Click here to view this Outlook as an Adobe Acrobat PDF.

December 2007

It is important at the outset to define the terms "financial firms" and "too much risk." By "financial firms," I mean commercial banks, investment banks, brokerages, and insurance companies that solicit and manage funds for the public. By "too much risk," I mean actions undertaken by managers of financial firms that result in substantial losses for the shareholders (owners) of such firms. On an aggregate level, I call "systemic risks" those that emerge when regulators and policymakers are forced to choose between either reinforcing (with bailouts) the venturesome investing that created the problem or allowing substantial damage to depositors and shareholders in financial firms, and possibly to the economy as a whole.

An even broader form of excessive risk is operative when central banks are forced to choose between abandoning their commitment to maintaining low and stable prices and allowing a recession to occur. This last, most poignant, manifestation of "too much risk" has arisen for the U.S. Federal Reserve. The outlook for the next six months suggests a strong possibility that growth will slow sharply because of credit problems tied to losses by financial intermediaries, while inflation will rise because of strong upward pressure on energy costs. Other central banks may find themselves in the same stagflation bind. European growth is slowing while inflation pressure is rising, while in Japan, growth has slowed sharply.

The Principal/Agent Problem

The principal/agent problem--where interests of managers of financial intermediaries diverge from those of their shareholders--has been clearly illustrated by the widely publicized cases of Merrill Lynch and Citigroup. Substantial errors in initially reported third-quarter earnings, when revealed, resulted in the "resignation" of the CEOs of both institutions--Stan O'Neal of Merrill Lynch and Charles Prince of Citigroup. For those who review these sad cases, it is tempting, but untrue, to say that financial firms take too much risk because they are managed by foolish people. The temptation arises from revelations in October and early November that both Citigroup and Merrill Lynch, to mention just the most glaring recent examples, had to revise third-quarter earnings reports downward by billions of dollars only days or weeks after initial results were first reported. The revisions were so large that the stock prices of both institutions fell sharply while financial sector stocks retreated broadly. Consequently, the CEOs of both institutions were forced to step down.

The terms of departure for both CEOs, especially Stan O'Neal's, undercut the charge of management stupidity while reinforcing the notion that both leaders had taken on too much risk for their firms. While investors in Merrill Lynch could check a website that records the scores of serious golfers and learn that Stan O'Neal played twenty rounds of golf between August 12 and September 30, 2007--as the first phase of the credit crisis was raging--they could also observe his $48 million bonus in 2006, which made him the second-highest paid CEO on Wall Street, together with his exit package estimated at $150 million, and conclude that this man was no idiot. He had negotiated a compensation package that paid him a total of more than $50 million in 2006 to take extraordinary risks and then paid him again in 2007 even when those risks resulted in billions of dollars of losses for Merrill Lynch shareholders.

O'Neal demonstrated that being a good golfer was more important to getting ahead at Merrill Lynch than doing anything about risk management, and Merrill's board of directors apparently agreed by awarding him a compensation package that paid off handsomely whether the risks turned out to be justified or not.

Citigroup's hapless CEO, Charles Prince, fared less well than O'Neal in the departure-reward department. However, in view of his apparent lack of attention to risk management as head of one of the world's largest banks,[1] he has not fared badly. According to a front-page Wall Street Journal article on November 9, Prince "got the news" on October 27 that Citigroup was facing billions in new losses ($10 billion for the third quarter alone to be exact) from deteriorating credit markets on top of the huge ($6.4 billion) write-down it had announced two weeks earlier. (The initial announcement of Citigroup's third-quarter losses was made on Monday, October 15, just as the Treasury Department's "super SIV" plan aimed at shoring up Citigroup's $80 billion exposure to off-balance-sheet "special investment vehicles" was announced.)

The shock to Prince, not to mention to his CFO Gary Crittenden, who passed the "news" to him on October 27, was that Citigroup's chief accountant had informed Crittenden that the company would be forced to disclose losses in its quarterly report to the Securities and Exchange Commission. One may infer, both in the case of Citigroup and Merrill Lynch, that the firms' outside auditors, who will have to sign off on financial reports at year end and who cannot be indemnified by their customers from prosecution for false and misleading statements, provided reminders that assumptions that had been made about the value of numerous assets, particularly those tied to mortgages, would not withstand the scrutiny of applying basic accounting principles regarding their valuation in financial statements.

In Prince's case, as head of Citigroup, the principal/ agent problem was compounded by what, if we are to judge by the Wall Street Journal report, can only be termed complete detachment from the realities of credit markets since July. Citigroup's exposure to substantial additional losses on mortgage-related assets should not have come as news on October 27, a full three months into the credit crisis.

Systemic Risk and Moral Hazard

The transition from the principal/agent problem, which concerns too much risk at the firm level, to the broader systemic risk problem of moral hazard and the threat of stagflation facing the Federal Reserve is tied to the underlying cause of the problems dogging the CEOs of financial institutions worldwide. The real estate that underlies mortgage-based assets that have been repackaged and releveraged into mortgage-backed securities and distributed worldwide to banks, investment banks, brokerage houses, and even to money market funds is a rotting asset. Its price has fallen in the U.S. market for residential real estate about 5 percent over the past year.

The fundamental problem for the financial sector is twofold. First, the trillions of dollars of subprime mortgages and higher-rated assets tied to real estate were created on the assumption that U.S. home prices do not fall persistently. They had not, until 2007, dropped on a year-over-year basis since the Great Depression. But now they are falling at a 5 percent annual
rate. Second, the drop in home prices looks likely to accelerate--probably to a negative 10 percent year-over-year rate or more--meaning that further write-downs will be required on mortgage-based assets.

A substantial part of the problem for the balance sheets of financial intermediaries, at least in the United States, lies with the flawed accounting procedures employed to deal with opaque or hard-to-price financial assets. These assets of banks and investment banks are placed in three categories: levels I, II, and III. Most problematic is level III, at which, in the absence of any actual prices or basis to infer value, assets are valued by reference to the owners' own models. In other words, banks get to say what their assets not priced in the markets are worth. Level III assets represent a substantial share of equity capital for Merrill Lynch (70 percent), Lehman Brothers (160 percent), Citigroup (106 percent), JPMorgan (45 percent), Morgan Stanley (255 percent), Goldman Sachs (184 percent), and Bear Stearns (156 percent). For these institutions, level III assets rose sharply during the third quarter, on average by 40 percent.[2]

The difficulty of valuing level III assets has serious implications. If financial institutions, faced with providing earnings reports, simply dump all of their questionable assets into the level III category and use their own "models" to value them, questions about the reported earnings and, in some cases, about the solvency of such institutions arise. This is especially true when the quantity of level III assets exceeds the institution's equity capital base. (The average ratio for the above-mentioned major financial firms is nearly 140 percent.) A 20 percent write-down of level III assets could wipe out almost a quarter of equity capital while sharply reducing reported earnings. The sharp rise in opaque level III assets during the third quarter, when mortgage security problems erupted, does not inspire confidence.

As time goes by, it is more likely that greater transparency will be achieved regarding the value of the heterogeneous collection of level III assets that looms large on the balance sheets of most financial institutions. The best way to establish such value--in fact, the only market-based way--would be to allow auctions of level III assets. Plenty of capital is available to purchase the level III assets at prices that accurately reflect the risks tied to activities that underlie those assets. Unfortunately, most efforts to date have been directed toward avoiding valuation of such securities rather than accurately establishing their value so that financial institutions can take the hit and then move on.

Policy Dilemma for the Fed

The fundamental dilemma facing financial firms owning mortgage-backed securities whose values are shrouded in mystery defines most clearly the notion of "too much risk." The dilemma following episodes of excessive risk-taking is this: solving the underlying problem entails a policy response that encourages taking more risk in the future and, ultimately, implies an even larger risk bubble yet to come.

More specifically, if government actions allow further delay in accurately valuing level III assets, or if government actions artificially inflate their value, then the excessive risk-taking that led to the large stock of level III assets in the first place is reinforced. Even more broadly, if, in the name of avoiding recession, the Federal Reserve acts to push up prices by sharply increasing liquidity in the economy, its long-standing commitment to inflation control will be compromised, as will the benefits of higher, more stable growth that have accompanied lower inflation since 1982. An inflation-generating response to excessive risk-taking that has resulted in the collapse of a housing bubble will encourage further risk-taking, larger future bubbles, and still higher inflation.

The Federal Reserve, and chairman Ben Bernanke in particular, has been clear that the Fed will not risk its goal of stable inflation even if doing so would be necessary to avoid a recession tied to a credit crunch resulting from excessive risk-taking in the mortgage sector. That said, the situation facing central bankers will get tougher in coming months. Bank of England governor Mervyn King has already discovered that with a poorly designed deposit insurance system, events can run out of control and force a response that may exacerbate moral hazard problems. As explained in the October 2007 Economic Outlook, Northern Rock, a mortgage lender in the United Kingdom, financed itself largely in the commercial paper market rather than primarily with deposits. Still, as a mortgage lender, Northern Rock was a depository institution and thereby served as a savings vehicle for many Britons. When panic erupted that Northern Rock would be unable to finance itself in the commercial paper market, lines formed in September outside of its branches and, in order to avoid a nationwide run on banks, England's financial authorities were forced to step in and keep Northern Rock afloat by guaranteeing the deposits--at least for a time--of Northern Rock's savers.

As U.S. growth begins to slow and perhaps even turn negative at the end of this year and early next year against the backdrop of an intensifying presidential campaign, the pressure on the Federal Reserve to reinforce riskier behavior by additional easing, even as inflation rises, will become intense. The best outcome would be a stable inflation rate that enabled the Fed to respond to the weak economic data and thereby to help cushion the impact on households of a continued fall in real estate prices. The worst outcome would see the Fed forced to ease by a credit crisis that unfolds more rapidly than an economic slowdown or moderating inflation appears. Financial markets have become highly volatile as they oscillate between pricing a benign scenario, in which the Fed eases with no U.S. recession and both inflation and credit markets are stable, and a disaster scenario, in which the Fed is forced by rising inflation pressures to delay easing until the credit crunch worsens and pushes the economy into recession.

Avoiding Systemic-Risk Traps

Answers to the question of why financial firms take too much risk suggest some changes going forward, both at the firm level and at the economy-wide level, to avoid the principal/agent problem and the systemic-risk, moral hazard problem. First, boards of directors of financial firms need to design compensation packages that align the interests of managers closely with those of the shareholders. Second, financial innovation is a good thing, but it needs to be monitored more closely, especially when it leads to a massive magnification of risk attached to aggressive lending that results in a real estate bubble.

Third, lenders who are allowed merely to originate mortgages and then to sell the resulting mortgage contracts into highly complex and highly leveraged packages of securities need to be given incentives to take some of the mortgage risk onto their own books. Fourth, banks should no longer be afforded the luxury of arbitrary model-based valuation of level III assets. Transparency on asset values is essential. The notion that credit-derivative obligations and other such mortgage-related securities would never be traded and, in fact, would only be held to maturity and valued at 100 cents on the dollar encouraged many banks to overinvest in that category.

Finally, central banks need to reinforce their commitment to stable inflation by refusing to cut interest rates while inflation is rising, even as growth slows sharply. The ominous rise in most measures of U.S. inflation expectations in mid-November suggests that the Federal Reserve will be sorely tested on this important point in coming weeks. Central banks also need to emphasize that commercial and investment banks that jeopardize their balance sheets and earnings by acquiring too many risky assets will not be bailed out by inflationary monetary policy.

Immediate Fed Problems

The Fed's dilemma over the next several quarters is underscored by its recent actions in September and October, which resulted in a 75 basis-point reduction in the federal funds rate undertaken to offset, in advance, the Fed's estimated negative impact of credit problems on economic growth. The Fed has, in effect, said that it expects growth to slow in the fourth quarter--perhaps to 1.5 percent or lower. Having announced that it has already taken steps in response to that expected slowing, the Fed may have to await even lower growth or a continued slowdown into 2008 before it eases further. Yet markets have already priced in further easing. The possibility of further easing is conditional on stable-to-lower core inflation over the fourth quarter of 2007. The risks that core inflation may rise--and along with it, inflation expectations--could further limit the Fed's easing response to slower growth.

The big risk, as noted already, is that the slowing economy will exacerbate credit problems as house prices fall more rapidly, requiring a still larger Fed rate cut to offset the negative impact flowing from intensifying credit problems to consumption, investment, and growth. Unfortunately, part of the reason that financial firms take what turns out to be too much risk is the bet that if credit problems intensify enough and threaten a dangerous, downward spiral, whereby worse credit leads to lower growth which in turn leads to worse credit problems, the Fed will cut rates further--irrespective of inflation risks.

The case for further rate cuts under such circumstances would be that a sharper-than-expected growth slowdown would contain inflation even if the Fed cut rates sharply in response to such a slowdown. That is a risky bet that no central bank would want to take, but, looking ahead, it is important to remember that such a difficult choice is a direct byproduct of too much risk-taking by financial firms. Financial firms, not central banks, need to learn to live with the consequences of bad risk management.

John H. Makin (jmakin@aei.org) is a visiting scholar at AEI.

 Click here to view this Outlook as an Adobe Acrobat PDF.

Notes

1. Comparisons among banks are difficult in view of the massive uncertainty about the value of securities tied to mortgage loans. We will return to this problem later in our discussion.

2. Fortune, November 12, 2007; figures are derived from the banks' financial statements.