A Theory of Two Big Balance Sheets
Friday, January 16, 2009
Alex Pollock explains the recent period of bubbles, busts, and bailouts—with lessons for improving the financial system.
How can we get any perspective on the financial bust we are experiencing, as we watch all kinds of asset prices plummet, famous financial firms fail, credit contract, and bailouts bloat the government’s balance sheet? We have to begin by considering the bubble which preceded and caused the bust.
This bubble was huge and the bust is severe, but the basic patterns are not at all new. They recur throughout financial history. Consider this observation:
“Nowhere does history indulge in repetitions so often or so uniformly as in Wall Street. When you read contemporary accounts of booms or panics, the one thing that strikes you most forcibly is how little either speculation or speculators differ from yesterday.”
This was written in 1923.
Are bubbles and busts caused by outside forces or “shocks” to the financial system—or are they intrinsic or “endogenous” to the nature of financial behavior? A notable argument for their being endogenous was that of Hyman P. Minsky:
“The successful functioning of an economy within an initially robust financial structure will lead to a structure that becomes more fragile as time elapses.”
Minsky (whom I knew well) liked to summarize his position with the paradoxical line, “Stability creates instability.” What he meant was that conditions of stability convince financial actors that greater borrowing and the increasing use of debt, especially short term debt, are safe. The amount of debt relative to equity, or “leverage,” rises. We see this in a basic way when home buyers come to have smaller and smaller down payments, and bigger and bigger mortgages relative to the price of the house.
Borrowers, lenders, regulators, politicians, and pundits become confident of future prospects as they learn by experience that more debt seems to work. Everybody makes more money for an extended period. “Short term financing of long [and illiquid] positions becomes a normal way of life.”
The rapidly increasing availability of debt tends to push asset prices higher, until, in the critical stage of affairs as envisioned by Minsky, the ability to repay the debt becomes dependent on the ability to sell the asset (most recently, a house) at a higher price to somebody else—who to buy it must have still more borrowed money from an optimistic lender. The system is now “fragile.” This game cannot continue indefinitely. Eventually the bust comes and the pyramids of leverage collapse.
So the “Great Moderation” of economic performance, with sustained growth, shallow recessions, and low rates of inflation, of which the world’s central bankers were so proud, created the conditions, from a Minskian view, for first bubbles and then financial panic.
The Pollock summary of Minsky is: Leverage is the snake in the financial market Garden of Eden—the temptation that leads to the fall.
In 1825, a major financial bust occurred in England. It included the first widespread default on their international debt by Latin American countries, which were newly independent. The contemporary banker and financial theoretician Lord Overstone described the cycle in these terms: “A state of quiescence—next improvement—growing confidence—prosperity—excitement—overtrading—convulsions—pressure—stagnation—distress—ending again in quiescence”—then of course we start over again.
But why should this be? Don’t we learn from experience? Doesn’t economic knowledge increase? Doesn’t having computers, vast amounts of data, and new mathematical models guide lending and investing decisions? Well, the idea that improved knowledge will keep us out of trouble is also not new:
“Disraeli had asserted that the boom of 1825 would not turn to bust because the period was distinguished from previous ages by superior commercial knowledge.”
Our 21st-century housing bubble, now deflated, was inflated despite—indeed partially because of—amazing computer power, reams of data, and sophisticated models operated by exceptionally bright analysts informed by Nobel Prize–winning financial theories. These computerized models created a sense of security, just as did the “superior commercial knowledge” of 1825.
As the great investment guru Benjamin Graham wrote in his classic The Intelligent Investor in 1949:
“The concept of future prospects…invites the application of formulas out of higher mathematics to establish the present value of the favored issues. But the combination of precise formulas with highly imprecise assumptions can be used to establish, or rather to justify, practically any value one wishes…. Mathematics is ordinarily considered as producing precise and dependable results: but in the stock market the more elaborate and abstruse the mathematics, the more uncertain and speculative are the conclusions.”
Two generations later, no thought could be more apt for the problems of complex mortgage securities, structured by abstruse mathematics applied to estimates of the behavior of human borrowers.
So here is where we are now:
“The overstaffed, mismanaged banking industry, now contracting as fast as it had expanded, needs to return to fundamental principles and sound underwriting practices. Sobered by the stock market crash of last October, banks should now be ready to provide investors with a more reasonable, competent and careful industry [rather than] the outright speculation by bankers which had destroyed profit margins…. A huge human price was being paid by the thousands of people now losing their jobs in London and New York.”
Quite accurate today, but this was written about the crash of 1987.
Short-Term Lenders and “Deleveraging”
Essential to understanding busts is the role of very risk-averse short-term lenders. Short-term lenders, such as buyers of commercial paper, bank depositors, investors in money market funds, interbank money market dealers, and investors in repurchase agreements, do not intend to take any real risk. They have no upside investment potential, they just want to employ funds in a very safe way for a modest return in the form of an interest rate.
In a financial panic, such as in 2007–08, such risk-averse lenders suddenly realize they are indeed seriously at risk—for example, as holders of deposits in IndyMac Bank, which failed, or as short-term lenders to Lehman Brothers investment bank, which became bankrupt. Everyone thinks, naturally: what other financial firms might collapse? How can they know who is really solvent and who is broke?
The result: the short-term lenders all become conservative at once and withdraw. They seek the greatest safety of Treasury bills. These may have very little yield, but keep principal secure. The money market will have reached the Will Rogers moment: it cares not about the return on capital, but the return of capital. There is a discontinuous drop in the ability of leveraged firms to continue the short-term borrowing which had until then seemed “a normal way of life.” They have to sell assets instead, into falling markets whose prices are driven lower by the distressed sales.
The losses on the sales raises further fears about their solvency and makes it even more difficult for them to borrow—the short-term lenders become even less willing to roll over their debt in the generalized fear and heightened uncertainty about who is broke and who isn’t.
As described by David Ricardo two centuries ago: “On extraordinary occasions, a general panic may seize the country, when every one becomes desirous of possessing himself of the precious metals [today’s cash and Treasury bills]—against such panic banks have no security on any system.”
Their lack of security in a funding panic “on any system” is what triggers government deposit guarantees (“insurance”) and also government bailouts.
Much of the lending which lubricates financial markets is on a secured basis, which makes it in ordinary times safe for the lenders, protected by the collateral, and safe as a reliable funding source for the borrowers. Such lending is administered by the margin clerks, whose job it is to make sure that the value of the collateral is sufficient to cover the debt at all times, with a sufficient margin for error: the collateral needs to be some amount greater than the amount owed.
When market prices are falling, the secured lenders need more collateral by definition and also higher margins for error. At some point, the borrowers run out of collateral to cover their borrowings—then instead assets must be sold to pay off the debt, putting further downward pressure on asset prices.
We may take as a motto summing up this situation the wonderful caption of a cartoon from James Grant’s Mr. Market Miscalculates. A Wall Street banker on the phone, looking harried with tie loose and shirtsleeves rolled up, says: “Fine. I’ll try it again. Pretty please, Mr. Margin Clerk, Sir….”
That cartoon is where we are at the moment: the result is “deleveraging”—financial firms trying to pay off debt, reduce the ratio of debt to equity (with equity depleted by losses), and correspondingly shrink their balance sheets.
Two Big Balance Sheets
Clearly any one firm can shrink its balance sheet by selling assets to someone else. But ask yourself this question: How is it possible for everyone to shrink at once?
Imagine the combined balance sheet of all the financial firms—all the balance sheets thought of as one big balance sheet. How is it possible for this aggregate balance sheet to shrink?
One way is an asset price collapse and widespread default on debts of all kinds. This is a “debt deflation,” the outcome no one wants, to say the least.
The only other way for the aggregate private balance sheet to shrink is for another balance sheet to expand: the government’s balance sheet. This includes the U.S Treasury, government agencies of various kinds, including the FDIC, Fannie Mae, and Freddie Mac, and of course the Federal Reserve. In other words, we have a kind of yin and yang of two big balance sheets: the aggregate private balance sheet can delever only by the expansion of the government balance sheet.
And this is of course what is happening. Since the beginning of the financial market panic in August 2007, the Federal Reserve Banks have expanded their combined balance sheet by about $1.4 trillion, or about 160 percent, looking ever more like a risk-taking commercial bank. The Federal Reserve System is now intermediating the risk between mortgage securities, loans to financial firms, and commercial paper on one side, and its own rapidly escalating deposits on the other.
What is the alternative in a panic? According to Walter Bagehot, the father of central banking theory, there is none. Here is his 1873 description of the problem:
“In a panic, there is no new money to be had; everybody who has it clings to it, and will not part with it. Merchants…are under immense liabilities, and they will not give back a penny which they imagine that even possibly they may need to discharge those liabilities. And bankers are in even greater terror. In a panic they will not discount a host of new bills [that is, make new loans]; they are engrossed with their own liabilities.”
Therefore, in Bagehot’s celebrated dictum, in a panic the Bank of England, or in our case the Fed, must “lend freely.” Both of them today, with a housing bust and a financial panic in both countries, are certainly doing so.
Bailouts and Taxpayer-Investors
Not only the central banks, but the treasuries of many countries are now expanding their balance sheets to keep the financial sector functioning. The United States, Britain, Ireland, Germany, Spain, the Netherlands, Belgium, Iceland, and others all display the recurring pattern of governments using the public credit to offset the losses of financial firms.
In the U.S. mortgage market, Fannie Mae and Freddie Mac have been put into the complete control of the government by the legal process of conservatorship. They have gone from being “government-sponsored enterprises” to government housing banks. Together with the government’s Federal Housing Administration, they now represent about 90 percent of current mortgage loan funding. They have recently announced a major program to reduce required payments on troubled mortgages.
To this we must add the Troubled Assets Relief Program (“TARP”) or Paulson Plan, by which the Treasury Department itself will make $700 billion in investments to try to stabilize the credit markets. This program has now moved from the original idea (not a very good one, in my opinion) of trying to buy troubled mortgage securities, to making equity investments in financial firms, which directly recognizes the fundamental issue of insolvency and fears of insolvency.
We have moved through the typical stages of governments faced with a severe financial crisis. First, there is delay in recognizing the extent of the losses while issuing assurances. A notable example, often pronounced in 2007: “The subprime problems are contained.” Then there is the central bank as liquidity provider or lender of last resort. But lending, however freely, provides by definition more debt. If a firm’s capital is gone, however much more you may lend it, it is still broke. What then?
When financial losses have been so great as to run through bank capital, which can happen unfortunately quickly in highly leveraged firms; when no one knows who is broke and who isn’t; when waiting and hoping and issuing assurances have not succeeded; when uncertainty and therefore risk premia are extreme—what is needed is not simply more liquidity, but more equity capital.
Thomson Hankey, a mostly unknown intellectual opponent of the celebrated Bagehot, made two basic arguments against government assistance, as restated by James Grant:
“No. 1, moral hazard: Let profit-maximizing people come to believe that the Bank of England will bail them out and they themselves will assume the leverage that will require them to be bailed out. No. 2, simple fairness: If Britain’s banking interest can claim a right to the accommodation of the Bank of England, why shouldn’t the shipping interest, the construction interest, the railroads, the agricultural interest? Shouldn’t all actors be equally entitled to benefit by any favors?”
Good points and questions, but the threat of financial collapse always trumps them.
The Paulson Plan, like similar efforts in other countries, is now focused on providing new capital. It already has the problem Hankey foresaw: lots of people who would like to get in line with hands out, palms upward.
Of course what is really happening in all such programs is that the taxpayers are being made into involuntary equity investors. How should they (we) be given fair treatment as investors?
In the aggregate, the investments the taxpayers are involuntarily making ought to have an expected positive return, in exchange for the risks they are taking. Secretary Paulson has suggested that the program might make a profit, and it might. There is both a yield on the preferred stock investments being made, as well as equity upside potential from warrants with strike prices reflecting depressed current stock prices. (Recall that the warrants the government got in the Chrysler bailout of a generation ago did indeed result in a significant profit.)
I propose two fundamental improvements in the bailout plan to make it reflect the interests of the real investors.
First: All the activities of the TARP program should be isolated in a separate accounting entity. All investments and other assets, all related debt and other liabilities, all expenses, and all income should be clearly measured as if TARP were a corporation. An audited balance sheet and income statement should be regularly produced. Then TARP’s performance could be judged not only by the administration as the program’s operator, but by the Congress as overseer, and most importantly, by the taxpayers as investors.
Second: 100 percent of any profit should be earmarked as explicit dividends to the taxpayer-investors. Such dividends might be in the form of cash or specific tax credits. This would be a well-deserved recompense to the majority of the citizens who bought houses they could afford, paid their mortgage loans on time, did not engage in leveraged speculation, paid their taxes, and then took all the risk of the bailout efforts.
Prudence, moderation, and virtue are their own reward, yes, but how about some dividends, too, if the bailout makes money?
Getting Back to Normal
The bubble inflates because enough people believe that the good times of expanding credit and rising asset prices will last forever, but of course they don’t. Similarly, in the midst of the bust, it seems like the bad times of scarce credit and falling prices will last forever, but of course they won’t. The whole point of the expansion of the government’s balance sheet and the bailout investments is to get to the other side of the bust, which we will, although not without further difficulties and losses.
Financial cycles cannot be avoided any more than business cycles can. After every big bust, various politicians and theoreticians announce mechanistic ideas and enact regulations to make sure the problems “never happen again”—but in time, they always do.
This is not to say we should not attempt to improve the financial system, and mortgage finance in particular. Here are four suggestions:
Leverage at the mortgage borrower level is expressed as the Loan-to-Value ratio or “LTV.” An LTV of 80 percent means the homebuyer has borrowed 80 percent of the price of the house and made a down payment out of savings of 20 percent. An LTV of 100 percent, not uncommon during the housing bubble, means the buyer has borrowed all the money. Did it make sense to make these loans? Only if you believed house prices would always go up.
There is a strong and reliable statistical relationship between LTVs and mortgage defaults: The higher the LTV, the higher the defaults. When a house price declines, the LTV automatically increases.
In a housing boom, as house prices rapidly inflate, the risk that they will subsequently fall increases. So in a rational system, LTVs would be lowered as house prices accelerated over their trend. What in fact happens in the boom is that as rising prices induce optimism in both lenders and borrowers, LTVs tend to rise, up to and including 100 percent—exactly the opposite of what should happen.
Reversing this perverse LTV behavior, if one could, would make for a sounder housing finance system.
2. Loan Loss Reserves
We should return to the old-fashioned idea of building loan loss reserves in good times. As an old banker told me long ago, “Bad loans are made in good times.” This is just the time to be building reserves against the inevitable vicissitudes of any firm bearing credit risk.
Spain, now involved in its own housing bust, is glad that in 2000, it forced its banks to begin a system of countercyclical “dynamic provisioning,” building up reserves far beyond current losses to prepare for possible future losses. This system, old fashioned as it is, is becoming the international model.
Unfortunately, in this country the SEC went in just the opposite direction from what the Spanish regulators did. It actively opposed building large loan loss reserves, because it regarded them as undesirable “earnings management,” which would understate profits in good times, while providing a cushion for bad years. The result was to overstate profits in the bubble and to have insufficient cushion in the bust.
The reality of all businesses built on credit risk is that they experience, on a cyclical basis, periods of high losses, often far beyond expectations—like now. If you don’t reserve against this reality, you create—as American official accounting did—illusory profits in the good times. These in turn trigger bonuses and also induce financial firms to reduce equity through buybacks of their shares. Alternately stated, in the good times financial firms book what are really insurance premiums for bearing risk instead as profit in advance, while the risk builds without prudent enough reserves. This is obvious, now that the risk has come home to roost.
3. Credit Risk Retention
We should combine mortgage securitization with credit risk retention by the mortgage originators that perform the credit underwriting and make the credit decisions.
A prime lesson of the 1980s savings and loan collapse was that for financial institutions to keep long-term fixed rate mortgages on their own balance sheets was extremely dangerous in terms of interest rate risk, although it was not a problem in terms of credit risk. The answer was to sell the loans to bond investors through securitization and divest the interest rate risk to those better able to bear it. As a side effect, credit risk was also divested.
In the wake of the mortgage bubble and bust, it is now loud and clear that divesting mortgage credit risk created huge problems of its own, breaking the alignment of incentives between the lender making the credit decision and the ultimate investor actually bearing the credit risk. Some commentators have referred to the good old days when the savings and loans kept the loans themselves—displaying either ignorance of the disaster that resulted or short memories.
The right synthesis of the historical lessons is for securitization to continue to address interest rate risk, while simultaneously encouraging retention of significant credit risk by the original mortgage lender. The superiority of this combination makes it well worth while to try to overcome the regulatory and accounting obstacles involved.
When we are past the bust, we must shrink back the inflated government balance sheet, so the private financial balance sheet can resume its normal functioning.
That will be the challenge of two years or so from now.
Alex J. Pollock is a resident fellow at the American Enterprise Institute.
Illustration by Darren Wamboldt/Bergman Group.