|
Treasury yields have pushed sharply lower since early April, as the European currency crisis has combined with fears of a double-dip recession in the U.S. As shown in Exhibit 1 below, the move in the 10-year Treasury pushed through the lower 2-standard deviation bands on two different occasions, most recently in mid-May. As I’ll discuss later, the downward push did not appear to trigger large-scale duration buying, which would have pushed the 10-year yield solidly through the 3.25% level.
The Treasury yield curve (2-10s) has flattened over the past two months, and at around +250 is well off its all-time highs of around +290 seen earlier this years. (See Exhibit 2.) I look for the curve to continue to flatten during the second half of the year. I think there are separate pressures on both short- and long-term Treasury rates. In my mind, the long end will be impacted a continued soft economy and job market, along with the continued lack of inflationary pressures. The short end of the Treasury market, by contrast, will be pressured by wider funding costs triggered by the situation in Europe. Exhibit 3 shows that LIBOR has widened sharply to other measures. The chart shows that the spreads of LIBOR versus both CMT and OIS have widened sharply since early April, when the situation in Greece began capturing investors’ attention. 
The spread between 2-year Treasuries and LIBOR, in turn, is a key factor in both the level of the 2-year yield and the shape of the curve. Exhibit 4, for example, shows the 2-10 Treasury spread superimposed over the spread between 2-year Treasury yields, with periods of sustained Fed tightening shown in the shaded areas. The chart suggests that there is a fairly strong relationship between the funding spread for 2-year Treasuries and the shape of the yield curve. Therefore, if the European situation muddles along or worsens, there will be continued pressure of short Treasury yields. In combination with the downward pressure on longer Treasury rates, I think the yield curve can flatten by another 50 basis points or more between now and the end of 2010. 
Agency MBS spreads have recently been fairly well contained. Exhibit 5 shows the 30-year current coupon spread over interpolated 5-10 year Treasuries and swaps, along with their averages over the second half of 2009. While the current coupon spread over Treasuries is off its lows (reached during the last stages of the Fed’s purchase program in March), it is inside of its H2 2009 average. Moreover, spreads to swaps are virtually on top of their H2 2009 average, and have not moved decisively since last fall. This suggests that agency MBS spreads have pretty much tracked the tightening in intermediate swap and (by implication) credit spreads over the last 6-9 months. It also raises the issue with respect to the factors driving the tightening in intermediate and long swap spreads over the past year; arguably, the technical factors impacting swaps are also affecting MBS valuations. 
One other interesting note is that spreads have not gapped wider since the Fed ceased their purchases in March. I’d argue that the steady source of demand by the Fed (for a nice round $15-20 billion per week) has been replaced by other sources of funds. These sources are less consistent and more opportunistic than the Fed, which is why the volatility of spreads has increase since last winter. Exhibit 6 shows that the 60-day standard deviation of the current coupon spreads, which had been declining steadily since the Fed purchase program began, ticked up immediately upon its completion. This is consistent with the notion of continued, albeit less consistent, demand for agency MBS, especially as long as GSE reform remains well over the horizon.
 |
Subscribe to this comment's feed
