Extreme, even absurd-sounding scenarios. Don’t feel foolish for including them in your reasoning. Sometimes one has to ponder the extremes to reach the utmost clarity.
The last week’s Swiss Franc move reminded to us that no one is safe in the markets. I could gloat and say “Look at me: I was wrong way around, but I actually did well because of my good risk management and portfolio construction!”
But, rather I am saying “Whew…”
In the world of leveraged finance one may not honestly say “My portfolio cannot blow-up.” Rather I would hear “My portfolio will not blow-up, unless U.S Government declares a full default… China goes to war with Japan… aliens invade…”
In the spirit of considering all scenarios, let’s talk about the bond markets.
One of the arguments I heard recently from people who wanted to be short U.S. government bonds is “With 10-yr note yield sub 2%, how much more can they rally? The risk is skewed to the rate upside.”
I take an issue with this approach. Let’s consider the true extremes.
1. All rates going to infinity. An event approximated by full default or hyperinflation. All future cash flows become irrelevant. All bond prices, regardless of maturity, go to 0.
The downside for a par bond is 100 to 0. Regardless of coupon.
2. All rates go to 0. An event approximated by establishing gold standard and no term premium. All future cash flows have the same value as present cash. Every bond is worth it’s face value plus all future coupon payments.
The upside of par 2-yr note with 0.5% coupon 100 to 101. Not much compared to the downside of 0.
However, the upside of 30yr par bond with 2.5% coupon is 100 to 100 + 2.5*30 = 175. Not so skewed, is it?
Another way to think of it: if you have a zero-coupon bond trading at 50, the risk symmetric with respect to these two scenarios (0 with infinity rates, 100 with 0 rates).
This counterintuitive distribution of risk, has to do with “bond convexity”. The convexity is caused by the fact that, as rates fall, the future gains and losses become more meaningful when discounted to the present. Thus those who bet on falling rates see the size of their position increase, as the market moves in their favor. The converse is true with betting on rising rates. These convexity effects increase dramatically as the maturity of the bond lengthens.
3. Rates go negative. Not possible, huh? Tell this to Euroland, Switzerland and Japan. So far the negative rates dominate only the portion of the yield curves within the 10yr mark and the convexity effects are subdued. But who can now deny the theoretical possibility of negative rates all across the curve?
Can future sovereign obligations become multi-fold more valuable, just by the virtue of being the future?
The U.S. Treasury curve seems in no such immediate danger. At least in nor much more danger than EURCHF floor was from being released.
However, the United States has a fundamental structural difference from European markets – the fixed prepayable mortgage market.
Imagine the prepayment and refinancing wave that would happen if the 30yr mortgages rates started to head towards 0. Below 0?
The mortgages originators will find themselves immensely short fixed income market and forced to keep buying, exarcurbating the move.
Now the dealers are not stupid (mostly), they are hedging (somewhat) not only their directional risk, but also the convexity. But are they really prepared for negative rates? The thing with derivatives: for every party there has to be a counterparty. So, no matter how you push the risk around, it remains in the system.
And when people are caught unprepared, things get funky. U.S. economy might not warrant negative rates, but can the runaway convexity move take us up there? Could we theoretically see bond futures rallying say, to 1000?
Those are not likely scenarios. And I am not saying that you cannot trade bonds on the short side. We are in the business of taking risk.
But don’t say that bond upside is limited. And if you blow-up by being short, do it with your eyes open.