MBS Remain Somewhat Insulated From Rates Market Volatility
After a few relatively quiet weeks, interest rate volatility exploded again last week, triggered by European turmoil and the MF Global bankruptcy. Using a relatively short (45 day) lookback period, realized volatility (i.e., the standard deviation of daily rate changes) for the 10-year Treasury increased sharply over the past 1 1/2; weeks, from 7.4 basis points to 8.4 bps. Without getting too technical, this reflects the sharp run-up and subsequent retracement in yields. For example, the 10-year yield bottomed out at just over 1.75% on 10/3; it reached as high as 2.40% on October 27th and is now hovering around 2%.
The spike in volatility was triggered by the announcement of an interim settlement to the European debt crisis, which subsequently crumbled when the Greek PM announced that it would be put to a vote with the Greek electorate. As with previous agreements, this "settlement" had the aroma of a highly clever solution that was unlikely to make a long-term difference; it's hard to understand the notion of a "voluntary" writedown of half the value of a group of assets without it constituting a "default." Until the vote is actually taken, more market volatility (as comments and statements by the players are parsed) is likely.
What's interesting is that the spike in realized volatility has not yet drifted over into either the options markets or the MBS sector. "Implied" volatility (which is essentially a "plug" that allows for options to be quoted on an apples-to-apples basis) has not drifted higher despite the see-sawing in market yields. In theory, implied and realized volatility should move together; however, in practice the two measures often decouple. This has arguably given some support to MBS valuations, since OAS models use implied vols as an input. The 30-year current coupon spread tightened about 8 basis points versus interpolated Treasuries last week, although it has widened a few basis points this week.
An interesting discussion swirling around the MBS community relates to the likelihood of a sharp pickup in prepayment speeds from the HARP II initiative. The latest discussions center on the waiver of rep & warranty claims on loans that are refinanced under the program. The discussion goes as follows: can large servicers limit their exposure to buyback claims from the GSEs by simply refinancing all eligible loans? It seems like a fairly tempting option. My view, however, is that originators are already having operational problems processing the current level of applications; I'm also skeptical that lenders can ystematically contact large numbers of eligible borrowers. Particularly with the turmoil surrounding BofA's withdrawal from the Correspondent Lending market, I don't think it heralds a significant spike in activity, even if refinancing large numbers of these borrowers would make economic (and legal) sense.
The Fed concluded the two-day meeting of the Open Market Committee with no change to monetary policy. However, they did leave the door open to more actions by the Fed to stimulate growth, and indicated that the purchase of more MBS was a viable option in the future. My gut is that another round of purchases would have little impact on housing, except if they were to commit to buying sizeable amounts of 30-year 3.0s. The fact that there is virtually no liquid market for 3% coupon passthroughs effectively puts a floor under primary mortgage rates; unless the base servicing requirement of 25 basis points is relaxed, there is currently no reliable outlet for loans with rates below 3.75%.