Submitted by Bo Peng
Careful with 30Y Treasuries
With Fed announcement of the second round of treasury purchase in the
2-10Y segment and corresponding yields at all time lows, it may be
tempting to pick the only cheap thing left in this space, the
longest-term treasuries. It seems natural to speculate that the 30Y will
follow, and the next natural thing for Fed to do should be buying the
Well, every obvious play in the market comes with some
caveat. The more obvious and sure thing it seems, the more worried and
paranoid you should be.
A few risks I can see about this seemingly obvious play.
1. Big players may be tempted to do a curve steepner between 10 and
30Y, long 10Y and short 30Y or some variations of it. Such bets paid off
handsomely in 09, then lost spectacularly since this April but if big
money piles on again, it would tend to drive up the 30Y yield.
2. One rationale behind a curve steepner trade, and a big risk for long
30Y, is inflation expectation. Perpetual stimulus and bailouts will
quite plausibly drive up inflation at some point in the future, and Fed
will be heavily biased to act too late rather than early against
inflation when it happens. And hyperinflation, if and when it happens,
tends to be in the typical black swan fashion -- sudden, violent, and
unpredictable in timing and scope.
3. Fed may be politically constrained in going further out the curve to buy treasuries.
4. Usually, the longer the maturity, the more interest rate risk or the
higher the leverage. This is usually a major rationale to buy 30Y if
you want higher leverage. But, because the curve is already so steep
between 5-10Y and 30Y, the deltas (interest rate sensitivity) are about
the same. In fact, gamma for the 10Y is higher (more negative) at
current levels, meaning that if yields decrease more (say 1% instead of
1bp), the capital gain on 10Y is actually higher than on 30Y. This makes
30Y less sensitive to rates, which is counter-intuitive. But the
inflation expectation is such that the curve may not shift up in
parallel (see point 1 above), and the 30Y bet could be a lose-lose
compared to 10Y -- if inflation shoots up, 30Y loses more because yield
goes up more at the long end; if deflation kicks in, 10Y gains more
because of higher convexity. And this is why the curve steepner is
tempting to begin with.
I try to constantly remind myself of
the LTCM bets and the Merrill bond-CDS basis bet and think thrice about
any obvious trades. And if you can't think of how it could fail, you'd
be safer to simply assume it's a trap. Unless you're trading other's
money and get paid on relative performance basis, you can always better
afford to miss some opportunities to make money than lose money.
1. There's always another opportunity somewhere else, down the road.
2. If something seems too good to be true, it probably is.
3. There's no free lunch.
Well, free lunch galore if you have discount window access or otherwise
win the government random bailout/handout lottos. The New American
Dream makes winning lawsuits look way too tedious and low on ROI and
Sharpe ratio, except the Old American Dream is morally far-superior on
relative basis. Forget about McDonald's hot coffee, people. That's so,
like, pre-crisis and amateur. Just stop paying your mortgage and credit
card and student loan and the money you borrowed from mom in senior
center, or better yet, join TBTF. Sooner or later our central-planning
state-capitalist commiekaze government will help you out.
Disclosure: Long 5-10Y TIPS and treasuries