I apologize in advance for the somewhat wonkish title and substance of this post. However, today's subject is getting a lot of buzz among fixed income professionals, so I thought you might be interested.
Although some of the features of mortgage-backed securities (MBS) make them an attractive asset class for many bond investors, they also have some unique characteristics that present a challenge for fixed income managers.
MBS are fixed income securities that are backed by home mortgages which all pay the same rate of interest. The vast majority of MBS are guaranteed for timely payment of principal and interest by the quasi-government agencies Fannie Mae and Freddie Mac. MBS guaranteed by Ginnie Mae carry the full guaranty of the United States Treasury.
As anyone who has ever had a mortgage in the U.S. knows, homeowners have the right to refinance their mortgages if interest rates fall. On the other hand, if mortgage rates rise, the homeowner's rate will not change (unless they have a adjustable rate mortgage).
MBS tend to trade at a higher yield than straight Treasury or Federal agency debt for a couple of reasons.
First, they are relatively cumbersome instruments: every month, payments of principal and interest are passed through to the security holder.
Second, unlike conventional bonds, the interest rate sensitivity of MBS rises as interest rates increase, but declines when rates fall.
This latter point is probably best illustrated by an example. Say you purchase an MBS that is secured by fixed rate home mortgages all paying a 5% interest rate. If rates go to 3%, many homeowners will find it attractive to refinance their existing mortgage. When this happens, the holder of the MBS will receive their principal back at par, even though they would prefer to keep receiving the higher rate.
On the other hand, if mortgage rates go to, say, 8%, homeowners will be perfectly content to hold onto their 5% mortgage, even though in this case the MBS investor would rather get their money back and invest at higher yields.
This "heads I lose, tails you win" challenge for MBS holders is one of the primary reasons that they offer better yields than other fixed income securities with similar credit characteristics.
MBS holders bear all of the interest rate risk, since their bonds are repaid at a time that is the least attractive from their perspective or do not get paid back when rates are rising.
In mathematical terms, this unattractive feature of MBS is called "negative convexity".
Without going into specifics (I can hear you breathing a sigh of relief), this simply means that MBS interest rates moves in the opposite direction of what would be preferable from an investment perspective.
OK, if you've made it this far, now we come to the reason I went through all of this.
There are $4.8 trillion of MBS outstanding. Traditionally these have been held by commercial banks, mutual funds and insurance companies, although investment banks routinely hold them in the course of their normal trading activity.
A number of large MBS investors try to hedge the interest rate sensitivity by either selling longer-dated Treasury securities (when rates are rising) or buying Treasurys (when rates are falling). Because the MBS market is so large (about 29% of the taxable bond market), any large movements in interest rates can be magnified by hedging activity.
Interest rates have been steadily rising since the end of April of this year:
Rates have been rising for a myriad of reasons, but to someone trying to hedge their MBS portfolio this is a real problem.
As yields have risen, the interest rate sensitivity (i.e. duration) of the MBS market has increased, and prices have been falling. As rise have risen the more the likely further MBS price declines are in store, so dealers will need to sell more Treasury positions to reduce their interest rate exposure.
As you might imagine, Wall Street is very concerned about all of this. Selling of Treasury securities by Chinese and Japanese investors hit record levels earlier this summer. If the MBS community needs to sell Treasurys as well, this will further push interest rates higher, and bond prices in general will continue to fall.
One of my colleagues reported yesterday that if the yield on the Treasury 10-year gets much higher than 2.9%, there will be the need for another round of large selling by MBS dealers. As you can see from the above chart, we're almost at the "tipping point".
But there is one further point to consider that may reduce the pressure on bond rates: the Fed.
Since the credit crisis in 2009, the Fed has been an aggressive purchaser of MBS. This highly unusual activity was undertaken to try to rekindle the housing sector, and there are signs that the Fed's efforts are working. However, this aggressive Fed action has made it the dominant player in the MBS market.
The Fed is not in the business of hedging their massive positions, since their focus is on policy, not total return.
If this true, some of the concerns about "negative convexity" may be overblown.
No comments:
Post a Comment