In this chapter, the focus is on the areas where my work offers the greatest departure from current practice and thinking. I will show that a reinvention of financial regulation can deliver a financial system that is less prone to crash and can do this without ossifying finance. It will achieve these goals because it is better at managing financial risk across the financial system.

In Chapter 2, I identified the two principal aims of financial regulation as containing systemic risk and protecting consumers protection. In Chapter 3, I argued that, today, understanding how to contain systemic risk is the greater challenge for policy makers. Consumer protection is an equally vital objective of financial regulation. In the past, consumers of finance have been undoubtedly let down by the inadequate regulation of unscrupulous providers. Yet, I believe regulators show greater understanding of what should be done in the area of consumer protection than in the area of systemic risk. Regulators are prioritizing protection of genuinely vulnerable consumers—especially from the deceitful and aggressive selling of financial products. They want to ensure that consumers are free of the perils of conflicts of interest by financial service providers while simultaneously supporting choice and innovation in the industry so that the average consumer can better manage their financial risks and uncertainties.

Exactly where along the continuum of protection and choice the balance should be is of course one of the critical challenges—in finance as it is in other areas. Regulators are unlikely to ever settle the point of balance, as it shifts over time with the ebb and flow of societal sensitivities. This is yet another process that has procyclical and political dimensions. In the backlash after a financial crisis, the line tends to be redrawn on the side of the greater consumer protection of all. Some activities are banned outright and responsibility and liability shifts to providers. In boom times, the line is often redrawn in the other direction¹ with regulators pressured to abolish “financial repression.”² The heated debate that surrounds consumer protection is inevitable. However, it is not an indication that regulators are as much in the wilderness with consumer protection as they are with systemic risk. We will explore consumer protection in greater detail in Chapter 7.

Several measures show that the Global Financial Crisis (GFC) that began in 2007 and reached its nadir after the collapse of Lehman Brothers on September 15, 2008, was the deepest and longest crisis since the Great Crash of 1929 and the ensuing Great Depression of the 1930s. That Depression ushered in the greatest number of structural reforms of the financial sector to take place at one time.³ Indeed, it paved the way for the most internationally agreed vision of collective stability—signed at Mount Washington Hotel in Bretton Woods, New York, in 1944.⁴ There is some evidence that systemic risks have grown steadily following the breakdown in the 1970s of this Bretton Woods system of pegged exchange rates and capital controls. Crises post-Bretton Woods have each seemed deeper or broader than the last. Before the GFC of 2007–9, we experienced the global crisis dubbed the “dot-com debacle” of 1999–2002, marked as one of the most debilitating postwar crises of the developed world. We should also note the devastating regional crises—such as the Asian Financial Crisis of 1997–98 and the earlier European Monetary System (EMS) Crises of 1992–93 and 1995. The EMS crisis was broader, if not deeper, than the preceding Latin American Debt Crisis of 1983–85.

Some argue that the trend of ever-increasing financial crises is actually part of a longer-term cycle. Whenever I hear discussion of trends and cycles, I am reminded of E. H. Carr’s wry observation that the way we view the past, and whether it resembles a trend or cycle, relates to where we are in the present.⁵ When times are good, we tend to see the past as a trend, with things just getting better and better over time. When times are bad, we tend to see cycles and draw the lesson that while things are bad today, they will be likely to turn up later. Our optimism is more enduring than our perspective. I am also struck by Professor A. C. Pigou’s observation in The Veil of Money⁶ of the changing views of money a little less than 100 years ago and their resonance with the last 20 years:

There is no need to draw a conclusion as to whether we are in a short trend or a super cycle to be concerned that the financial system, and financial ­innovation, appears today to be the source of the amplification and spread of economic risks rather than a force for their absorption and mitigation. Financial crises have for too long been a frequent, severe, and contagious phenomenon of our lives. Reasonable observers cannot confidently state that financial regulation is on the right track and all that is required is nudging bank capital adequacy requirements up a couple percentage points, tighten up exemptions to the requirements and maybe some additional tweaking.⁷

In fairness, many in the developing world have been making similar ­observations for some time as they watched “emerging markets” experience increasingly deep and contagious crises. In the post-Bretton Woods era, the sequence of international emerging market crises started with the Latin American Debt Crisis followed by the Tequila Crisis,⁸ before moving on to the Asian Financial Crisis, with a host of deep but less international crises in between.⁹ During these emerging market crises, the default response of wealthy countries declared that the system was not a problem. The fault lay with these poor countries who lacked the necessary fiscal and political discipline to prevent themselves from being so badly hit. The inevitably bitter medicine they offered up directly, or through agencies they influenced, like the IMF, was real wage cuts through austerity and devaluation and mandatory opening up to international banking.¹⁰ Reforming the international system and ring-fencing local bank capital only rose to the top of the agenda, along with an acknowledgement of the adverse implications of pursuing austerity simultaneously, when the GFC engulfed the rich nations. What we consider to be self-evidently right for others is often not what we consider right for ourselves.

No one said life was fair. However, it is worth noting the connection between the international political economy in the twenty years prior to the GFC and aspects of that crisis. Having been severely burned by crises in the 1980s and 1990, and feeling that their concerns and complaints about the working of the international financial system went unheeded, ­emerging markets sought ways to reduce their dependency on the ­international financial system and its institutions. Faced with improving trade positions as a result of exchange-rate depreciations during the Asian crisis and increased ­consumption in rich countries, they channeled surpluses into ­rainy-day reserves and funds. An initially reinforcing, but ultimately unstable feedback loop emerged.¹¹ Reflecting the superior liquidity and dominant role of the US dollar in international financial transactions, the national savings of these countries were ­concentrated in US Government instruments. This ready ­buying of government debt loosened the reins of fiscal discipline in the United States, facilitating fiscal adventures such as the Bush Tax Cut and wars in Iraq and Afghanistan. Low interest rates on government debt created a demand for private investment and consumption that spilled over to other countries as reflected in a wide US account deficit.

Stronger spending and growth in the world’s largest economies added to trade surpluses and international reserves in the main emerging-market ­exporters. Every economic textbook and macromodel states that a large current account deficit, at a time of low unemployment, is a sign of excessive demand that should be curtailed by tighter fiscal, monetary, and regulatory policy. To do otherwise could lead to inflation or bust. But that is unpopular and politicians have elections to fight. Moreover the inflation was coming in the form of asset prices rather than items in the consumer price index. It was far better for politicians, central bankers, and investors in these rising asset prices to assert that this time is different. It was better to worry about ­deflation in Japan and to view the large current account deficits in places like the United States and, United Kingdom as signs of structural investment booms driven by the superiority of American and English bankers and their financial system over which policy makers were powerless. The resulting asset market bubble inevitably burst.¹²

It was easier for policy makers in the spending countries to focus on the ­conduct of other countires. To do otherwise would have forced them to ­examine their failure to conduct adequately tight fiscal, monetary, and ­regulatory policies. In reality, global reserve increases played an aggravating part in the story of the past 20 years. However, they were more a symptom of excessive borrowing than a cause of the crisis.¹³ Greater faith in an international lender of last resort that does not require troubled countries to throw out the baby with the bathwater in return for emergency assistance that might have provided greater diversity in the investment of savings. This, in turn, might have tempered the boom that led to the bust.¹⁴ It is all a huge “if.” Foreigners are always easier to blame but the scale of the boom meant that the main protagonists had to be local to the world’s largest economies rather than on the periphery. Perhaps the simple point is that there is an international dimension to financial regulation that we have so far neglected. We will consider this again in Chapter 14, but first let us focus on what must be done on the home front.

Financial Regulation and Procyclicality

In Chapters 3 and 4, I argued that while the financial sector may have more than its fair share of crooks, financial crashes are not caused by the antics of a few.¹⁵ The antecedent of financial crises is often some prior event or ­reasonable belief—frequently centered on the arrival of a new technology that initially drives a collective underestimation of risk and boom. Market participants are drawn into areas widely perceived to be safe. Confidence and conviction is as widespread as it is intense creating a haughty spirit, often ­celebrated in popular culture making everyone feel they can make a fortune in the ­markets.¹⁶ When the safe turns out to be risky, widespread shock, dislocation, and despair ensues. In short, the problem is a collective miscalculation of risk that sustains the financial boom and sows the seeds for the inevitable bust to come.

The “procyclicality” of the financial cycle is why the Basel II shift toward more market-sensitive risk measurements, such as banks’ own internal risk ­models or public credit ratings, was fundamentally flawed. It was doubling up on ­danger inevitably leading to disaster. Credit availability is the major determinant of asset prices, especially in housing markets.¹⁷ As a boom develops, rising valuations and falling assessments of risk push down Basel II–type capital requirements, incentivizing banks to lend more by pushing up asset valuations. The history of rising asset prices pushes down perceived risks.¹⁸ Those banks that don’t increase their lending in the boom lose market share and face ­slipping stock prices.¹⁹ At the top of the boom, when banks should be most cautious, the constraining forces of bank regulation and capital requirements are weak,²⁰ with reported capital at historically high ratios to risk and flattered by inflated valuations and deflated risk assessments.

Equally troubling is when this process shifts into reverse during a bust. Collapsing valuations and rising risk assessments push up capital requirements, reducing lending and borrowing—even though the credit mistakes were made previously. This deleveraging shrinks asset valuations further, reducing lending yet again and deepening the recession. Banking regulation should act to check the financial cycle but the switch to a so-called risk-sensitive approach in Basel II inevitably, and predictably, amplified the crisis.²¹

While many may find it self-evident that you cannot prevent market failures with the greater use of market prices,²² there are those whose Panglossian view of markets leads them to counter that the underlying problem of misestimation of risks is really an issue of misinformation. If only market participants had better information they would make the best-possible judgment. This is reinforced by the frequent cry of indignant participants post-crash, exclaiming that if only they had been forearmed with such-and-such information, they would not have joined the party the night before.

It was only after the GFC that the mortgage originators’ lack of due diligence regarding the income of candidates for subprime mortgages was fully revealed. The true extent of short-term external borrowing by Asian corporates only became known after the Asian Financial Crisis of 1997–99 had taken hold. The level of interlocking directorships and their contribution to conflicts of interest, insider dealing, and panic-inducing uncertainty for the outsiders only became clear as the Japanese property bubble burst in the 1980s. The same was true in the United States during the Panic of 1907.²³ The idea that market failure is caused by inadequate information is behind the near-religious faith placed in transparency and in moves made toward greater disclosures, mark-to-market accounting, the encouragement of credit bureaus, credit ratings, and more.

In general, the more information, the better, and the disposition of greater transparency should only be tempered by specific cause.²⁴ Information asymmetries are at the heart of consumer protection issues. More extensive disclosure of potential conflicts of interest and greater transparency are important contributions to consumer protection from unfair practices. They are also devices to preserve the integrity of, and confidence in, the financial system. However, my close-up experience of several financial booms leaves me unconvinced that it is a lack of disclosure that is the root cause of systemic risks. Enhanced transparency would not have made much difference. This is an inconvenient notion for those on the post-crisis hunt for that smoking gun of who knew what and when they knew it.

Generally in today’s world, information is more widely disseminated and more quickly revealed than ever before. Yet this explosion of information has not coincided with weaker financial crises. The opposite has occurred. As we have discussed, booms are caused by a collective belief that a new technology will transform the future. Invariably, this comes with the assertion that old relationships lack relevance and there is a new paradigm so investors should zero in on the new metrics. If an investor during the dot-com bubble of 1997–99, insisted on finding out more about the profit and loss of one of the instant multimillion-dollar start-ups—Pets.com, e-Toys.com, or Webvan.com²⁵—she would have been laughed out of town as a dim-witted dinosaur. In a voice reserved for teaching young children, it would be explained that the new metric is obvious, only the stupid fail to understand the value of the number of “eyeballs”²⁶ and those not on board with the new metrics, will miss out on huge gains. This collective delusion was prolonged—over three years in the case of the dot-com bubble—because the starry-eyed evangelists appeared right, hip, and rich. The heretics in hair shirts appeared wrong, staid, and more impoverished with every passing year. Avoiding embarrassment in front of one’s peers is a powerful insulation that protects bubbles from bursting early.

One of the defining aspects of the GFC was that the ­underlying technological revolution was the least tangible of all previous ­revolutions. Even tulip mania had tulips.²⁷ The notion this time around was that the cleverness of bankers and the sophistication of their computer models allowed risks to be better managed.²⁸ Hubris is unbridled in financial booms. Added to the belief in better risk management was a traditional housing market boom like those that underscore most financial booms. Many of the new risk management products that ended up exploding were instruments aimed at diversifying real estate risk. Although they are frequent, housing booms also come with a host of reasons why traditional borrowing limits no longer apply and why property prices in a particular area must keep on outpacing the rise of incomes, rents, GDP, reservoir capacity, bridge capacity or car parking spaces. In London, the stories tend to hover around planning or other genuine restrictions on new housing supply or the appeal of London to the world’s latest set of tax-averse, newly-enriched tycoons.²⁹

During the GFC, little attention was paid to income ­verification because property prices were expected to rise strongly. If ­property prices are expected to rise by 20 percent per annum, a loan of 99 percent of the value of the property, with the slimmest of equity protection for the lender at the outset, would have a buffer of 40 percent in a mere two years. By 2005 and 2006, almost no one had experienced a 40 percent decline in property prices for almost two decades, long beyond the memory of the fresh-faced borrowers, real-estate agents, and lenders. Lenders easily convinced themselves that, if after a couple years borrowers could not afford repayments, they would be able to repossess the property, resell it, and still make a full recovery of their loan, interest and costs. This held so as long as the property value had not fallen by more than a seemingly impossible percentage. It is a short jump from this thinking to adjustable rate mortgages (ARMs), where repayments start off low before rising after two years, or even Northern Rock’s Tomorrow Mortgages that were 120 percent of the value of the property being mortgaged.

Upfront mortgages and real estate commissions created incentives for aggressive selling that must be addressed. However, booms and busts exist on a scale too huge to be easily reduced as the sole work of greedy commission agents. Booms are aided and abetted by the average person who is buoyed by a belief that it all makes sound financial sense. What appears obviously foolish in the cold light of hindsight seemed to almost all as clever the night before. Financial regulation has to act as a countervailing force to these long swings of market optimism. It cannot rely on statistical measures of current risk or market estimates of future risk. To do so would make financial regulation procyclical, in danger of augmenting market failures rather than dampening it.

This was the grave error of Basel II. Many conscientious souls, who spent a major part of their career bringing Basel II to fruition, understandably find it tough to admit to the fundamental flaws of the exercise. Some argue that it was not Basel II at fault but rather mark-to-market accounting and bankers’ pay. They remind us that Basel II was not yet in full force in the boom years leading up to the GFC.³⁰ While mark-to-market accounting and bankers’ pay did play an adverse role—something we ­discuss in Chapters 8 and 10—this argument is disingenuous. The philosophy of Basel II was about a more market-sensitive approach—of which the adoption of mark-to-market accounting and market-sensitive risk management models were key elements. Tied up with this philosophy was the notion of enhanced reliance on market discipline that would make intervening in market valuations, market risk assessments, bank business models, and market-determined pay both unnecessary and distortionary.³¹ Moreover, this philosophy had triumphed in Basel long before the implementation of Basel II.³² It is found in the amendments to the implementation of Basel I as early as the 1996 amendment on market risks, which introduced the idea of banks using their own internal risk models based around short-term developments in market prices.³³

Macroprudential Responses

A common response to the crisis and failure of bank regulation has been the establishment of new systemic risk committees. Consisting of the wise and well-connected, they are tasked with determining if capital adequacy requirements should be raised or left untouched.³⁴ Charles Goodhart and I have been among those pushing hard for a more macroprudential approach and support steps being made in this direction. Of course, there are quite a few challenges that come with this emerging approach to macroprudential policy. It is another inconvenient truth of financial crises that many central bankers had the discretion to tighten lending limits during the boom times and chose not to exercise that discretion.³⁵ The collective inability of humanity to escape the preoccupations of the present is not easily overcome by anointing a special few to do so. A crucial lesson of the crisis is that we need more rules to rein in credit growth during a boom. Greater discretion is not required. A rule based on bank profitability or credit growth could determine when capital requirements are raised during a boom or relaxed during periods of financial stress.³⁶

I favor rule-based countercyclical capital requirements but they are not without complication. An ill-designed approach could lead banks to respond by concentrating their lending in the booming sector and away from other sectors as the booming sectors are best able to turn a profit even after paying higher capital charges. Raised capital requirements could be confined to the booming sectors or banks made to lower loan-to-value ratios to ­borrowers in those sectors. This might appear inelegant and ad hoc (banks may try to get around the definitions of the curbed sector). Banking regulation is often ad hoc. Bear in mind that it would be a temporary measure, regularly updated in response to the shifting borders of the booming sector. Arguably, a macroprudential policy is more about concentrations than aggregates of ­lending. Capital requirements can rise with increased concentrations of risk on a bank’s balance sheet thereby encouraging diversification as well as ­creating a capital buffer.³⁷ This is more elegant than simple countercyclical shifts in capital requirements and the long-term covariance of credit risks should be managed. However, on its own this does not solve the underlying problem. Statistical correlations of risk are, like almost everything else, procyclical. The same world appears to be a diversified and liquid place in a boom and a concentrated, illiquid one in a crash.³⁸

And the challenges do not end there. There must be a way of reducing capital requirements in the postboom era, allowing capital to be released and buffers used up. Lowering capital requirements just as people are recognizing that the world is a riskier place than they thought is just as difficult to achieve as raising them beforehand when the world looks to be a safer place. The politics of discretionary countercyclical action is harder than the economics. This explains why it is talked about rather than implemented. It emphasises the need for a more rule-based approach.³⁹ The rule must be seen as inviolable so that in practice it only rarely gets set aside.

Even more challenging is that we would be raising and lowering ratios of capital that are themselves based on a fundamentally flawed calculation. The current system of bank regulation requires capital to be set aside against risk-weighted assets. The more risky the assets, the higher the weight and the more capital that must be set aside. This appears elegant and logical. Who is against “risk sensitivity”? The statisticians and bank supervisors have, through long use, established this rule as a norm. But that doesn’t make it right. If a lender believes a loan is risky—loans are assets for a lender—under ­normal banking practice this is addressed by requiring additional guarantees or ­collateral. The lender is not going to award the loan on the same terms as others and just put aside more reserves. They might also lower the loan-to-value ratio at which they will lend and charge a higher rate of interest. Consequently, the expected return, including what is recoverable if the loan fails, already takes into account the risks of lending. Capital is needed, then, not to ensure the safety of banks if a risky loan fails (already covered by higher collateral and reserves funded by higher interest rates charged on risky loans) but to protect the bank from a systematic underestimation of risks that overwhelms these reserves.

Basing the capital requirement on the original estimation of the riskiness of a loan implies that a bank is more likely to underestimate the risk of a loan that it already considers risky. The degree of error in risk estimation is proportional with the original estimation of risk. This is unlikely to be the case. Banks do not make loans that at the outset they consider risky, and when they do these risks are mitigated with collateral requirements. The greatest danger to a bank’s ability to survive is if a large proportion of loans that it considers safe turn out to be risky. But under the risk-sensitive approach, banks carry least capital against their greatest danger: previously considered safe loans turning risky. This is one of the fundamental flaws of the so-called risk-sensitive approach of Base

In Sending the Herd off the Cliff Edge,⁴⁰ I identified another fundamental flaw. By setting capital adequacy requirements against risk estimates, banks are encouraged to concentrate lending into areas that statistical measures of the past suggest are safe yet offer slightly better returns than other safe areas. This concentration pushes the price of these assets up to the point where they become overvalued and vulnerable to a mass exit, turning what was safe into something risky. In a world of uncertainty, common data sets and common risk models, the observation of safety creates risk.⁴¹ These are interesting areas that I would love to discuss more but this book is confined to seeking solutions. It is to these solutions that we now turn.

One nod to the challenges previously noted is the reintroduction of the ­leverage ratio alongside risk-sensitive capital adequacy requirements. This limits the amount of lending in aggregate—across seemingly safe and risky borrowers—relative to capital. It also completely restricts capital to loss-absorbing assets like cash reserves, shareholder funds, and the like. Of the second-best solutions to complicated problems this belt-and-braces approach has merit.⁴²

Another nod to the challenges of the risk-weighted approach to estimating capital adequacy is regulators’ rising use of “stress tests.” Regulators are asking banks to assess the effect of certain specific scenarios—like the fall of the price of houses by 40 percent—on their capital buffers. If a bank reports that its capital levels fall below the regulators’ threshold level then capital levels must be increased. This is more useful conceptually than practically. Banks make part of their profit through maturity transformation—borrowing short and lending long—so will fail in any scenario in which liquidity disappears for a long time. Revealing this potential instability could itself set off a bank run. Finance is full of such troublesome dynamics. Consequently, the stress tests applied tend to be narrowly short of the kind that all banks would fail. At the edge of the permissible, these stress tests end up capturing the degree to which a bank’s funding is not dependent on external liquidity conditions. It is better to measure and influence that directly rather than indirectly. Reducing this funding risk should be the touchstone of endeavors to keep the financial system safe.⁴³ With the current approach to macroprudential policy fixated on capital, we will be left with a system that is either overly dependent on error-prone measurements of risks⁴⁴ or demands ever more capital. The focus should be on risk managing the system, a concept already alluded to and which we now give more detailed consideration.

Risk Managing the Financial System

The GFC settled the debate—for now—on the need for a macroprudential dimension to regulatory policy. However, it may have done this without clearly establishing common ground on the meaning of ­“systemic risk” and how best to manage its various causes. In the absence of ­common understanding and a common framework, macroprudential regulation is in danger of reverting to an enhanced microprudential exercise,⁴⁵ with ­macroprudential merely meaning that we have a wider set of risks to consider and to provide capital against. As the economy slows, perhaps under the ­burden of more unproductive capital, the collective amount of risk rises, requiring more capital.⁴⁶ At times, the new capital adequacy regime appears more procyclical, and not contracyclical as intended. Many bankers complain that lending today, at a time of record-low interest rates and depressed valuations, is being constrained by regulation.⁴⁷

In financial booms, when liquidity is plentiful, there is a belief that all risks can be extracted and measured by the volatility and return of their traded price and totted up into a single measure of risk. Banks could then increase or lower their activity to match this risk with their risk appetite. This vision of risk is convenient for the statistical models used by banks and, partly through their influence, colored the approach to risk taken by the Basel bank regulators. But this idea is wholly incorrect. There are many different types of risk and their difference means they cannot be tallied together and the aggregate amount notched up or down to match a level of capital or risk appetite. From a financial regulator’s perspective, the main risks with systemic implications are liquidity, credit, and market risks. These risks are different from each other, not because we give them different names and describe them differently, as we might describe different colors, but because they must be hedged differently. It is meaningless to add them together.

The liquidity risk of an asset is the risk that, if forced to sell the asset tomorrow, a deeply discounted price would be accepted to entice an unwilling buyer as compared to the price that might be achieved if there was more time to find a buyer. Hedging liquidity risk does not come from owning a diverse range of equally illiquid assets but by having the time to sell—perhaps through long-term funding or long-term liabilities. I live in a Californian-styled glass and steel house that happens to sit in a leafy, Victorian suburb of London. Many people find the house fascinating and part of my neighbors’ Sunday ritual is to walk their friends past the house, look at it, point at some aspect and talk about it. But with its walls of glass, most cannot envisage themselves actually living there. If I needed a quick sale I would have to accept a larger discount than others who may live in more conventional homes. It is not that the market price is lower but that there is not one singular market price. To achieve the highest price requires more time with uncommon and illiquid assets. Somewhere in Berlin is an avant-garde couple who want my house but I have to wait for them to visit my London suburb before I can sell it. This liquidity risk would be major had I funded the house purchase with a three-month loan that I rolled over every three months, hoping that at the point of each of these rollovers, the loan company did not have a change of heart about the housing market. It is much less of a risk with a 30-year mortgage. The liquidity risk of an asset can be hedged by having long-term funding.

The credit risk of an investment is the risk that a borrower will default on its payments of interest and or principal. Credit risk is not hedged by having long-term funding because it rises the more time there is for a default or some other unforeseen disaster to take place. If, at their birth in 2000, I gave my twin boys a gift-wrapped bond issued by one of the high-flying companies of the day, like Jarvis Construction Company, without the possibility of selling it until they were 18, I would have given them something that is a cross between an asset and a time bomb. The likelihood that something cataclysmic will go wrong at any time during a period of 18 years is bigger than the likelihood that something will go wrong over the next five, or one year, or one week, or one day.

If I could sell the bond at any time within the 18 years, it would sharply reduce the credit risk. I might have sold it before the Potters Bar rail accident in 2002 that killed seven people at a time when Jarvis had the rail maintenance contract. Unnerved, I might have sold the bond shortly afterward and avoided the second rail accident one year later that harmed the company’s reputation. I may have sold it when the company dived headlong into debt-backed infrastructure projects in 2004. I could have sold it a few years later just prior to the GFC and avoided the bonds becoming worthless in 2010 as the company collapsed under the weight of too much debt in an era of reluctant lenders. In an uncertain world time increases credit risk.⁴⁸ The way to hedge credit risk is to spread exposure across a large number of diverse credits, where the risk of one borrower not being able to afford interest payments is uncorrelated with another. Combining railway construction and maintenance bonds or loans with the credit risk of renewable energy firms, for instance, would reduce aggregate credit risk compared to a portfolio of lending to only one of these risks.

The market risk of an asset is the risk that its long-run market price will fall—like the risk that oil prices will fall on the announcement of a new oil ­discovery. To hedge market risks requires either diversifying risks across time—by ­having long-term funding of those assets to eliminate the need for an inopportune time to sell—or across assets or some combination of these two activities.

Risk-Absorptive Capacity

Arrayed against these different risks are individuals and firms with characteristics that give them intrinsic capacities for naturally hedging different risk types. A long-term pension or life insurance fund, for instance, receiving a regular set of contributions or premiums for 20-odd years in return for a lump sum payment, has a capacity to absorb liquidity risks, though no particular ability to spread credit risks.

The right place for a risk is where there is a capacity to absorb that type of risk. One critical advantage of placing risk where it can be best absorbed if it erupts is that it becomes less dependent on its size being measured correctly. Incorrect measurement of risk is at the heart of financial crises and something almost all financial market participants succumb to through the economic cycle. If risks in the financial system are in the wrong place, then no reasonable amount of capital will save the system.

Not only does placing risk where there is a natural capacity to absorb it improve the resilience of the financial system, it also makes good investment sense. Imagine two pots of identical amounts of savings at the beginning. One is invested in a portfolio of blue-chip equities and another in a portfolio of government bonds. We return 20 years later. The likelihood that the pot invested in equity markets has outperformed the pot in bonds is high.⁴⁹ But this higher long-run return comes with substantially greater short-run volatility than the investment in bonds. If there is capacity to ignore the short-run volatility without suffering from, or having to manage that risk, this is ­nirvana. This might be because, for example, it is a pension maturing in 20 years, allowing for liquidity-risk and market-risk premiums which give the equity portfolio its higher long-run returns. However, not all savers have that luxury hence these risk premiums exist.

In a situation that dictates assets be converted into cash at short notice for a medical emergency, an investment or allocation to bonds or cash would be necessary. A bank that is funded by depositors who can withdraw their funds without notice does not have the capacity to lock its funds away without access and therefore cannot earn the liquidity premium. What it does have is the capacity to absorb individual credit risk by virtue of its superior ability to spread credit risks across thousands of borrowers. Some may fail to make payments on their debts but most will honor their debts and pay interest. It can earn the credit risk premium. The safest return for investors is the return available for choosing a type of risk that they have a superior capacity to absorb.

If all the risks in the financial system were placed where there was the best capacity to absorb them, then both the financial system and individual borrowers and savers would be at a Pareto-optimal combination of safety and return. This means that more safety is possible but only by reducing returns and more return is possible but only by reducing safety. Crucially, this return is higher than what is sometimes called the risk-free rate of return, i.e. the return on assets with the greatest liquidity and credit quality and least market risk, such as short-dated US government bonds. This is therefore not a proposal for narrow or repressed forms of finance. The return society can safely earn is greater than that approach would allow. It is the risk-free return, plus the liquidity premium earned by those with a capacity to earn it, plus the credit risk premiums earned by those with a capacity to earn them, plus the market risk earned by those with a capacity to earn it.

We need capital for those junctures with a mismatch between risk capacity and risks being taken for a variety of possible reasons. This capital requirement would not only act as an additional, if expensive, loss absorber, but its expense would exploit the power of markets to pursue lower costs to good use. It will incentivize good investment practice as investors first match the risks they take with their risk capacity. Socially positive innovation is incentivized as investors seek to maximize all of their risk-absorbing capacities. Furthermore, this promotes systemically strengthening risk transfers as those with credit risks but little absorptive capacity for such risk trade them for other risks for which they do have absorptive capacity.

These risk transfers will play the greatest role in reducing systemic risk It is critical therefore that this capital regime applies equally across the widest definition of the financial sector—from insurance firms to banks to hedge funds. In the next chapter we will examine in greater detail how this approach would work from the perspective of the non-bank financial sector, like insurance and pensions.

Note how different these risk transfers would be compared to the systemically dangerous transfers that contributed to the GFC and which made ring-fencing parts of the financial system seem attractive. Risk transfers are motivated by a firm’s need to reduce capital requirements. Capital that is available for absorbing losses is capital that is not earning returns making it expensive to hold. If capital requirements are based on mismatches, risks will be driven to where they are best matched. This was not the scenario prior to the GFC. As we have discussed earlier, banks were required under Basel1 and 11 to hold capital for the credit risks they took but not their liquidity risks. This incentivized them to shed credit risk—the one risk they have superior capacity to absorb. It also made them take on liquidity risk which is the very risk they have little capacity to hold but for which they had no capital charge for holding.

Little that happened during the boom was not incentivized by the regulatory regime. Banks sold their liquid credit risks to pension funds and insurance companies. Sometimes this occurred directly or via special purpose vehicles (SPV). Matters were made worse because banks offered these SPVs a liquidity backstop.

The spread of mark-to-market accounting to corporate pension schemes and other developments made it tough for these funds and other long-term savings institutions to hold illiquid assets, which, by their very nature cannot be priced accurately at a high frequency. This is discussed more intensely in the next chapter and in Chapter 8. For now we note that these long-term savings institutions sold banks the illiquid assets they had a superior ­capacity to hold and purchased from the banks liquid credit risks that they had no particular advantage in holding. They paid for liquidity they did not need and they did so through lower returns. Risk transfers amplified the GFC because they went in the exact opposite direction to the matching of risks to risk capacity.

How can capital requirements be based on risk mismatches? Risks need to be separated on the basis of how they can be hedged. Most risks will be separated into liquidity, market, and credit risks. In the case of liquidity risks, pools of illiquid assets can be matched with pools of funding with different maturities. For instance, if an institution has raised funds through a 12-month time deposit,⁵⁰ this funding could be matched against assets that can easily be turned into cash within 12 months. This can include, for instance, bond ­instruments with a less than 12-month maturity. Some instruments with longer maturities could be classified as being effectively more liquid, perhaps because they are backed with liquid collateral or are considered a safe asset in times of stress—such as long-term government bonds. Bank supervisors would have to approve the classifications.

Risk-absorbing capital would have to be furnished against the liquidity risk of less-liquid equities, bonds and loans, unless there are additional longer-term pools of funding. Matching the liquidity of pools of assets with pools of funding is easier than trying to do the task asset by asset. It also provides for changes in funding and will help to minimize and isolate those liquidity risks that ­cannot be easily matched with funding. These unmatched risks could either be sold to someone with more longer-term funding (for whom the assets would be more valuable) or capital could be put up against them.

Within the banking sector, regulators have taken one step toward this goal through the introduction of the net stable funding ratio. The goal is to match a bank’s long-term assets with an equal or greater amount of long-term or stable funding. This is one of the new regulations that bankers are most ­vehemently opposed to, yet from a systemic risk point of view it is one of the most important pieces of new regulation. Regulators must not retreat. Of course banks can only effectively shed liquidity risks if there is a more appropriate entity to hold these risks. It is critical that those with a capacity to hold liquidity risk are rewarded for doing so by basing their capital requirements on unmatched rather than aggregate risks. Yet, under the noses of the new systemic risk committees, the new regulation of insurance and long-term savings institutions looks set to achieve the opposite, through an emphasis on short-term valuation and risk that makes little economic sense for them.

The wrong formulation and proposal of the Solvency II regulation for life insurance companies and pension funds might force the natural holders of the financial system’s liquidity risk to try and avoid it. We take up this issue in greater detail in the next chapter. It is worth noting here that forcing long-term institutions to behave like short-term ones will be the biggest contributor to systemic risk since the original version of Basel II. It is a betrayal of moves toward a more macroprudential approach to ­regulation. Macroprudential should partly mean a realization by regulators that the financial system’s resilience is about where risks reside across the financial and economic system as a whole and not just at banks.

In the case of credit risk, capital should be put up against the degree of ­long-run diversity of credit risks. A higher concentration would require higher capital, rewarding those with more diversified lending and automatically raising ­capital requirements as the boom lifts asset prices while it drives up ­covariances. Regulators could adjust the amount of capital to be held against liquidity and credit risk so that we begin at the same current level of capital but with completely changed incentives, ones that could lead to lower capital but safer lending.

Conclusion

The central problem we are facing is a collective underestimation of risks that leads to the euphoria of the boom and the despair of the inevitable bust. This underestimation is often triggered by a new discovery that appears set to change the world as we know it. For a while, this perception becomes reality as asset prices strengthen and a rising tide lifts all boats. Booms are the triumph of optimists. The secondary issue is that resulting from this ­process market-based measurements are procyclical. At the top of the boom, seconds before anarchy descends, market measures of risk are near record lows, short-term correlations appear low and bank-capital ratios therefore appear to be near record highs. Risk-sensitivity is a bankrupt idea.

We can make the system safe in a way that does not rely on an ability to ­measure the size of risks. Risk sensitivity must be replaced with the notion of risk-absorptive capacity. There are different risks which individuals and firms have varying capacities to absorb. If I place a risk where it can be safely absorbed were it to erupt then there is no need to be accurate at measuring that risk. Capital requirements should be levied on the mismatch between risks held and the capacity to hold those risks. This will provide a buffer to absorb risks and it will incentivize risks to go to where they can be best absorbed and most safely held. The systemic benefits will only be fully realized if this approach is extended across the entire financial system and not limited to banks.

It would lead to systemically strengthening flows of credit risks to banks from life insurers and pension funds and liquidity risks from banks to life insurers and pension funds. This is the polar opposite of what occurred prior to the GFC. The increasing reliance on measures of risk for markets and valuation in banking regulation in the decade prior to the GFC made a procyclical financial system even more procyclical. Financial regulation must act as a counterveiling force to the financial cycle. It cannot hinge on contemporaneous measures of risk and value. Sanity is not statistical.⁵¹

___________________

¹In 1996, as the dot-com euphoria was getting going, the US Congress passed legislation to remove some restrictions on retail access to hedge funds. In retrospect, this would have been the time to tighten them.

²Financial repression includes such policies as maximum loan rates or minimum deposit rates. These policies are designed to protect vulnerable consumers from usury. However, since they stop the market for loans or funds clearing at whatever price they clear at, they do reduce the total possibilities available to all borrowers and savers.

³In the United States, the 1929 Great Crash and subsequent Great Depression of the 1930s gave birth to the Banking Act of 1933 (more commonly known as the Glass-Steagall Act). The bill was designed “to provide for the safer and more effective use of the assets of banks . . . to prevent the undue diversion of funds into speculative operations and for other purposes.” It separated commercial and investment banking. The crisis also brought into fruition the Securities Act of 1933 and the Federal Deposit Insurance Corporation.

⁴See in general, Eric Helleiner, Forgotten Foundations of Bretton Woods: International Development and the Making of the Postwar Order (Ithaca, NY: Cornell University Press, 2014).

⁵Mr. Carr is not without his critics, most notably, Sir Geoffrey Elton who responded with The Practice of History (New York: Cromwell, 1968). See Edward Hallett Carr, What Is History? (Cambridge, UK: Cambridge University Press, 1961).

⁶Arthur C. Pigou, The Veil of Money (London: Macmillan, 1949).

⁷Besides the introduction of the long-term stable funding ratios, which are being delayed anyway, much of Revised Basel II (more popularly known as Basel III) could be characterized in this way.

⁸The “Tequila Crisis” of 1994–95 centered on Mexico and the pressure on the exchange rate band—relating to a prior borrowing binge—that “broke” at the end of 1994. The cumulative 55 percent devaluation of the Mexican peso had adverse consequences for its Latin trading partners and competitors.

⁹See Stephany Griffith-Jones, Ricardo Gottschalk, and Jacques Cailloux, eds., International Capital Flows in Calm and Turbulent Times: The Need for New International Architecture (Ann Arbor: University of Michigan Press, 2003).

¹⁰This is how the IMF acquired its moniker, “It’s Mostly Fiscal,” and why one cannot underestimate the degree of resentment many in developing countries feel toward the organization.

¹¹This international payment cycle was dubbed “Bretton Woods II” by Michael Dooley, David Folkerts-Landau, and Peter Garber in “An Essay on the Revived Bretton Woods System” (Working Paper 9971, Cambridge, MA: National Bureau of Economic Research, September 2003).

¹²See Claudio Borio and William White, Whither Monetary and Financial Stability: The Implications of Evolving Policy Regimes (Jackson Hole, MI: Federal Reserve Bank of Kansas City, 2003).

¹³See Claudio Borio and Piti Disyalat, Global Imbalances and The Financial Crisis: Link or No Link? (Working Paper 346, Basel: Bank for International Settlements, May 2011).

¹⁴See Stephany Griffith-Jones, Jose-Antonio Ocampo, and Joseph Stiglitz, eds., Time for a Visible Hand: Lessons from the 2008 World Financial Crisis (Oxford, UK: Oxford University Press, 2010).

¹⁵It doesn’t always look that way, as financial booms and the greed they unleash increase the number of swindlers, the size of the swindles, and the propensity of the ordinary public to be swindled.

¹⁶In a story that has been distorted by time and moment, Joe Kennedy, father of President Kennedy, claims to have sold his stocks in the winter of 1928, less than a year before the Great Crash, after his shoe-shine boy told him to buy Hindenburg shares—makers of the Zeppelin. “You know it’s time to sell when shoeshine boys give you stock tips. This bull market is over,” Quoted in Ronald Kessler’s “The Sins of the Father: Joseph P Kennedy and the Dynasty He Founded”, (Warner Books, March 1997).

¹⁷See Table 1, in Fabio Panetta’s remarks to the De Nederlandsche Bank special conference, titled “On the Special Role of Macroprudential Policy in the Euro Area,” www.bancaditalia.it/interventi/intaltri_mdir/en_panetta_10062014.pdf, June 10, 2014.

¹⁸Simply put, banks are required to put aside capital—approximately 8 percent of risk-weighted assets. If reported risks fall, the amount of capital required falls. Invariably what happens in a boom is that reported risks fall at the same time as assets grow, so that the dollar amount of capital may well rise to record levels. What we observe precrisis is a rising leverage ratio (total assets to bank equity) and postcrisis, when risk is reestimated, is that capital has fallen to dangerous levels.

¹⁹This is what former CEO of Citibank Chuck Prince was getting at in his ill-fated, but prescient remarks to the Financial Times on July 6, 2007 that “When the music is playing you have to get up and dance.” The imperative was reinforced by stock-related compensation for bank managers. Unfortunately for Mr. Prince the music had stopped but he was still dancing.

²⁰See Charles Goodhart, “Procyclicality and Financial Regulation,” Establidad Financiera 16 (Banco de España, 2012).

²¹An increasing number of people claim to have correctly predicted the crisis. But predicting that something in the economy and society is amiss and predicting the inevitability of a banking crisis are different. Contrary to public opinion, there were a number of people who identified the direct causal link between the increasing market sensitivity of banking regulation and a future banking crisis. The problem was that few wanted to hear it. See: (1) Jón Danielsson, Paul Embrechts, Charles Goodhart, Con Keating, Felix Muennich, Olivier Renault, and Hyun Song Shin, Financial Markets Group, LSE, “An Academic Response to Basel II” (Special Paper 130, Financial Markets Group, May 2001); and (2) Avinash Persaud, “Sending the Herd off the Cliff Edge,” World Economics 1, no. 4 (2000), pp. 15–26.

²²See Avinash Persaud, “The Inappropriateness of Financial Regulation,” Research-Based Policy Analysis and Commentary from Leading Experts, www.voxeu.org/article/inappropriateness-financial-regulation, May 1, 2008.

²³“By 1890 just 300 trusts [in the United States] controlled 5,000 companies. Financial practices aided insiders while relevant information to investors and other outsiders remained problematic and usually hidden.” Gary Giroux, Business Scandals, Corruption and Reform (Santa Barbara, CA: Greenwood Press, 2013).

²⁴There are examples of where greater transparency, especially in the greater frequency of reporting, can lead to greater market instability. See Benu Schneider, ed., The Road to International Financial Stability: Are Key Financial Standards the Answer? (New York: Palgrave Macmillan, 2003). Similarly, disclosures can bring harm to private asset holders—making them out as targets for physical harm for example—and so a public interest test should be brought to bear on the form and extent of disclosures. I once started receiving hate mail at my home address from an anti-vivisection group, that had caused serious harm to some of their other targets, because I was sometime previously a director of a division of State Street Bank and another division of the bank was at the time of the letters acting as a custodian for an asset manger who held shares in a company that had tested its products on animals. The custodian’s name was down as the owner of the shares.

²⁵Webvan.com is a particularly egregious example. It was valued at $1.2 billion in 1998 and just $2.5 million by 2001.

²⁶Or the number of visits to online shops.

²⁷Arguably, some of the ventures during the South Sea Bubble of 1720 in England were equally intangible. The allure of the discovery of South America for Western Europeans drove that bubble higher. One prospectus during the South Sea Bubble read: “A company for carrying on an undertaking of great advantage, but nobody to know what it is.” See Charles Mackay, Extraordinary Popular Delusions and the Madness of Crowds (Lexington, KY: Maestro Reprints, 1841/2014).

²⁸Hans Christian Andersen’s story, The Emperor’s New Clothes, comes to mind.

²⁹Less often stated by the boom’s cheerleaders, but equally important, is how much easier it is to borrow money to invest in housing, with the attendant tax advantages, than for almost any other investment. Lord Adair Turner, former Chairman of the UK’s Financial Services Authority, among others, has frequently emphasized this driver of property and financial booms. See Jerin Matthew, “Lord Turner: Britain’s Property Frenzy Will Lead to Another Financial Crisis,” International Business Times, March 27, 2014, www.ibtimes.co.uk/lord-turner-britains-property-frenzy-will-lead-another-financial-crisis-1442030.

³⁰A sophisticated version of this defense of Basel II can be found in the response to my general arguments by Jesus Saurina, the highly intelligent, conscientious director of the Stability Department of Banco de España, in the IMF’s Finance and Development (June 2008), pp. 29–33.

³¹Market discipline was the essential component of the third of Basel II’s three pillars.

³²See Basel Committee on Banking Supervision, A Brief History of the Basel Committee (Bank of International Settlements, 2013), www.bis.org/bcbs/history.pdf.

³³The supervisory-approved methodologies were based around “value at risk” models that estimated the potential loss a bank may face from recent volatility and correlation of market prices. These models are procyclical, with short-term volatility and correlations being low during quiet times, encouraging the buildup of risk but spiking sharply higher during periods of crisis, dictating a sell-off of risky assets and causing further increases in volatility, correlations, and so on. See Avinash Persaud, “The Folly of Value-at-Risk: How Modern Risk Management Systems Are Creating Risk” (lecture, Gresham College, London, December 2, 2002).

³⁴On December 16, 2010, the European Systemic Risk Board (ESRB) was established and given responsibility for the macroprudential oversight of the EU’s financial system and the prevention or mitigation of systemic risk to the financial system. On April 1, 2013, the UK established an independent Financial Policy Committee (FPC) at the Bank of England charged with a primary objective of identifying, monitoring, and taking action to remove or reduce systemic risks with a view to protecting and enhancing the resilience of the UK financial system.

³⁵Federal Reserve Chairman, William McChesney Martin, famously said in a speech given on October 19, 1955: “The Federal Reserve, as one writer put it, after the recent increase in the discount rate, is in the position of the chaperone who has ordered the punch bowl removed just when the party was really warming up.”

³⁶Charles Goodhart and Avinash Persaud, “How to Avoid the Next Crash,” Comment Page, Financial Times (January 30, 2008). This is a proposal for time-varying capital adequacy requirements based on the acceleration in credit growth. An alternative approach focused on bank profitability is favored by the Swiss National Bank among others.

³⁷This could be measured by estimating the covariance of the risks the bank is running. See Tobias Adrian and Marcus Brunnermeier “CoVaR” (unpublished mimeo, Princeton, NJ, 2011).

³⁸For a discussion of the endogeneity of liquidity, see: (1) Anastasia Nesvetailova, “Liquidity Illusions In The Global Financial Architecture,” (Cheltenham, UK: Edward Elgar, 2012); and (2) Marco Lagana, Martin Penina, Isabel von Koppen-Mertes, and Avinash Persaud, “Implications for Liquidity from Innovation and Transparency in the European Corporate Bond Market,” (ECB Occasional Papers 50, August 2006).

³⁹See Charles Goodhart, “Is a Less Pro-Cyclical Financial System an Achievable Goal?” National Institute Economic Review 211, no. 1 (2010), pp. 81–90.

⁴⁰Persaud, “Sending the Herd off the Cliff Edge.”

⁴¹In The Fall and Rise of Keynesian Economics (Oxford University Press, 2011, p. 51), John Eatwell and Murray Millgate flatteringly refer to this as the “Persaud Paradox.” There are now a number of references to the Persaud Paradox, including Pablo Triana’s fascinating study on value-at-risk, The Number That Killed Us: A Story of Modern Banking; Flawed Mathematics and a Big Financial Crisis (Hoboken, NJ: Wiley, 2011).

⁴²This approach is well described in Anat Admati and Martin Hellwig’s, The Bankers’ New Clothes: What’s Wrong with Banking and What to Do About It (Princeton, NJ: Princeton University Press, 2014).

⁴³Funding liquidity has receded from the center of attention today. However, when the GFC first emerged in 2007 it did so as a crisis in funding liquidity. The institutions most in trouble, like Northern Rock, HBOS and Bradford and Bingley in the UK, were those that had relied most on the short-term money market funding— which dried up—of long-term mortgages.

⁴⁴Scaling up or down these error prone risk assessments by countercyclical mechanisms is unlikely to solve the underlying problem that banks are most harmed by previously considered safe assets becoming risky rather than risky assets being risky.

⁴⁵I am particularly worried about this being the unintended outcome of the move to the European Banking Union. See Avinash Persaud, “Vive la difference,” Guest Article, Economist (January 26, 2013).

⁴⁶There is some research that questions the intuition that more capital means less lending. See Thomas Hoenig, “Safe Banks Need Not Mean Slow Economic Growth,” Financial Times (August 19, 2013).

⁴⁷To be fair, bankers are not only referring to capital adequacy requirements when they say this, arguing moreover that anti-money laundering and anti-terrorism finance rules also make them reluctant to do any new business.

⁴⁸This effect of time can be seen in the differential pricing between long and short-dated credit default swaps on the same credits.

⁴⁹For an excellent survey of long-run returns in different markets and instruments, see Elroy Dimson, Paul Marsh, and Mike Staunton, Triumph of the Optimists: 101 Years of Global Investment Returns (Princeton, NJ: Princeton University Press, 2002).

⁵⁰Depositors suffer tough penalties for breaking the deposit before 12 months.

⁵¹The words of Winston Smith in George Orwell’s 1984.