Yesterday, bonds fell sharply due to stronger-than-expected housing price and consumer confidence reports. That reflects the belief that the economy is mending, and as a result, the Fed will deliver on its promise to dial back and then end QE. Ten year Treasury yields rose to the 2.10%-2.11% level. Various commentators claim that rates will zoom higher either right over that point or at 2.25%. Russ Certo of Brean Capital claim’s there’s a “technical vacuum” at 2.11% that will lead 10 year rates to gap up to 2.25%. And Bruce Krasting about an apparently widespread concern among investors, that convexity in mortgage-backed securities market could produce a nasty back loop, or convexity vortex, at 10-year Treasury yields of around 2.25%.
The guts of his argument, starting with a quote from a hedgie buddy:
Some familiar with it say the vortex is 19 bps away..2.2% on ten year treasury, 3% on the CMM..if breaks, MBS holders subject to extension and duration risk. Would now have to increase convexity hedging. Would lead to price gaps and significant selling. With shortage of treasuries due to bernank and co. and low liquidity, could be very disruptive.
When mortgage interest rates fall, the probability that an individual will re-finance a mortgage increases. When mortgage interest rates increase, the likelihood of a re-financing of the mortgage goes down. Therefore, in a rising rate environment, the average life of a pool of mortgages increases. For example, if a bond fund held Mortgage Backed Securities (MBS) with an assumed 10-year average life, AND interest rates rose, the average life of the MBS portfolio would be extended for a few years. This is convexity. The last thing that a bond manager wants in a rising rate environment is to have the average maturity of the portfolio extended, as this adds to the losses. As a result, MBS players hedge their portfolios against “duration risk” by shorting Treasuries (ten-year paper). The higher rates go (and the speed that rates are increasing) forces more and more of the convexity selling.
Krasting did say this source was a perma bond bear, so he consulted a perma bond bull, who remarked:
I don’t disagree – I would guess we have a huge concentration of mortgages that would go out of the money at 2.25% 10yr UST, slowing prepays, extending servicer portfolios, bringing on longer duration UST selling ……
So we have bulls and bears agreeing. Must be true, right?
Maybe not. First remember we’ve had some dire bond market calls in the past produce some short term perturbations but none of the expected follow-through. The freakout over the S&P downgrade of US Treasuries saw the bonds increase in price shortly after the event took place. Reader AU, commenting on the Krasting post, recalled November-December 2010. The QE2 FOMC meeting coincided with the midterm elections, and bond prices fell in the following weeks. There were similar concerns that convexity would beget more selling, but after the correction, prices settled down.
I consulted our house mortgage maven, MBS Guy, who was skeptical of the convexity-driven meltdown thesis:
Vortex is a great fear mongering word. But a few basis point move just isn’t that severe in any scenario. Yes, some bonds will go down in price and that will cause portfolio shifts. That’s really pretty normal. Investors aren’t guaranteed falling rates forever.
Also, the key to really answering this question is knowing how much MBS was issued at the lowest rates and what percentage this represents of total outstanding. I believe it is probably less that 10% – I don’t have the numbers but every research desk on the Street does.
As a note, I’ve always been a skeptic on the risks of MBS convexity. I think it is one of the great exaggerated risks of the market the fear of which is based largely on the Fed rate raising episode of 1994.
Three factors that will be some counterweight to a slowdown in rate-refi’s:
1. Home purchases are increasing, and if refi’s slow, banks are very likely to loosen underwriting to increase originations. New purchases aren’t likely to be enough to offset the refi slowdown, but enough to change the dynamic Krastinh is describing.
2. As property values increase, more borrowers will be above water again, which will allow them to finally refi at more attractive rates providing a counterbalance to in the money refi’s
3. There’s a $900 billion or so non-agency market that is far more credit sensitive than rate sensitive. Much of this market is adjustable rate, and most pays at above market margins. In theory, these bonds might even increase in value if refi’s slow down (although the prepay rate has been very slow for these borrowers already).
I would note also that this stuff really highlights how unnatural the 30-year fixed rate mortgage is. I don’t see why we should be working so hard to perpetuate its existence when it is the cause of so much potential instability.
Marshall Auerback also pointed out that the danger isn’t the Fed but the state of the economy:
The real problem is the withdrawal of fiscal support. If the market begins to fall, everybody will say that this has been exacerbated by the threatened end of QE, even though I think it is more to do with the sharply declining US budget deficit, which is sucking more and more income out of the US economy.
The impact of QE is more ambiguous: To reiterate what I have said many times in the past, quantitative easing involves the central bank buying assets from the private sector – government bonds and high quality corporate debt. So what the central bank is doing is swapping financial assets with the banks – they sell their financial assets and receive back in return extra reserves. So the central bank is buying one type of financial asset (private holdings of bonds, company paper) and exchanging it for another (reserve balances at the central bank). The net financial assets in the private sector are in fact unchanged although the portfolio composition of those assets is altered (maturity substitution) which changes yields and returns.
In terms of changing portfolio compositions, quantitative easing increases central bank demand for “long maturity” assets held in the private sector which reduces interest rates at the longer end of the yield curve. These are traditionally thought of as the investment rates. This might increase aggregate demand given the cost of investment funds is likely to drop. But on the other hand, the lower rates reduce the interest-income of savers who will reduce consumption (demand) accordingly. So it’s unclear to me that it’s had any kind of beneficial impact and, to the extent that it has engendered risk speculation in the commodities complex, it’s actually derogated from economic growth.
Now some readers may point out that Wall Street was all agog over the data releases on Tuesday, and therefore Marshall’s concerns are overblown. But go look at them more closely. Even though consumer confidence spiked higher, it’s now up to 76, with 100 as the base line from 1985. I have got to tell you 1985 was not a hell-bent-for-leather economy. In fact, Reagan broke with his free market ideology that year to press for the implementation of the Plaza Accord to tank the yen and help American manufacturers. And if you look , more people expected the number of jobs to decrease rather than increase, saw business conditions as poor rather than good, and saw jobs “hard to get” rather than “plentiful”. The only area where sentiment was net positive was on income expectations, where the proportion expecting gains in the next six months was 16.6%, versus the 15.3% that anticipated a decrease.
Similarly, even though the year-to-year gains in the Case Shiller index exceeded 10%, there are two things to keep in mind. First is that buying is concentrated in low-priced homes (and at least in NYC, we are also getting tail-winds from foreign investors, as the Russians in particular are loath to buy in Europe with austerity creating political instability). I heard a report from a gent in Atlanta who had a lovely-sounding home in Atlanta (on a golf course, great condition) that all of one person had looked at in its full year on the market. He said everyone wanted to buy out of foreclosure. Second, even though housing is relatively strong, it’s still not at an absolute level that is robust enough to drive a recovery as it has in the past. And Scott asks how does one reconcile a housing rally with falling lumber prices?
So while Krasting argued that the vortex risk might not be as severe as imagined because the Fed would resume QE if the bond markets became unhinged, you may see the Fed resume QE (assuming it does wind it down sooner rather than later) because it realizes it has misread the real economy. The central bank seems to be suffering of a combination of confirmation bias (needing to believe its patent medicine is actually working) and concerned about political blowback if it does not seem keen to unwind QE (even though Audit the Fed was cut back from its original scope, it still got further than the Fed liked, for instance). Of course, it is also possible that the Fed is playing the same game that Penelope played with her suitors, pretending action is imminent and keeping them drunk in the meantime. And perhaps nervous investors are channeling this myth, since her paramour-wannabes got slaughtered for taking advantage of her hospitality for so long.