Thank you, Mr. Chairman. Financial markets have become much more turbulent since the last meeting—especially in subprime mortgages and associated securities, in U.S. and global equities, and in foreign exchange markets. The good news is that markets have generally remained liquid and well functioning, with a minor exception on the New York Stock Exchange on February 27. Moreover, there are few signs of significant contagion from the subprime mortgage market into the rest of the mortgage market or from subprime mortgage credit spreads to corporate credit spreads more generally. In general, the debt markets have been mostly unruffled by recent developments.
I plan to focus my attention on four major market developments. First, the substantial turmoil in the subprime mortgage market—I talked about the risk that this market might unravel at the January FOMC meeting; that certainly occurred more quickly and more forcefully than I anticipated. Second, I want to talk a little about the decline in U.S. equity prices and the accompanying rise in actual and implied price volatility. Third is the sharp correction in the so-called “carry trade” in foreign exchange markets. The low interest rate currencies such as the yen and the Swiss franc have appreciated, with the greatest moves coming against their higher-yielding counterparts. Finally, I’ll talk a bit about the sharp downward shift in market expectations about the path of the federal funds rate target over the next year and a half. Two key questions motivate my comments. First, is the market turbulence driven mainly by fundamental developments, or does it reflect mainly a shift in the risk appetite of investors? Second, what is the ongoing risk of contagion from the market area that has experienced the most stress—the subprime mortgage market—to other markets?
Regarding the subprime mortgage market, the deterioration appears driven mostly by fundamental developments. As you know, the delinquency rates for subprime adjustable-rate mortgages have risen sharply. In contrast, as shown in exhibit 1 of the handout, there has been little change in delinquency rates for fixed-rate mortgages. Most significantly, delinquency rates for the 2006 vintage of subprime adjustable-rate mortgages have climbed unusually quickly. As shown in exhibit 2, the last vintage that went this bad so fast was the 2001 vintage, and that had a much different economic environment—one characterized by a mild recession and a rising unemployment rate. The deterioration in the quality of subprime mortgage credit has led to a sharp widening in credit spreads for the ABX indexes. The ABX indexes represent the cost of default protection on a basket of collateralized debt obligations that are backstopped mainly by subprime mortgages. As shown in exhibit 3, although this widening has been most pronounced at the bottom end of the credit quality spectrum (BBB-minus and BBB), it has rippled upward to the higher-rated tranches that are better protected. Exhibit 4 shows how the credit deterioration initially registered in the ABX indexes as market participants sought to buy protection. In milder form, this deterioration also registered in the underlying collateralized debt obligations and asset-backed securities. The widening of the credit spread in the ABX indexes was probably exaggerated by the fact that there was an asymmetry between the many that were seeking loss protection and the few that were willing to write protection. This can be seen in two ways. First, as shown in exhibit 4, the spread widening was more pronounced in the ABX index than in either underlying collateralized debt obligations or asset-backed securities. Second, as shown in exhibit 3, the ABX spreads have come down a bit from their peaks even as the underlying market for subprime mortgages, as reflected in the ongoing viability of many mortgage originators, has continued to deteriorate.
The deterioration in the subprime market has undermined the economics of subprime mortgage origination and securitization. This is especially true for those mortgage originators with poorer underwriting track records. Their loans can no longer be sold at a sufficient premium to par value to cover their origination costs. In addition, the costs that they must incur to replace loans that have defaulted early have increased sharply. In several cases, these difficulties have caused banks to pull their warehouse lines of credit. Several of the large monoline originators are bankrupt, distressed, or up for sale—they are highlighted in red in exhibit 5. Moreover, several of the diversified lenders, such as HSBC, have indicated that they are tightening credit standards and pulling back from this sector. The result is that the volume of subprime mortgage originations is likely to fall sharply this year—perhaps dropping one-third or more from the 2006 rate of slightly more than $600 billion. This tightening of credit availability to subprime borrowers is likely to manifest itself through a number of channels. These channels include (1) a drop in housing demand, as borrowers who would have been able to get credit in 2006 no longer qualify under now toughened underwriting standards; (2) an increase in housing supply, as the rate of housing foreclosures increases (notably, the Mortgage Bankers Association reported last week that the rate of loans entering the foreclosure process in the fourth quarter of 2006 reached a record level of 0.54 percent, the highest level in the history of the thirty-seven-year-old survey); and (3) additional downward pressure on home prices, which in turn threatens to increase the magnitude of credit problems, delinquencies, and foreclosures.
In considering these channels, it is important to emphasize that the credit strains in the subprime sector are unlikely to have peaked yet. The reset risk on the adjustable-rate portion of the subprime loans originated in 2005 and 2006 will be felt mainly over the remainder of 2007 and 2008. Most of the adjustable-rate loans are fixed for two years at low “teaser” rates. When yields adjust upward once the teaser rate period is over, some borrowers may have insufficient resources to service these debts. The good news—at least to date—is that spillover into the alt-A mortgage and conforming mortgage areas is very mild, both in terms of credit spreads and in terms of loan performance. Although there has been some rise in delinquency and foreclosure rates for higher-quality residential mortgages, these rates are still low both qualitatively and historically. Moreover, there is little evidence that the subprime problems have hurt mortgage loan volumes. For example, the Mortgage Bankers Association index of mortgage applications for purchase has increased in the past three weeks.
Turning next to the U.S. equity market, it is less clear-cut whether the decline in prices and the rise in volatility are fundamentally based. As several observers have noted, equity valuations do not appear to be excessive. If that is the case, then why have equities been more turbulent than corporate and emerging-market debt, for which spreads remain unusually narrow? Although this point is legitimate, two fundamental developments that make U.S. equity prices less attractive deserve mention. First, equity analysts have been reducing their earnings forecasts for 2007. Although the top-down view of the equity strategists for the S&P 500 index has not changed much, on a bottom-up basis, earnings expectations have dropped sharply. As shown in exhibit 6, the aggregate forecasts of the individual sector analysts now indicate a growth rate in S&P 500 earnings for 2007 of about 6 percent, down from about 9 percent at the beginning of the year. In contrast, S&P 500 earnings have grown at an annual rate of more than 10 percent for four consecutive years. It should be no surprise that falling earnings expectations could weigh on equity prices. Second, uncertainty about the growth outlook has increased. This shows up clearly, for example, in our most recent primary dealer survey. Because greater uncertainty about the growth outlook presumably implies greater risk, the rise in uncertainty should—all else being equal—result in lower share prices. In contrast, it is easier to explain the modest widening of corporate credit spreads. In theory, lower share prices and higher volatility imply a greater risk of default, which should imply wider credit spreads. Corporate credit spreads have behaved in a manner consistent with this. Josh Rosenberg from the research group at the Federal Reserve Bank of New York recently investigated this issue. He found that the spread widening in the high- yield corporate debt sector was consistent with past periods in which the implied volatility for equities rose sharply. Exhibit 7 summarizes one key result. The widening in the BB-rated corporate spreads in the week after the February 27 retrenchment was of a magnitude similar to that of other instances in which implied equity-price volatility as measured by the VIX index rose sharply. In the most recent episode, the VIX index rose 848 basis points, and the BB corporate spread rose 27 basis points. This rise compares with an average rise of 21 basis points in the BB spread in the ten cases in which the VIX rose most sharply. The rise in the most recent episode is well within the range of historical experience.
In many other areas in which asset prices have moved sharply, risk-reduction efforts appear to have played the biggest role. For example, in the foreign exchange markets, the biggest currency moves were in the currency pairs associated with so- called carry trades, such as the yen and Swiss franc for the low-yielding currencies and the Australian and New Zealand dollar for the high yielders. Exhibit 8 indicates the change in the yen versus the Australian dollar, the New Zealand dollar, the euro, the British pound, and the U.S. dollar during three separate periods—the week before the February 27 stock market selloff, the week of the stock market selloff, and the past two weeks. The high-yielding currencies appreciated the most during the run-up to the February 27 selloff, fell the most during the February 27 week, and have recovered the most against the yen over the past two weeks. The changes in speculative positioning in foreign exchange future markets tell a similar story. Exhibit 9 shows the change in the share of the open interest position held by participants in the noncommercial futures market. Over the past few weeks, net short positions as a percentage of the overall open interest in the yen have dropped, and long positions in the British pound and Australian dollar have dropped.
An examination of how Treasury yields, stock prices, exchange rates, and credit spreads have moved also indicates that risk-reduction efforts have been important. Exhibit 10 shows the correlation of daily price and yield movements in 2007 before February 27. As one can see, the correlations were quite low. In contrast, the correlation matrix in exhibit 11 shows the correlation of daily price moves for the period beginning on February 27. Most of the correlations have climbed sharply, suggesting that risk positioning is driving price and yield movements.
Finally, short-term interest rate expectations have shifted substantially since the last FOMC meeting. As shown in exhibit 12, near-term expectations have shifted, with market participants now expecting a modest reduction in the federal funds rate target by late summer. However, the federal funds rate futures curve is still above the curve at the time of the December FOMC meeting. In contrast, longer-term expectations have shifted more sharply, with a larger move toward easing. As shown in exhibit 13, the June 2008/June 2007 Eurodollar calendar spread is now inverted by about 60 basis points. This calendar spread is more inverted than it was at the time of the December 2006 FOMC meeting.
Compared with the shift in market expectations, the forecasts of primary dealers have not changed much. Exhibits 14 and 15 compare dealer expectations with market expectations before the January FOMC meeting and before this meeting. The horizontal bold lines represent market expectations. The blue circles represent the different dealer forecasts. The green circles represent the average dealer forecast for each period. The two exhibits illustrate several noteworthy points. First, the average dealer forecast has not changed much since the January FOMC meeting—the green circles in the two charts are in virtually the same position. Second, the amount of dispersion among the dealers’ forecasts has not changed much—in fact, the range of the blue circles is slightly narrower currently. Although many dealers now mention that their uncertainty about the growth outlook has increased, that does not appear to have been reflected in their modal forecasts. Third, there is now a substantial gap between the dealers’ average forecast and market expectations—the gap between the horizontal bold lines, which represent market expectations, and the green circles, which represent the average dealer’s view, has increased. Why is there a large gap between the dealers’ forecasts and market expectations? I think there are three major explanations. First, the dealers’ forecasts are modal forecasts and do not reflect the downside risks that many dealers now believe have emerged in the growth outlook. Second, dealer forecasts often lag behind economic and market developments. Only when “downside risks” grow big enough to pass some threshold are dealers likely to alter their modal forecasts. Third, some of the downward shift in market expectations may represent risk-reduction efforts. An investor with speculative risk positions that would be vulnerable to economic weakness might hedge these risks by buying Eurodollar futures contracts. This hedging could push the implied yields on Eurodollar futures contracts lower than what would be consistent with an unbiased forecast of the likely path of the federal funds rate.
Nevertheless, the potential gap between market expectations and the Committee’s interest rate expectations may pose a bit of a conundrum for the Committee. If the Committee were to shift the bias of its statement in the direction of neutral, market expectations with respect to easing would undoubtedly be pulled forward and might become more pronounced. After all, most dealers expect that the Committee will not change the inflation bias of the January FOMC statement. In contrast, keeping the bias unchanged in order to keep market expectations from shifting further in the easing direction might be inconsistent with the Committee’s assessment of the relative risks regarding growth and inflation. If the Committee were to keep the bias unchanged even when its views had changed, the communication process might be impaired.
On a housekeeping note, I wish to bring to the Committee’s attention the changes to the “Morning Call” with the Trading Desk. They were discussed in a memo distributed to the Committee last week. Under the new format, which we plan to implement on Thursday, the call will be open to all members of the Committee, and you will have the option of participating in the 9:10 a.m. discussion of reserve management issues, the 9:20 a.m. portion covering recent developments in global markets, or both portions. The March 15 memo outlines the new procedures for joining these calls. Finally, there were no foreign operations during this period. I request a vote to ratify the operations conducted by the System Open Market Account since the January FOMC meeting. Of course, I am very happy to take questions.