February 27, 2008
Why Washington’s rescue cannot end the crisis story
By Martin Wolf
Last week’s column on the views of New York University’s Nouriel Roubini (February 20) evoked sharply contrasting responses: optimists argued he was ludicrously pessimistic; pessimists insisted he was ridiculously optimistic. I am closer to the optimists: the analysis suggested a highly plausible worst case scenario, not the single most likely outcome.
Those who believe even Prof Roubini’s scenario too optimistic ignore an inconvenient truth: the financial system is a subsidiary of the state. A creditworthy government can and will mount a rescue. That is both the advantage – and the drawback – of contemporary financial capitalism.
In an introductory chapter to the newest edition of the late Charles Kindleberger’s classic work on financial crises, Robert Aliber of the University of Chicago Graduate School of Business argues that “the years since the early 1970s are unprecedented in terms of the volatility in the prices of commodities, currencies, real estate and stocks, and the frequency and severity of financial crises”*. We are seeing in the US the latest such crisis.
All these crises are different. But many have shared common features. They begin with capital inflows from foreigners seduced by tales of an economic El Dorado. This generates low real interest rates and a widening current account deficit. Domestic borrowing and spending surge, particularly investment in property. Asset prices soar, borrowing increases and the capital inflow grows. Finally, the bubble bursts, capital floods out and the banking system, burdened with mountains of bad debt, implodes.
With variations, this story has been repeated time and again. It has been particularly common in emerging economies. But it is also familiar to those who have followed the US economy in the 2000s.
When bubbles burst, asset prices decline, net worth of non-financial borrowers shrinks and both illiquidity and insolvency emerge in the financial system. Credit growth slows, or even goes negative, and spending, particularly on investment, weakens. Most crisis-hit emerging economies experienced huge recessions and a tidal wave of insolvencies. Indonesia’s gross domestic product fell more than 13 per cent between 1997 and 1998. Sometimes the fiscal cost has been over 40 per cent of GDP (see chart).
By such standards, the impact on the US will be trivial. At worst, GDP will shrink modestly over several quarters. The ability to adjust monetary and fiscal policy insures this. George Magnus of UBS, known for his “Minsky moment”, agrees with Prof Roubini that losses might end up as much as $1,000bn (FT.com, February 25). But it is possible that even this would fall on private investors and sovereign wealth funds.
In any case, the business of banks is to borrow short and lend long. Provided the Federal Reserve sets the cost of short-term money below the return on long-term loans, as it has for much of the past two decades, banks can hardly fail to make money.
If the worst comes to the worst, the government can mount a bail-out similar to the one of the bankrupt savings and loan institutions in the 1980s. The maximum cost would be 7 per cent of GDP. That would raise US public debt to 70 per cent to GDP and would cost the government a mere 0.2 per cent of GDP, in perpetuity. That is a fiscal bagatelle.
Because the US borrows in its own currency, it is free of currency mismatches that made the balance-sheet effects of devaluations devastating for emerging economies. Devaluation offers, instead, a relatively painless way out of a slowdown: an export surge. Between the fourth quarter of 2006 and the fourth quarter of 2007, the improvement in US net exports generated 30 per cent of US growth.
The bottom line, then, is that even if things become as bad as I discussed last week, the US government is able to rescue the financial system and the economy. So what might endanger the US ability to act?
The biggest danger is a loss of US creditworthiness. In the case of the US, that would show up as a surge in inflation expectations. But this has not happened. On the contrary, real and nominal interest rates have declined and implied inflation expectations are below 2.5 per cent a year. An obvious danger would be a decision by foreigners, particularly foreign governments, to dump their enormous dollar holdings. But this would be self-destructive. Like the money-centre banks, the US itself is much “too big to fail”.
Yet before readers conclude there is nothing to worry about, after all, they should remember three points.
The first is that the outcome partly depends on how swiftly and energetically the US authorities act. It is still likely that there will be a significant slowdown.
The second is that the global outcome also depends on action in the rest of the world aimed at sustaining domestic demand in response to a US shift in spending relative to income. There is little sign of such action.
The third point is the one raised by Harvard’s Dani Rodrik and Arvind Subramanian, of the Peterson Institute for International Economics in Washington DC, (this page, February 26), namely the dysfunctional way capital flows have worked, once again.
I would broaden their point. This is not a crisis of “crony capitalism” in emerging economies, but of sophisticated, rules-governed capitalism in the world’s most advanced economy. The instinct of those responsible will be to mount a rescue and pretend nothing happened. That would be a huge error.
Those who do not learn from history are condemned to repeat it. One obvious lesson concerns monetary policy. Central banks must surely pay more attention to asset prices in future. It may be impossible to identify bubbles with confidence in advance. But central bankers will be expected to exercise their judgment, both before and after the fact.
A more fundamental lesson still concerns the way the financial system works. Outsiders were already aware it was a black box. But they were prepared to assume that those inside it at least knew what was going on. This can hardly be true now. Worse, the institutions that prospered on the upside expect rescue on the downside. They are right to expect this. But this can hardly be a tolerable bargain between financial insiders and wider society. Is such mayhem the best we can expect? If so, how does one sustain broad public support for what appears so one-sided a game?
Yes, the government can rescue the economy. It is now being forced to do so. But that is not the end of this story. It should only be the beginning.
* Manias, Panics and Crashes, Palgrave, 2005.
Kent Janér (guest): Largely, I agree with Martin Wolf’s analysis of what went wrong and what should be done in the future to prevent the by now very familiar pattern of boom and bust in regulated financial systems.
There is one aspect that I think merits more attention than it has been given, an aspect that also has some important short term effects - the equity base of the financial system. I think the equity base is currently being mismanaged, and regulators could have some tools to improve the situation.
As everyone knows, there are much more losses in the financial system than have so far been declared. I think close to USD 150 bn has been reported at this stage. That could be compared to for example the G7 comment of 400 bn in mortgage related losses, and 400-1000 bn in total losses probably covering most private sector forecasts. At the same time new risk capital has been raised to the tune of roughly 90 bn USD (ballpark number).
A back of the envelope calculation shows that a large part of the equity of the financial system has been wiped out, much more than has been reported. The market knows, the regulators know and the banks themselves certainly know that even though they are far from bankrupt, they are on average in truth operating at equity/capital adequacy ratios clearly below both legal requirements and sound banking practices.
Currently, the banks are repsonding by reporting losses little by little, keeping up the appearance of reasonable capitalization. At the same time, they try to reduce their balance sheet, especially from items that carry a high charge to capital. This way they hope (but hope is never a strategy) that time will heal their balance sheet; earnings will over time be able to offset continued writedowns. High vulnerabilty to negative surprises, but no formal problems with minimum capital adequacy ratios and control of the bank, “only” weak earnings for some time.
That is all very nice and cosy for bank´s directors, but not for the economy in general. If a small part of the banking sector has specific problems and rein in lendig, so be it. That probably has little impact on the rest of the economy. However, if the entire financial sector postpone reported losses and contract their balance sheet, that is another question altogether. The cost to rest of the economy could be very high indeed.
So, what should be done? Pretending that banks are OK and sweat it out over time is dangerous to economy as a whole, but so is being too harsh on the banks right now.
I actually think there is an answer - the banks should be made to recapitalize quickly and aggressively. Accepting new equity capital would minimize social cost of their current mistakes. There is an obvious practical problem with that, the price at which that capital is available is not necessarily the price at which current shareholders want to be diluted. So, in essence, the banking sytem continues to push the cost of their mistakes to others by not coming claen on their losses and recapitalize, rather they try to muddle through by not declaring their losses in full and pull in lending to the rest of the economy.
I think regulators should be tougher here, banks that clearly are below formal capital adequacy ratios with proper mark to market should be armtwisted to accept new money. I am also looking with dismay on the fact that even some of the weaker banks are still paying dividends to their shareholders - on a global scale I think the financial system has paid out more in dividends since the start of the crisis than they have raised in new capital. So, a likely situation is that banks with failed business models in the first part of the crisis distribute capital to their owners and somewhat later asks the taxpayer for help…
Kent Janer runs the Nektar hedge fund at Brummer & Partners AB in Sweden
Posted by: Kent Janér | February 27th, 2008 at 3:06 pm | Report this commentMartin Wolf: I believe that Mr Janer has raised an important point. Assume, solely for the sake of argument, that true losses in the banking system are indeed $1,000bn (these being bygones, of course). Then the system as a whole is severely undercapitalised, since a well-informed colleague tells me that the Tier One Capital of the world’s 1,000 largest banks was $3,365bn in 2006.
There are three possible policy responses: leave things as they are, which increases the incentive for banks to “go for broke”, reduces the equity cushion in the event of further disaster, and leaves lenders ignorant of the true position of their banks; second, encourage banks to shrink balance sheets, in order to raise capital ratios, which is a sure-fire way to create a depression; or force banks to come clean on losses and raise capital at the same time.
The last is obviously the best outcome. A particularly attractive feature is that it would greatly dilute the equity of badly managed banks, which is exactly what one wants to happen. Shareholders have to understand that this is risk capital and oversee management accordingly, even in good times.
Incidentally, the capital detailed above supported $74,232bn in lending. So a loss of $1,000bn in capital (remember this is definitely NOT a prediction) could, if applied mechanically, lead to a $22,000bn contraction in outstanding lending. That would certainly lead to a deep depression. Obviously, this cannot be allowed to happen. But it brings up the point that this is in significant respects a solvency crisis.
Finally, Mr Janer’s warning that banks might distribute dividends, before admitting the extent of their losses, and then run to the state for rescue is too plausible to be funny.
Posted by: Martin Wolf | February 27th, 2008 at 7:20 pm | Report this commentRob Goodson (Guest): I agree with Martin Wolf that the desired policy response to tremendous, but currently undisclosed, banking losses is to force banks to recognize their losses and simultaneously re-capitalize. The stronger banks would eventually thrive. Trust would (slowly) be restored. Spreads between LIBOR and Central Banks’ lending rates would narrow. Whatever contraction in worldwide lending that might occur would be mitigated and reversed as sentiment shifts from “where is the bottom?” to “the bottom is in place.” There would likely be a moderate to severe shock to the stocks of those negatively impacted, and possibly to the world stock markets as a whole. But once the bank insolvencies are public, and recapitalization occurs, lending at a more sustainable pace and price would resume (one would hope with better risk management practices in place).
Unfortunately, there are many political and practical impediments that I fear will make the forced disclosure and recapitalization policy response improbable or incomplete, and instead lead to a de facto first policy response: leave things as they are.
Politically, the U.S. is in an important election cycle, making any regulatory crackdown or forced disclosures unlikely for at least a year. Furthermore, governments are largely reactive rather than pro-active to financial crises. Thus, unless and until there is a major stock swoon or complete seizing of the credit markets, forced disclosures or forced bankruptcies seem unlikely.
As many political impediments as there are to full bank loss disclosures, the practical impediments may be greater. The secondary markets for residential mortgage backed securities (RMBS) are illiquid and the stated values are extremely low for many tranches of many RMBS. While many banks have credit default swaps (CDS) that theoretically hedge exposure, the published descriptions of the CDS market make it appear so murky that no one can be assured that the CDS are backed by anything other than a shell company. Thus there is a practical problem of truly valuing the potential losses. Moreover, forced liquidations of the RMBS or other securities of insolvent firms could lead to a spiraling decline in value and cause additional forced sales and insolvencies.
I hope that the world financial institutions’ collective losses do not approach 1,000bn. But whatever the ultimate number, the higher the real losses and impairment of banks’ capital, the more likely that the policy response will be to leave things as they are. That, in turn, could lead the U.S. on a collision course with 1990 Japan.
Posted by: Rob Goodson (Guest) BA, Economics, Harvard | February 28th, 2008 at 1:03 am | Report this commentKent Janér (guest): Rob Goodson suggests the most likely, but not best, response to the current crisis from politicians and regulators is continued laissez-faire. Let us hope that prediction will turn out to be too pessimistic.
Banking is a regulated business, ultimately controlled and backed by governments. Especially in businesses that are regulated the interest of the industry and society as a whole may not be the same all the time. Perhaps now is one of the times when the interest of the banking sector and wider society are out of sync. If so, authorities should not yield too much to the banking sector.
The financial sector has shown very high profitability over a number of years, rewarding their owners handsomely. Banks have been cashing in on the boom phase of a financial system that was gradually overextending itself. During the past few years too much lending was done relative to the capital base of the banking system. This very rapid expansion of loans was an important reason why banks created off balance sheet vehicles to carry the loans; they simply did not have enough capital to put it on their balance sheets. One could say that a basic control mechanism for credit creation, regulatory bank capital, was circumvented with the help of SIVs, conduits etc. and aggressive accounting practices.
On top, lending standards were lax since the loans were going to be repackaged into structured products and sold to someone else. Availability of funding and market liquidity was taken as given, a free public good to be used. The rating agencies were not doing their job and regulators were not paying enough attention to the shifting structure of the system.
We are now in the bust phase. Losses appear, asset prices fall, funding becomes harder to find, liquidity dries up, lending contracts and the economy weakens. There obviously is a price to pay for mistakes made. Many families will have to leave their homes; some of those families should never have borrowed in the first place. Some corporations will run into trouble, as will some investors and speculators. The banks themselves will take heavy losses for their past mistakes. There are important lessons to be learnt from this for future regulations. There are also probably things that could and should be done to ease problems for homeowners.
At this stage, the most important issue is to minimize further losses to society from an awkward starting point. A recapitalization of the financial sector is urgently needed. The alternative is allowing banks to get their balance sheets in order by contracting lending, a very unpleasant strategy. Society needs a banking sector that lends and provides liquidity in a prudent way.
Not having a functional financial system for a couple of years entails unknown, but probably very high, costs to society as a whole. Ironically, that will not help the banking sector itself. For example, future losses from consumer related credits are likely to be much higher in a credit contraction than a recapitalization scenario, because the former implies slower growth and higher unemployment.
I thus do not see much of a risk for the broader financial system and asset prices should authorities toughen their stance towards banks with weaker balance sheets. On the contrary, it would help both the growth outlook and asset prices as risk premium would fall. Importantly, recapitalization implies less fire sales of already cheap assets, thus lower risk of cascading defaults of leveraged investment vehicles.
In this episode it is the financial sector that is dragging down growth prospects for the rest of the economy. Some segments of the credit markets are totally bombed out, implying depression like losses from lending. On relative valuation, credit instrument that are close to banks like leveraged loans stand out as cheapest among peers. These are clear signs that banks are balance sheet constrained.
If banks are not willing to invest in / lend through the cheapest segments of the securitized credit market, do we really expect them to lend to corporates and households at less attractive terms?
Today’s situations stands in sharp contrast to the situation a couple of years ago. Lenders now get well paid relative to risk for extending loans. Still they are not doing it since they are constrained by low equity and swollen balance sheets. Weak banks prefer trying to stretch out their problems over time. This leaves a dark cloud hanging over financial markets. Living with a weak banking system carries a lot of risks. Shocks to the system that could be handled with ease by a stable banking system could now send it into tailspin with massive liquidation trades going through the market at fundamentally too low prices. Such a situation does not encourage risk taking from outside the banking sector either.
It could be added that wholesale recapitalization certainly would include raising a lot of new risk capital, but it is not the only way. Many banks could and should sell non-core parts of their business. And, if they on realistic assumptions are near or below formal capital adequacy rules, they should not be paying dividends. Strong players should be buying weaker ones, something made more difficult today by regulators in effect allowing weaker players to try to muddle through.
Regulators made the mistake of allowing rapid expansion of not so good credit, hidden off balance sheet. Now it looks as though they will allow the weak banks to play the current situation their way, resulting in a banking system that is not lending as it should. It also leaves a dark cloud of uncertainty and fragility hanging over the economy and financial markets. There was a high cost to society for the first mistake. The cost could be higher for the second.
Kent Janer runs the Nektar hedge fund at Brummer & Partners AB in Sweden
Posted by: Kent Janér | February 29th, 2008 at 10:53 am | Report this commentMartin Wolf: I find myself in agreement with Mr Janer once again.
Posted by: Martin Wolf | February 29th, 2008 at 8:05 pm | Report this commentNouriel Roubini: In his February 27th column in the Financial Times Martin Wolf considers further my “12 steps scenario to a systemic financial crisis” that he had very thoughtfully presented in his previous column.
In the new column Wolf makes a series of interesting points. First he argues that my “analysis suggested a highly plausible worst case scenario, not the single most likely outcome”. Second, he argues that, even in this worst case scenarios, the expected losses for the financial system (that I estimated to be about $1 trillion) and thus the potential fiscal costs of a bailout of the financial system would be only 7% of GDP, an amount that a large economy such as the US can easily afford: as he argues the US public debt would increase in that scenario only to 70% of GDP. And the yearly costs of servicing such increased debt would be – in his view – “a mere 0.2% of GDP in perpetuity. This is a fiscal bagatelle”.
What should we make of his argument? For now I will not debate whether my scenario is plausible or not; let us leave aside this debate for now and let us concentrate on the fiscal costs of a bailout of the financial system in this “worst case scenario”. Will it be only 7% of GDP or more? I will argue in this article that such fiscal costs of a financial crisis could be much higher than 7% of GDP, as high as 19%, even leaving aside the even bigger losses in the net worth of the private sector that a severe financial crisis would imply.
Here are a number of reasons why the overall financial losses and the fiscal costs of a bailout of the financial system could be much higher than 7% of GDP.
First of all 7% of GDP still represents $1 trillion of greater stock of public debt. And $1 trillion is not spare change, even for a large country like the US. The S&L crisis cost the US budget about $120 billion (closer to $250 billion in today’s dollars). A cost of $1 trillion is at least four times as high. Also consider the political consequences of asking the US tax payer to allow the increase of the explicit debt of the US government by $1 trillion in order to bail out the financial system. Currently even much more modest proposals to use a few billions of public money to deal with the foreclosure crisis are met with political resistance; let alone thinking of adding another $1 trillion to the US public debt.
Second, in past episodes of fiscal bailouts of a financial system in crisis, such as those described by Wolf in a chart in his column, such bailouts have been usually associated with a formal or effective nationalization of most or all of the banking system. In a financial crisis the entire capital of the banking system is often wiped out and, ever after shareholders have lost all of their equity, the value of the assets of the banks is below the value of their liabilities (deposits and other insured debts). Thus, a fiscal bailout and recapitalization of the banking system requires an effective nationalization – until banks are cleaned up, recapped and sold back to the private sector – of the banking system. So in this $1 trillion (7% of GDP) fiscal bailout scenario, you would get an effective nationalization of a good part of the US banking and financial system. Is this nationalization a consequence that the US – the beacon of free market capitalism – is comfortable with? As it is, already now the partial recapitalization of the US financial system is occurring via foreign government-owned entities – the sovereign wealth funds – providing the new capital. So, regardless of whether it would be the US government or some foreign governments to take control of US financial institutions we would be on the way to an effective domestic or foreign nationalization of the US financial system.
Third, in this financial crisis scenario the real losses for the net worth of the US private sector are much higher than $1 trillion. With home prices having already fallen by 10% relative to their peak, $2 trillion of the value of equity in US homes (housing wealth) has already been wiped out (14% of GDP). Since home price are almost certain to fall by another 10% in 2008 and beyond a 20% cumulative fall in home prices is equivalent to $4 trillion of losses (or $28% of GDP). And an eventual cumulative fall in home prices of the order of 30% is highly likely at this point; that would wipe out $6 trillion of housing wealth (or 48% of GDP). 48% of GDP is a much bigger loss than 7% of GDP.
Fourth, in an average US recession – even in a mild one such as those in 1990-91 and 2001 - the S&P500 falls by an average of 28% and often it takes five to ten years for the index to recover its pre-recession level. Given that the US stock market capitalization was about $20 trillion at its peak in 2007 that 28% loss would wipe out another $5.6 trillion (or 39% of GDP) from the net worth of the private sector.
Fifth, commercial real estate was – like residential real estate – in a bubble that is now starting to go bust as reckless lending and underwriting practices similar to those in residential mortgages were experienced in commercial real estate mortgages. Prices of commercial real estate are expected to fall by 10% to 20% that would wipe out a few additional trillion dollars from the value of such assets.
Sixth, and most important, my estimate of $1 trillion of losses for the financial system was based on the view that total credit losses on mortgages would be around $300 to $400 billion while the other losses of the financial system (on consumer debt, commercial real estate loans, leveraged loans, downgrade of the assets insured by monolines, loans to non financial corporations and holdings of corporate bonds, credit default swaps) would add up to another $600 to $700 billion. But the total credit losses on mortgages may end up being much higher than $300 billion and as high as $1 trillion to $2 trillion.
The reason is as follows: today there are already over 8 million households with negative equity in their homes (i.e. the value of their homes being lower than the value of their mortgage debt); if home prices fall by another 10% (20% cumulative) the number of households with negative equity would be over 16 million; and if home prices fall by a cumulative 30% about 21 million households would have negative equity. As discussed in a previous column of mine in this scenario of negative equity households have a strong incentive to walk away from their homes and saddle the banks or holders of the mortgage with the loss deriving from the difference between the mortgage value and the home value.
Based on work done by Calculated Risk, I have argued that in this scenario of massive “jingle mail” the losses for the financial system would be at least $1 trillion and as high as $2 trillion. To get $1 trillion of losses one does not have to make heroic assumption: even assuming only a 20% (rather than 30% fall in home prices) and even assuming that only half of the 16 million households with negative equity “walk away” you easily get losses of the order of $1 trillion (assuming that the average mortgage is $250k and that the loss to the creditor is about 50% with the latter assumption based on the fall in the price of the home, the legal and other costs of foreclosure and the additional costs of selling a home in a very illiquid market). With larger home price depreciation and thus larger numbers of households walking away the losses can be as high as $2 trillion.
These are staggering amounts that would totally wipe out all of the capital of the US banking system and lead to a systemic banking crisis. So, in the worst case scenario if the losses on mortgages are $1 trillion (or in an extreme case $2 trillion) rather than the current estimate of $300 billion and, if on top of these losses the additional credit losses from consumer credit, commercial real estate, leveraged loans, etc. are another $700 billion the total losses for the financial system would add up to at least $1.7 trillion ($1 trillion plus $700 billion) and as high as $2.7 trillion ($2 trillion plus $700 billion). $1.7 trillion adds up to a maximum fiscal bailout cost of about 12% of GDP (not 7%) while $2.7 trillion adds up to a maximum fiscal bail-out cost of about 19% of GDP.
Now 12% of GDP is not spare change and, certainly, 19% of GDP is a staggering amount (ten times higher than the fiscal bailout cost of the S&L crisis). Of course some of the losses would be taken by the shareholders of the banks and financial institutions; but a clean and fair fiscal bail out of the financial system – that reduces the fiscal costs – would imply first wiping out all of the shareholders of these institutions (as their capital/equity is fully destroyed by such losses) and a formal nationalization of a good part of the banking and financial system. And thus in such a scenario one has to ask again the question: what are the implications of a potential nationalization of a good part of the financial system of the most advanced capitalist country in the world?
Some final and additional considerations.
First, if millions of households exercise their legal option of walking away from homes with negative equity, even before any government intervention the losses for the financial system would be larger than the equity in the system. So banks and other financial institutions would be insolvent and forced to be taken over by the government leading to the need for a fiscal bailout (as most bank liabilities are effectively safe given deposit insurance).
Second, proposals for a fiscal resolution of the foreclosure crisis do not necessarily prevent this effective bankruptcy of the banking system. For example, if as proposed by folks such as Mark Zandi and Alan Blinder, the government were to buy via auctions distressed mortgages and RMBS, the only way to limit the fiscal cost of such government intervention would be to buy such mortgages at their true market value that is equal to the value of the property once the full downward price adjustment has occurred. In this best case the government does not lose much money on such a scheme of taking over mortgages and repackaging them in lower loan value and lower interest rate mortgages for distressed borrowers. However in this scenario: first a good part of the mortgages are effectively nationalized; and second the full losses deriving from the difference between the market value of the home and the value of the original mortgage are then absorbed by the creditor financial institutions. Thus you end up again with a situation where the equity of these financial institutions is wiped out, they become insolvent and the government is then forced to take them over and incur the fiscal costs of bailing them out and recapitalizing them.
Conversely, if the mortgages are bought – via auctions – at a price that is higher than the eventual market value of the home, the fiscal costs of this effective nationalization of the mortgages becomes potential very large (as the government takes the downside risk of the further fall in home values and eventual default of households with negative equity) while the financial sector’s equity is less than fully wiped out. Thus, in this scenario the weaker financial institutions go belly and are nationalized – in spite of this partial bailout of the shareholders of these entities – while the stronger ones do not go belly up only because the public sector provides a massive bailout (equivalent to the difference between the secondary market value of the mortgage and the true underlying value of the home). In summary, regardless of whether you nationalize mortgages (in the scenario of the government buying a whole bunch of such mortgages) or nationalize the banks you end up with massive losses (again possibly as high as $1 trillion to $2 trillion for mortgages alone) that are – in one way or the other – fiscalized.
In conclusion, it used to be said that “a million here, a million there and soon you are speaking about real money”. In this case we need to amend the saying to “a trillion here, a trillion there and soon you are speaking of staggering amounts of money and massive fiscal bailout costs”. Martin Wolf says that a fiscal bill of 7% of GDP is modest and affordable. But the analysis above suggest that the fiscal bill of bailing out a banking and financial system that suffers a systemic crisis would be at least 12% of GDP and as high as 19% of GDP. Even for a rich country like the US 19% of GDP (or $2.7 trillion of additional public debt) is not spare change nor is a “fiscal bagatelle”. And saddling every US household with an additional $30,000 of debt in perpetuity is not small burden either. And all this is true even leaving aside the other $10 trillion plus of losses in the net worth of the US private sector (fall in the value of residential real estate and commercial real estate, and in the value of the stock market) that a severe recession and financial crisis would imply.
The wise Wolf is himself – at the end – very aware of the political economy of a financial system where in good times the profits are privatized and in bad times the losses are socialized. This is not a politically sustainable regime. As he correctly puts it:
“This is not a crisis of “crony capitalism” in emerging economies, but of sophisticated, rules-governed capitalisms in the world’s most advanced economy. The instinct of those responsible will be to mount a rescue and pretend nothing happened… Worse, the institutions that prospered on the upside expect rescue on the downside. They are right to expect this. But this can hardly be a tolerable bargain between financial insiders and wider society. Is such mayhem the best we can expect? Is so, how does one sustain broad support for what appears so one-sided game?”
This is indeed a one-sided game where financial insiders privatize profits while the massive losses of their reckless behavior – searching dangerously for yield, gambling for redemption, being subject to distorted incentive not to monitor their lending and risky investments - are systematically socialized during a crisis. This is actually “crony capitalism” of the worst kind, as bad as the one that plagued emerging market economies and led to their severe financial crises in the last decade.
PS: In this article I have left aside the complex issue of what are “inside” assets versus “outside” assets (or net worth) of an economy. Financial losses are in part a redistribution of wealth from creditors to debtors and thus “inside” assets that should not affect the net worth of an economy. But they can have financial consequences on net worth for several reasons. First, a financial crisis that leads to a credit crunch can cause a recession that reduces the value of output/income that is generated by the economy; this is a real loss of income and welfare. Second, the part of loss of net worth that is driven by an asset bubble and excessive investment in assets (such as residential and commercial real estate) that eventually go into a bust is a real wealth loss as it signals that too much was invested for many years in assets with low returns. Thus, the sharp fall in the market value of such assets (residential and commercial real estate and value of equities) reflects this low return on such capital goods, another factor that reduces long-term consumption and economic welfare.
Posted by: Nouriel Roubini | February 29th, 2008 at 9:45 pm | Report this commentMartin Wolf: I think Nouriel’s post is so important that I plan to devote a column to it in the not too distant future.
Posted by: Martin Wolf | March 3rd, 2008 at 6:24 pm | Report this commentChristopher Steane (guest commenter):
The main constraint on credit creation is not bank capital. It’s the hugely higher cost of maturity transformation.
But how can this affect the system as a whole? What are those banks who have liquidity doing with it? Are they keeping it very short, thereby causing concern about wholesale funding? Or are money market funds keeping maturities very short? How does the process of credit contraction work?
Bank A decides to call in a short term loan to a hedge fund. The hedge fund sells an asset to Bank B and returns the money to Bank A. Bank A repays a depositor, who puts the money on deposit with Bank B. In illiquid markets, the hedge fund may lose capital by its forced sale. But Bank B correspondingly gets the asset more cheaply. If it’s a loan asset, the return to maturity for Bank B will be higher than it was for the hedge fund. Bank A has delevered, as has the hedge fund. All other holders of the same asset feel poorer (and if they mark to market, have seen capital destruction). But if it is a loan asset, then anyone who marks to market now has a higher yield to maturity.
As assets reprice for risk, holders take losses. In effect, the price for holding risky assets has gone up. But has the risk of these assets also gone up? The credit risk may or may not. No one knows today how risky any particular loan is. You can tell only in arrears. What has gone up is the return to liquidity. It is true that the value of a loan today is based on what you can be paid for it. But tomorrow, when the borrower repays, the value is then certain. And if you don’t like the price the market is paying today, and keep the loan until tomorrow, your mark to market loss today, if you take it, will be exactly offset by a higher return on your risk. But to do this, you have to have the funding to do it: and the premium you are paid is for this use of liquidity.
Banks have expected losses and unexpected losses. The purpose of capital is to protect against the latter. And under a mark to market system, they need it against the volatility of market prices for loans, which are driven as much by liquidity shortage as they are by credit risk. What makes the bank’s shareholders poorer is (1) changes in the external world which increase the likelihood the banks portfolio of loan assets will not pay out contractually (2) increases in bank’s funding costs - not the interbank rate (normally passed though to borrowers), but the cost of managing the liquidity mismatch. If all assets are match funded, it’s only (1) that causes losses to shareholder wealth.
MTM accounting however requires bank capital to absorb MTM negative valuations. It means that assessment of a bank’s asset portfolio and capital adequacy is made by reference to market prices, which in an illiquid market reflect only the most distressed seller. There is a lot of smart capital which thinks risk margins are too high already, but it’s not deployed because of the belief that distress may drive them higher yet. And if MTM accounting is so widespread as to force further distress selling, they may well be proven correct. MTM accounting for loan books is like applying bankruptcy valuations to businesses. It may be objective, but it’s not relevant unless the business is insolvent. I don’t believe Warren Buffet measures his decisions by reference to market values every five minutes.
The bad news for the ultra bears is that the liquidity of major banks is supported by central banks as a quid pro quo for regulation: and that the fears of central bankers about moral hazard are a combination of noise and irrelevance. Who is suggesting that the owners of Bear Stearns and Northern Rock, and the senior managers, have not paid a very full price for their misjudgements and flawed risk management? Shareholders have borne the enormous losses at Societe Generale. While individual small banks may be allowed to create losses for depositors, system-wide losses are not going to happen. And public rescues for banks are depositor protection not protection for shareholders or managements.
Those who cannot see beyond their envy at bank compensation packages miss the mark. There is indeed a problem of perceived equity. The problem is that the financial system is much bigger than the sports or entertainment industries, but the dynamics of individual compensation are similar. We don’t normally criticise individuals, whether football players, film stars or journalists, for maximising their incomes in a free market. And as for the argument that banking is a regulated industry, most of the problems in financial markets currently are not to do with the regulated entities: indeed, regulation has driven activity into unregulated areas, and that’s where the problem has been created. Then we criticise bank managements for creating the wrong incentives: but they operate in a free labour market: and it is not the case that the most generous remuneration packages align with the worst managed organisations: look at Goldman Sachs. Banks have been losing some of their most skilled individuals to private equity and hedge funds for years. You can criticise hedge fund remuneration, but no where is there any law requiring anyone to to have their money managed by hedge fund managers.
What’s constraining banks is not capital, it’s liquidity. Of course, the banking system is a closed loop and the deposits created by new lending must go somewhere. But if they go to money market funds, or institutional investors, who place them in money markets at short maturities, they are not as good to the banks as smaller retail deposits. So expect retail deposits to be bid up relative to interbank rates as banks compete to build retail deposit bases. And expect margins on wholesale lending, if they continue to be related to interbank rates, to rise to reflect the cost of retail deposits, as well as to reflect the cost to banks of more cautious liquidity policies.
For the near future, capital requirements for banks will be higher, and the returns to intermediation will have to rise to attract and remunerate the capital after paying the higher costs of maturity transformation. For years, the returns to intermediation have been pressured by capital market disintermediation. This was driven in large part by regulatory capital rules (why would anyone invent a SIV if not for Basel 1?). Now it will take quite a long time for people to forget the hard lessons about the cost of liquidity and the risks of maturity transformation. Of course, if the regulatory response is to impose liquidity rules beyond those the market would, then a further round of disintermediation will eventually follow: but Central Banks will be more attuned to the systemic risks posed by non bank intermediaries. In any event, this regulatory dilemma won’t go away.
The comment makes two mistakes. First, it says the problem is that banks do not have enough capital and proposes forced recapitalisation. Second, it ignores the real problem, which is not capital but liquidity. As to the first, no one knows the extent of the ultimate losses in mortgage portfolios: there is a bid / offer spread referred to by Martin Wolf at the beginning of his column of $300bn to $2000bn. How much of this is in the banking system, as opposed to hedge funds / institutional investors / insurance companies? No one knows. How much of it should be taken into account by bank managements? Presumably the commentators would have them look at market values. However, mark to market is little help in assessing ultimate losses if its the liquidity risk premium which has shot through the roof. And MTM assessment of credit risk can be overdone. If banks in 1982 had been forced to mark to market the LATAM lending then the whole system would have needed rescuing then. While some of the problem went away through banks earning enough to create adequate provisioning, also the ultimate net losses were lower than feared. If you look at the experience of fully funded mortgage securitisations in the UK in the early 90’s, the margin earned on the healthy mortgages over time paid for the losses, and the securitisation vehicles had enough liquidity to bridge the gap, and did not have to mark to market. History does not teach that full marking to market at distressed sellers’ prices is a good way to handle the problem.
The problems of the current market have been created by the absence of regulated bank activities, not by their presence. For years, the market has determined that the requirements for bank capital in Basel 1 are too high. This has created the rush to disintermediation. The credit creating function of banks - through maturity transformation - has been replaced by market liquidity mechanisms. And now that those have gummed up, because no one trusts the market values set by distressed sellers and illiquidity, everyone looks to the banks to save them. But the US is now very short of wholesale bank capacity: it has atrophied over the last 10 years in the face of disintermediation. And not just for major corporates. Look at mortgages. A broker originates a mortgage and arranges for it to be funded initially by an investment bank, which collects it temporarily in a warehouse with thousands of others. These form the assets for a new securitisation. 90% of the amount of the securitisation is in the form of highly rated securities which are sold to a leveraged hedge fund. This funds its investment through the repo market with its prime broker, another investment bank, which in turn funds itself in the repo market from a money market fund. The money market fund is a collective vehicle for thousands of individual depositors. Where in this process of credit creation is the regulated banking sector? Nowhere. How has the maturity transformation, from the immediate access offered by the money market fund, to the 25 year mortgage maturity, taken place? By an assumption on the part of the hedge fund and the money market fund that there will always be a liquid market for this paper, in which a 25 year investment can always be converted into cash without (material) loss of principal. What is wrong with this? The assumption that there will always be liquidity.
Aye, there’s the rub. Enough liquidity. That’s the conclusion of both of the prior paragraphs. The fact is that the banking system can’t immediately ride to the rescue, but not because it does not have enough capital. Of each dollar lent, only 5 to 10% is the bank’s own money. The rest comes from deposits. And given what happened to Northern Rock and Bear Stearns, the first thing the management of every bank on the planet is concerned about is the liquidity of the bank. And in this regard, as noted above, one week deposits from an institutional depositor form a weak base to build 3 or 5 or 10 year lending. It supports the bank’s short term repo book, but that does not in itself create credit. Banks lending money for years to create credit need to know they will have the deposits to fund this. And until they do, the presence of more capital is an unnecessary burden because they can’t use it and the stock market will penalise them. The market, of which the commentators are so fond because of the transparency and discipline, is precisely why the US banks are no longer in the business of wholesale intermediation: the returns have been lousy and those banks which have not diversified out of lending have been taken over by those whose share prices reflect the returns from the new business of disintermediation. This is not a judgement of value but it is a statement of fact.
What to do? First, be very cautious about introducing new regulations. The old ones played a big role in creating the problem. Second, expect that banks rebuild liquidity and give them every incentive to do so. Third, accept that returns to intermediation will increase as the margins paid by wholesale borrowers increase. A 1% increase in margins on all bank lending will do wonders to pay for the increased costs of liquidity and to attract capital. Fourth, get on with fixing it rather than look for scapegoats. The current disruption is, in my view, the most serious I have seen in a 30 year banking career, and will take time to resolve. The banking system is a necessary tool: what is required is to put it to work to fix problems created by over-enthusiastic adoption of financial innovation. But allocating blame for the mess is extremely difficult, and most of the comment I have heard, both from politicians who can be expected to be ignorant and from commentators who should know better, is ill informed.
One can imagine the political approach to this problem. At the core is the issue of individuals who borrowed too much to fund purchases of houses. Among all the commentators now braying, do any have well thought through proposals? All seem to imagine that regulators can master the technical details when they can’t. A simple mind might advocate controls on the proportion of money someone needs to put up as equity before getting a mortgage. Can we imagine what the politicians would say? Perhaps no one advocates this for the very good reason that it wouldn’t really work, because it would simply stimulate the growth of second mortgage lenders offering to fund the equity proportion. If we made second mortgages for this purpose illegal, then the lending would be unsecured, and possibly more costly. Can we make borrowing and lending between consenting adults illegal? There’s a reason that banking is the world’s second oldest profession. So can we (1) stop looking for the scapegoats and (2) concentrate on realistic practical proposals to get the intermediation system working well again.
I have been working as a lending banker for 30 years and am now the Global head of Structured Finance at ING Bank. However, my comments are made strictly in a personal capacity
Posted by: Christopher Steane | March 22nd, 2008 at 9:48 am | Report this commentMartin Wolf: I think Mr Steane has made many interesting points. I certainly sympathise with his desire not to seek scapegoats. But one cannot ensure that mistakes are not repeated if one does not try to work out what those mistakes were. So I do not see this as an attempt to find scapegoats, but rather as an attempt to identify who or what was responsible for the present mess. But it also seems to me to be politically impossible and morally reprehensible to ask the great mass of taxpayers to pay the costs of a rescue of the reckless rich. After all, someone benefited hugely from the securitisation practices that caused the illiquidity Mr Steane describes. Finally, should we make borrowing and lending between consenting adults illegal? Yes we must, if an agency of the state is called upon to guarantee the value of the loans ex post.
Posted by: Martin Wolf | March 29th, 2008 at 6:36 pm | Report this comment