Fixing the MBS market, improving QE2 | FT Alphaville

Fixing the MBS market, improving QE2

Are you one of the gazillions of people not concerned by surging MBS fails?

Perhaps you should be.

Fails to deliver of Mortgage-Backed Securities (MBS) reached an all-time record late last month, with dealers failing to receive some $1,000bn worth of the securities.

These fails happen when an MBS seller doesn’t deliver a promised security at the promised time. The reason these have been increasing of late seems to be a combination of extremely low interest rates, and the (glaring) fact that the Federal Reserve has cornered much of the agency market with its quantitative easing.

And fails can be a concern since one fail to deliver tends to have a knock-on effect on other dealers — leading to a daisy chain or round robin of fails. This can mean dealers (suddenly) find themselves having to hold a higher amount of capital.

RBC Capital Markets’ Michael Cloherty has this to say on the subject:

Fails create counterparty credit risk that boosts capital needs (thereby limiting the flow of credit to others). In addition, fails are starting to become a significant problem for the [Collateralised Mortgage Obligation] machine—we see structuring desks paying up significantly to ensure delivery of mortgages that they can repackage for a different investor base. For a Fed that is looking to lower private sector rates, hurting the CMO demand for mortgages is not helpful.

Finally, allowing fails related to QE1 to cross the $1T mark will not help improve the market’s reception of QE2 (as Chairman Bernanke’s Washington Post editorial shows, the Fed clearly cares about the public perception of QE2).

And the Federal Reserve does have some options here.

Back in March 2009, when the US Treasury repo market was in the throes of its own bout of fails, the Fed decided to penalise non-deliverers by introducing a fail to deliver fee. There’s some talk of the Fed doing the same thing for MBS now.

However, the last time the MBS market was tight (back in the summer), the US central bank decided to go for a coupon swap. That is, the Fed offered to swap $9bn worth of 30-year 5.5 per cent Fannie Mae coupons, to ease market liquidity. MBS fails did come down after that, but of course, now they are back on the rise.

Nevertheless, Cloherty figures they’ll try to do it again:

Introducing a fee on a huge number of fails caused by QE1 is not going to help the Fed win the hearts and minds of market participants at a time when the Fed can use all the support for QE2 that it can get. When the Treasury fails fee was finalized, fails had already dropped back to more normal levels.

A coupon swap first. Instead, we think the Fed will take the route that is operationally easiest (and politically least risky) by first doing coupon swaps, selling some of the coupons that have the largest fail problems and buying 3.5s and 4s (where the current production is). We wouldn’t expect that to be done in illiquid year-end markets, but the Fed could go a long way to solving the fails problem by doing a series of swaps early in the New Year. The longer the fails problem lingers, the larger the impact on market liquidity, so we would not expect the Fed to linger too long.

A coupon swap does not permanently solve all of factors behind fails, so it is reasonably likely that the [Treasury Market Practices Group] follows up with a fail fee in the second half of 2011. That doesn’t solve all of the delivery problems either, but the combination should help ease some of the strains in the MBS market.

Fingers crossed.

Related links:
That cornered – failing – MBS market – FT Alphaville
Oops, MBS settlements failing again – FT Alphaville
The introduction of the TMPG fails charge for US Treasury securities – NYFed