?” about any number of possible events. Access to liquidity remains the single most important issue for any financial services firm… We place major emphasis on assuring our access to liquidity. The cornerstone of our approach is a “cushion” we maintain in the form of cash and highly liquid securities that averaged some $38 billion in 2003.’

It was a policy that stood the company in good stead; Goldman Sachs emerged as perhaps the strongest investment bank in the world in the aftermath of the financial crisis; ready to take over the shrunken franchises of its previous competitors.

Lehman and Bear Stearns had the advantage of being ‘first-movers’ in fixed-income markets. The 2003 annual report of Lehman Brothers wrote: ‘The firm’s fixed-income business produced record revenues of $4.4 billion, an increase of 68 per cent versus the previous high achieved in 2002.’ Lehman’s also reported its continued push into mortgages. ‘The company acquired a controlling interest in Aurora Loan Services (ALS), a residential mortgage loan originator and servicer.’ The company’s profit ‘increase was principally driven by strong institutional customer flow activity, particularly in mortgage related products’.

By 2005, Lehman was forging ahead in fixed income, mostly related to mortgages. The company’s report for that year said:

‘[In] fixed income capital markets the notable increases were in commercial mortgages and real estate, residential mortgages and interest rate products.

Revenues from our residential mortgage origination and securitisation businesses increased in 2005 from the robust levels in 2004, reflecting record volumes and the continued benefits associated with the vertical integration of our mortgage origination platforms. We originated approximately $85 billion and $65 billion of residential mortgage loans in 2005 and 2004, respectively. We securitised approximately $133 billion and $101 billion of residential mortgage loans in 2005 and 2004.’

In 2004, Lehman had an 8 per cent market share of originating non-agency mortgages for mortgage-backed securities (MBS), and a 12 per cent share of securitising these mortgages. Bear Stearns was the firm that specialised most in mortgage products. It was consistently rated the top securitiser of MBS securities from 2000 onwards.

Given the extreme gearing, and the lax standards that emerged in the mortgage markets of both America and Europe after 2005, it is absolutely no surprise that these the two companies were also the most spectacular investment banking failures. They led the field for so long, and were so wedded to MBS, it would have been amazing if they had survived the crisis.

Mortgage originators also grew fast. Countrywide’s earnings before tax were $842 million in 2002 and nearly tripled the following year to $2.4 billion. From 2003, increasing competition from other companies and increasing costs – from compensation and rents plus ‘other operating expenses’ – meant the bottom line remained virtually static from 2003 till 2006 at around $2.5 billion. Profits struggled to improve even though net interest income doubled between 2003 and 2006 from $1.4 billion to $3 billion.

The servicers did not have the field to themselves for long. Other companies were quickly filling the space to take advantage of the riches available. Some were created by investment banks. In 2003, Bear Stearns bought Texas subprime originator EMC and its acquisition immediately contributed to the fixed-income department’s profitability. The subsidiary quickly took on a balance sheet persona of its own. In 2000, EMC’s principal holdings were $4.2 billion; by 2005 it had holdings of $69 billion. Lehman Brothers also began acquiring companies to feed it with mortgages.

Monoline bond insurers are specialised companies that use their financial strength to guarantee bonds, and in return they take a fee. One monoline company, Ambac, initially benefited from the expansion of asset-backed bond issues but quickly found increasing revenues depended on taking on increasingly risky underwriting. Revenues rose 46 per cent at Ambac between 2002 and 2004, but in the subsequent two years growth fell to 15 per cent, then 13.5 per cent. Insurers had to move to underwriting products that were much more highly structured – and also more risky, such as collateralised debt obligations (CDOs) – to maintain their revenues. As we now know, monolines were among the principal victims of the crisis, which exposed their weakened positions after years of declining revenues.

One of the consequences of the explosion in origination was that banks frequently lost money on lending mortgages, but made it back through creating and selling the securities or by holding them and relying on cheap funding. For instance, a two-year, fixed-rate mortgage in the US and the UK was frequently offered with an interest rate below the cost of two-year bank funding. Mortgage borrowers received lending on better terms than the banks who were lending to them. The bank would not mind if it was convinced it could finance itself at a lower rate in the short term, or sell the securities, or create a CDO. After nearly a decade of this, the amount of short-term wholesale funding in the banking system became unable to support the number of assets held on – and off – bank balance sheets in any sustained financial disturbance.

The Great Moderation

Throughout the period of explosive growth in mortgage lending, global financial markets displayed remarkable calmness. Economists and central bankers congratulated themselves on their contributions to what became known as the ‘Great Moderation’; high, stable growth, low inflation, flat yield curves, contracting spreads, falling insurance costs and lower volatility. Almost every measure of investor appetite for risky assets showed investors were convinced by the Great Moderation story; spreading their investments into previously shunned areas. By 2005, even Africa, a continent usually shunned by mainstream investors, was feeling the benefit of investment flows from Western investors.

We’re losing money, buy more bonds!

Under the surface, calm banks were feeling increasing pressure. The flat yield curve and low bond spreads caused margins in America and Europe to fall to the lowest point in 16 years. US bank margins would have been even lower except banks were slow to pass on interest rate rises to depositors. Banks became more reliant on increased securitisation fees and off-balance sheet funding from structured investment vehicles. Just at the time banks would normally aim to attract depositors with higher rates, which would shore up funding if the economy began to slow and defaults rise, financial innovation was taking banks away from depositors. This was to prove disastrous in mid-2007 when the wholesale market seized up.

If low funding costs helped banks to move into securitised portfolio management, the competition in banking also kept rates down and low rates offered to depositors became a way to remain competitive. The low rates reflected increased pressure that bank balance sheets were experiencing even in 2006 in the face of an unusually flat curve and intense competition.

Originators responded to the competition by selling more subprime mortgages, to maintain profitability through volume. A large number of commercial banks chose to sell low-yielding, but high-quality, mortgage assets in 2006 to relieve funding pressures. Some also purchased non-securitised mortgage loans, which offered higher income than bonds. Some banks, such as Bank of America in late 2006, chose to limit the amount of securitisation of their new mortgage loans because of a lack of returns. Yet, rather than opting for a strategy of safety, Bank of America exchanged low-yielding, high-grade bonds for even riskier assets (loans) in an attempt to enhance returns without adding to its securitisation holdings. The gain in yield was small – calculated to be only 0.2 per cent on the amounts they exchanged. Pressure to maintain profits pushed US banks into making or acquiring loans more aggressively, even as the credit cycle was turning sour. From 2006 onwards, many mortgage originators, borrowers and banks were dealing in wildly optimistic expectations that the world would always be benign.

The Bank of England governor Mervyn King summed up the developments:⁴

‘Investors, including banks, overlooked the fact that higher returns could be generated only by taking higher risks. As a result, money was lent on easier terms. That helped push up further asset prices that had already risen as real interest rates were falling. It also led to an explosion in the size of the financial sector as new instruments were created to satisfy the search for yield. As well as lending to households and businesses, banks lent to other banks which bought ever more exotic instruments created by the financial system itself. The effect was to replicate the original risky loans many times over.’

Originate and distribute, the case of UBS

Europe followed America, in America and in Europe. Home-grown investment banks – UBS, Deutsche Bank, Barclays – could enter the US mortgage market directly, and often benefit from a cheaper funding base back home. This was certainly the case for the Swiss bank UBS.

UBS, which I know well after working there for twelve years, had been concerned that its investment banking division was falling behind US competitors in fixed income. It had also been slow to develop brokerage to hedge funds, an area that had been dominated by Goldman Sachs and Morgan Stanley. A report from management consultants in 2005 on how to close the gap with these competitors led to an ambitious five-year plan aimed at closing the gap. Memos were sent round announcing the plan. The most important message was that the bank would take much more risk. All members of the fixed-income business were asked to seek ways in which the bank could deploy its capital in a more aggressive manner. This was a huge departure for what had been a conservative bank whose main business relied on long-standing wealth management. The bank also had experience of the damage caused by aggressive trading in the losses sustained on its investments in Long Term Capital Management.

Part of the new plan called for the establishment of an internal hedge fund, Dillon Read Capital Management (DRCM), set up to ensure ambitious staff were not lured away by high-paying funds. With its low-cost funding from huge deposits in Swiss francs, the bank assisted the start-up of the fund and also took from the investment bank important business areas in mortgage origination, commercial real estate and credit arbitrage. There had been no special aim to emphasise real estate within the DRCM and the transfer of the investment bank personnel was agreed relatively late in the day. At the time, the business case made sense because demand for mortgage funds was high. There were also few other areas available that offered attractive returns after years of increasingly congested investment in US markets.

As a separate initiative, the investment bank had identified securitised products as the main area in which it was deficient, compared with the three main competitors, together with an increase in the servicing of hedge funds, development of emerging markets and commodities. In all the expanding business areas the bank emphasised its conservative history; presentations included prominent reference to the company’s ‘superior credit rating and balance sheet’.

Losses were first noted within DCRM on subprime positions from 2006 mortgages. The fund had chosen the very last moment to enter the subprime market. Already there were worrying signs of the end. In fact, the mortgage analysts of UBS, regarded as the best in Wall Street, had already published a report asking ‘2006: is this the worst vintage ever?’ A flood of mortgage products had entered the market that enticed borrowers to refinance by suggesting these were the last subprime mortgages available, while mortgage originators competed to sell more and more. Lenders began to increase the size of loans compared with the value of the properties. Documentation requirements were also relaxed considerably.

Subprime securities as ‘central bank money’

At face value, one of the more surprising revelations of UBS’s subprime holdings were those held in the portfolio used to fund the bank’s balance sheet. This function typically invests only in the very best assets. From 2003 onwards, the portfolio managed by this group acquired increasing amounts of mortgage-backed bonds, including subprime. The decision to expand the mortgage-backed holdings was taken because these bonds were accepted by central banks as collateral. The bonds were also readily saleable. In other words, from the point of view of UBS, the mortgage-backed bonds were indistinguishable from ‘central bank money’, the most liquid and valuable medium of exchange. Indeed, the reason UBS held on to its subprime bonds for so long, despite losses, was that they were accepted by central banks.

The use of asset-backed bonds as collateral at central banks, particularly the Fed, was widespread. In the aftermath of the financial crisis, the central banks of the US and Europe became probably the largest repositories of these bonds, as private owners refused to have anything to do with them. UBS had made the decision to switch to asset-backed bonds because the credit rating downgrade of Japan – prompted by its interminable recession and banking crisis – had made the bank’s holdings of Japanese government bonds appear more risky. The holdings of subprime at the Swiss bank were a response to the global nature of the bank, and also to the effects of an earlier banking crisis in Japan.⁵

From February 2007, it became obvious that subprime mortgage-backed securities had become illiquid. Far from offering greater security, they offered none. Holders resorted to estimating expected cashflows and making assumptions about the likelihood of default. The assumptions often tended to be wildly optimistic, sometimes with management encouragement.

UBS, together with all the other affected banks, suffered from assuming recent history indicated normal conditions, whereas the calm economic environment of the Great Moderation was actually a historical aberration.

The problem of data was especially acute for subprime finance because there was no long-term history of such lending, especially through a testing recession. The behaviour of subprime during 2001 was taken as a benchmark even though it was clear that recession had been limited and was accompanied by a great deal of official support for housing, particularly that offered by the Fed’s low interest rates.

Probably because of its late arrival on the scene, UBS became the greatest European casualty of over-enthusiasm for subprime; losing a total of $50 billion in writedowns in the 18 months that followed August 2007. The lessons from the downfall of an ambitious global bank are rooted in its conservatism – it took years to be persuaded to bolster its mortgage holdings – and its global presence. The focus on mortgage-backed securities, especially subprime securitisations, was driven by a desire to catch up with competitors. The bank’s five-year plan of 2005 shows it aimed to increase its presence in commodities, emerging markets and mortgage securitisation on the advice of consultants. It was a move symptomatic of a crowded investment space where investors jostled for returns.

Once the financial crisis began, the writedowns at UBS spread beyond subprime securities to other holdings. It seemed that in each new quarter the portfolio of affected securities widened. From high-risk CDOs the mark-downs spread to better quality mortgages, then to bonds insured by monoline bond insurers. In the end, the losses on subprime at UBS amounted to only half the total writedowns.

UBS was just the most public humiliation among European banks. Others achieved less widespread publicity, but large losses, both in American and in European asset-backed securities, were common. And as the market losses widened, so the number of securities that were refused by other banks increased, leaving only central banks to accept them, unwillingly, to forestall a broader problem. Thus, central bank money was made to accept even more of the toxic investments that no-one else wanted. If the demand returns, then the central banks will be judged to have acted wisely. If the defaults rise, the losses will either cause more bank problems, or the central banks will have to absorb them.

The decision of UBS to increase its asset-backed portfolio because it was accepted at the Fed illustrates the widespread regulatory forbearance of such assets before the crisis. Leverage, risk models, product innovation, balance sheet expansion, off-balance sheet contingent liabilities were all known about and in most cases encouraged by regulators and central banks. There was no attempt to slow or reverse the trend to securitisation and increasing complexity. To have suggested such a course would have invited ridicule.

Commercial banks aim to make a profit in a competitive environment; if the only way they can make a profit is to engage in risky investments, then the alternative for some may be commercial eclipse. Commercial banks are expected to take risks to make money; regulators are expected to stop them, but none of them did.

It was an attitude summed up by Greenspan in the late 1990s: ‘If we wish to foster financial innovation, we must be careful not to impose rules that inhibit it.’⁶ The ECB, in its support for financial integration, followed a similar ‘hands off’ approach. The European Commission encouraged the new products. In 2005, European Commissioner Charlie McCreevy⁷ lauded the benefits of financial innovation, quite rightly:

‘Financial innovation can help to boost returns on savings. It can help to transform wealth into a stream of annuities. We have to put it to work. My job is to provide a European regulatory environment which sustains financial innovation; which allows fund managers to develop new investment propositions; which respond to investor needs. In ways that are sensible.’

It is only after the event that banks – and apparently regulators – can determine what innovations were sensible, and what concentration of innovation is sensible.

How to place ‘value at risk’

One area that certainly needs radical surgery is the risk systems of banks. Many of the techniques used for ‘slicing and dicing’ risk into securities can be criticised. But greater criticism should be reserved for the systems to measure overall risk within banks. Risk systems were supposed to measure and control the balance sheets of banks through data analysis, variance calculations, measurements of correlation, and expected returns. These systems shared many of the characteristics of the securities that would cause problems; they were model-based and reliant on an assumption about the future that lay within certain constrained parameters. They also relied on the quality and the circumstances of the data used. Risk models measure the past moves of a portfolio and calculate the likelihood of losing money in future. They are called ‘value at risk’ (VaR) models. Risk systems based on VaR cost banks money to create and maintain, then proved not only useless, but an incentive to dangerous behaviour. It is probable that the use of VaR systems was responsible for the loss of much more money than the total writedowns on CDOs.

VaR works by analysis of a history of securities’ behaviour and comparing this with other securities. It then works out risk measures for the portfolio run by the bank that indicate the degree of risk being run. VaR attempts to show the probability of a given loss. Many people assume this indicates the maximum loss but that depends on what probability is used and what volatility assumptions are used. On bank trading floors, risk control was alternately a constant source of amusement and annoyance.

There are several dangerous aspects to this system. First, there may be a 99.5 per cent chance of not losing a large amount of money, but the small 0.5 per cent chance of losing an extremely large amount of money may have such severe consequences that it outweighs the remaining 99.5 per cent probability. And so it was for many banks. Their VaR systems suggested the sort of market behaviour that followed the subprime bust was possible only perhaps once in the history of the world. Such implausibly low odds of catastrophe illustrate the limited use of the models.

If the probabilities were wildly wrong, it is easy to see how this was so. VaR systems often used data with a history of five years. This is less history than a typical economic cycle and certainly a great deal less than the long expansionary period of the Great Moderation period, which ran from 1994 to 2007.

The terrorist attacks of September 2001 on the World Trade Center introduced enormous dislocation into the price histories of securities. The way that VaR systems operate concluded that the period that followed the attacks contained greater risk than the period before the attacks; a classic case of shutting the door after the horse has bolted.

There was a more fundamental problem in VaR that stemmed from the way that severe dislocations are seen – such as the 2001 attacks. Because the system used a set period of history, a time comes when the dislocating event drops out of the history. For some banks, including UBS, this occurred in October 2006, five years after the attacks. As the period after the attacks had been the most benign market conditions known, the removal of the attacks caused risk limits to jump overnight. In some cases, risk-taking was able to double. Bearing in mind the bull market in credit was approaching a time when it would have been sensible to reduce risk, it is highly likely that the application of VaR models contributed to the last mad period of investment in risky securities.

High pay policies

Banks’ pay policies have attracted a lot of attention. The UBS report showed that rewards at banks took no account of the low cost of funds offered by the bank, nor of the long-term risks involved. Its employees could buy an asset that yielded only slightly more than the cost of funds and, if invested in large enough quantities, the dollar returns would determine a high reward for the employee. The bank assumed its access to cheap funding would survive financial market disturbance. Similar incentives proved the undoing for many others, including Lehman Brothers, Bear Sterns, Northern Rock and Depfa. The investments themselves would not have looked attractive without access to low-cost funds, so the added skill in many investments was largely irrelevant. As I showed earlier, the provision of these funds often originated in countries such as Switzerland and, more frequently, Japan, where interest rates had been kept at ultra-low levels for several years.

Incentives to engage in riskier securitisations were also high because they offered bigger fees and long-term investments provided upfront bonuses for employees. According to the UBS report, bonus payments ‘for successful and senior IB Fixed Income traders, including those in the businesses holding subprime positions were significant. Essentially, bonuses were measured against gross revenue after personnel costs, with no formal account taken of the quality or sustainability of those earnings.’

Employees were given a free call option on shareholder and taxpayer capital. In the event, they were also given liberty to play with the system of trust that underpins payments. The alternative, losing high quality staff to more amenable competitors, was not acceptable and remains likely to stall any long-term change to the culture of bonuses.

According to figures published by the Securities Industry and Financial Markets Association, the pay costs of broker dealers at companies such as Goldman Sachs, Merrill Lynch and Morgan Stanley in 2003–08 amounted to $550 billion.⁸ Add to that figure the cost of compensation in the banking sector plus mortgage originators and the losses estimated by the IMF for US banks ($1 trillion) through writedowns look roughly equal to the amount of money paid to the workers in the industry. Considering the disruption caused and the long-term damage likely to growth in the rest of the economy, it would have been better value to have paid the money direct to all the brokers and bankers and asked them not to come to work for six years.

The cost to the wider society has been colossal, in lost opportunity, trust and future growth, running well above the cost of writedowns to buyers of ‘toxic assets’. The cost of banking collapse is transferred to the wider society to protect the payment systems. The most transparent measure of society’s cost is the increase in government debt in the countries affected. The US Federal deficit in 2009 is expected to be nearly $2 trillion in 2009 alone or 14 per cent of gross domestic product (GDP), with more debt ahead to ameliorate the damage to the economy. The UK is likely to see government debt expand from 42 per cent of GDP to 75 per cent over the next four years, a total of $1.2 trillion over that period. Spain’s and Ireland’s government debt will expand from 36 per cent and 21 per cent of GDP to over 70 per cent, or about $650 billion.

A still greater loss may be in the trust in a system that seemed to have helped transform the world for the better. A loss of trust in the financial system could undermine support for more obviously beneficial institutions, such as democratic government in less stable countries, and possibly even in the developed world. The political storm that shook Britain over parliamentarians’ expense claims in the spring of 2009 would probably not have shown the venom it did without the backdrop of rising unemployment and falling purchasing power by the British public, both caused by the financial crisis.

Banks writedowns are a loss of Western wealth

The banking system of Europe and America lost up to $2.8 trillion according to the IMF. Greed, concentrated risk in poorly researched products, lack of appropriate data on subprime investors, leverage through financial products such as CDOs, poor judgement, misaligned incentives, competition, lax lending standards. All of these factors played a part.

There is another effect too. The losses were an explicit measure of the loss of competitiveness to Eastern economies, particularly China. Banks work as intermediaries – go-betweens between lenders and borrowers – but the losses of Western banks express, in some sense, the loss of the societies in which they are lodged. There are losses for reserve managers too, if they invested in subprime mortgage products. But most reserve managers did not put a high proportion of their investments in these types of securities; reserve managers have emerged relatively well from the crisis. At one level there has been a transfer of wealth, through the official organs of reserve management policy, from Western countries to those countries following high reserve accumulation policies.

There is further damage to Western wealth creation from the crisis through the probable lower rate of growth in the West that will follow the crisis, plus the lower levels of debt that will be tolerated, or desired, by the population. Both will limit future growth levels.

These effects occur just as many countries in the West face ageing problems in their societies and would like to deliver as much growth as possible. Just when society as a whole would ideally be requesting more credit to pay for pensions and healthcare for its elderly, the amounts used to save the banks will severely limit the spending plans of governments. On the other hand, the amounts to be spent on the ageing population are so colossal that the extent of spending on the banking system pales into insignificance. And perhaps that is one of the more bitter lessons we have to learn; that for all the disruption we have suffered in the financial crisis, it is nothing compared with the sacrifices to come.