Convexity vs Overseas Money Flows vs Real Yields vs The Fed

The move in the bond market today was pretty violent. I briefly wondered how notable the move was relative to history, and I found that we have seen much worse moves in the past. Examples:

  1. Early 2004, when the market realized that the FOMC was going to have to tighten.
  2. Summer 2003, when we got a self-sustaining selloff in the market when we realized that fed funds would not be 1% forever, and mortgage hedgers forced an overshoot.
  3. Spring 2003, when the FOMC hinted that it would cut to 1%, and actually did it.
  4. 2002, when the FOMC was aggressively cutting the fed funds rate.
  5. And more… I could go back a lot further, but moves of this level of violence occur about once a year on average. Since the start of the FOMC tightening cycle, things have been calm, and predictable, lulling investors into a false sense of security. That security has just gotten shaken.

Over at RealMoney, except for Tony Crescenzi, there aren’t many who know much about
bonds. A lot of the writers there are paying attention to bonds, but it is basically a black box, or a technical model to them. In this piece, I will try to explain the four factors playing tug of war with the long end of the bond market, and what the prevailing effect is likely to be over the short and intermediate terms.


The largest component of the bond market is residential mortgage backed securities [RMBS]. Residential mortgages shorten when rates fall, because people refinance; when rates rise, fewer people than previously expected refinance, and people take longer to pay off their mortgages. A bunch of similar residential mortgages get gathered into a trust; participations in that RMBS trust are sold to institutional investors.RMBS is a complicated security to manage from an interest rate standpoint. Originators of the mortgages have to hedge them before they sell them off to investment banks who will turn them into RMBS. Many bond managers like to own RMBS for its high credit quality, liquidity, and attractive yields, but the problem is this: when interest rates move, the RMBS does what you don’t want to see happen. When rates fall, the bonds prepay, and you have to reinvest at lower rates. When rates rise the bonds pay slowly; you’d like to have more cash to reinvest at the higher rates, but the RMBS pays slowly.

This effect of shifting cash flows working against the owner of the mortgages or RMBS is called negative convexity. Basically, the owner has written a covered complicated bond call option. Anyone option writer receives a premium (extra yield), but must sell his investment if it goes into the money. If the investment stays out of the money he keeps the premium and the depreciated investment.
The mortgage originators and the RMBS holders want to limit the effect of the options that they have written, so they hedge. They can do that in three ways. I’ll use the example of hedging rising rates:

  1. Sell or short long Treasuries to lower overall exposure to rising rates.
  2. Sell RMBS that is lengthening, and replace it with RMBS that is less sensitive to rising rates.
  3. Enter into a swap where you pay a fixed rate for a number of years, and receive (floating) three month LIBOR over the same period.

Now, when rates rise or fall by a lot, mortgage hedging tends to reinforce the move, making falls and rises sharper. The effect depends on the configuration of the rates that people are paying on mortgages across the US. Negative convexity is highest on mortgages the are near the current mortgage rate for new loans. (At the money) At present, most prime mortgages are now below the current mortgage rate. I estimate that 20% are above and 80% below. What that implies in the current situation, is though mortgage hedging has had a big impact in the move so far, its effect should diminish dramatically from here, and almost disappear if the ten year gets up into the 5.50-6.00% range. Compare this to an era like 1994, where at the start of the move 10% of mortgages were below the current mortgage rate, and 90% above; it’s no wonder that move was so devastating. From the convexity effect, it should not be that bad here.

The Fed

We finally reached the fifth stage of grieving on the expected FOMC rate cut. The market has accepted that no cut is coming in 2007. This makes me edgy because that had been my view since December, and when the crowd crashes my lonely party, I have to ask, “Okay, now what? When and what is the next FOMC move? What is the crowd missing?”
Honestly, I don’t know, at least not yet. The FOMC is happily balanced between a variety of concerns, and I think it leaves them in a state of noisy inaction. Right now I think they will stand pat for as long as they can justify it. The effect on the bond market been to de-invert the curve, with the long end rising significantly. From here, though, unless the probability significantly increases that the FOMC will tighten, the effect on long rates is done here and now.
Overseas Money Flows

The current account deficit must be matched by a capital account surplus. The books have to balance! The only question is at what price (exchange rate) that the market clearing balancing will occur at. Partially due to the new view of FOMC policy, the dollar has strengthened over the last month, as foreign investors think that US rates will remain high for a while. But the varying currency policies of different countries will have an impact on the overall path of the dollar, and interest rates. Here are some bits of news affecting the question:

  1. South Korea continues to let the won rise, depressing the economy, but limiting local inflation. Also, they don’t have to buy as many US dollar assets.
  2. India has acted similarly with the rupee, which is having an effect on its competitiveness, but is reducing local inflation.
  3. Thailand had to let its currency appreciate against the US dollar because of rampant speculation.
  4. Other countries like Russia and some of the Gulf states are reducing exposure to the US dollar.
  5. Not everyone is reducing exposure to the US dollar, though. Most of the Gulf states still recycle petrodollars at the same rate as before, and China still uses the dollar as its linchpin in its currency system (which is just a dirty crawling peg). Now, this is leading to inflation in China, and a potential crisis, but so far, nothing is imminent, and that should continue to have a depressing effect on US interest rates, and retard the fall of the dollar.
  6. Growth is proceeding better outside of the US, which should put downward pressure on the dollar and upward pressure on rates.
  7. The two main carry trade currencies, the Japanese Yen and the Swiss Franc, are still at low levels, with low but increasing interest rates. There is atill a lot of speculative borrowing going on in both currencies, both locally and abroad, which is depressing the exchange rates. Eventually this will end with the Yen and the Swaissie rallying, but that could take a while.

So long as enough foreign entities are comfortable providing goods and services to the US, which taking back financial promises, there will continue to be downward pressure on rates, and the dollar will muddle. Woe betide us when they tire of giving us such nice treatment.
Yields and Real Yields

Having been a bond manager at a major insurance company I have experienced being a yield buyer. A yield buyer is willing to buy more bonds at higher yields, all other things equal. A real yield buyer is willing to but more bonds as the inflation-adjusted yield rises. Well, we are at levels on the ten year Treasury yield that we have not seen in five years. Yield buyers should become increasingly interested, which should moderate any increase in yields. Similarly, certain types of bond issuance should slow, as some projects will no longer justify paying the higher yield to finance the deal.
Real yields are another matter. As real yields rise, defined benefit pension plans and endowments buy more. But how much is inflation rising? It’s hard to say; our trade with developing nations decreases our inflation rate; some suggest that that might be coming to an end as inflation rises in the developing countries. My suspicion is that inflation is higher than the government says it is, but is not increasing much at present.

My upshot here is that up 0.25-1.0% in yield from here, assuming modest increases in inflation, we would see a lot of incremental buyers for bonds. We would also see fewer issuers.

Of my four effects, convexity is a spent force soon. The FOMC is likely on hold, so that effect is largely over, unless I am wrong. If rates rise, I expect yield and real yield buyers to arrive. The wild card is the foreign investors; if more countries revalue their currencies upward versus the dollar, reducing exports to the US, and reducing the need to buy our debt, then interest rates could easily rise by another percent. I don’t think that will happen, but it is the major risk here.

Absent negative foreign developments, I will get more bullish on the long end around 5.50% on the ten-year Treasury. Until them, I with with my positions, and clip the coupons; I don’t expect the rise to be severe, but will invest more on weakness.

Full disclosure: long TLT, FXF, FXY, and the rest of my bond portfolio is over at

PS — It is not obvious to me that the current correlation between the stock and bond markets will maintain. I expect stocks to do better, relative to bonds, in the short term.