Diary of a Financier

T-Bond Yields & My Convexity Miscalculation

In Capital Markets on Fri 4 Feb 2011 at 08:29
  • I have small losses in a 4-part pair trade to capture a bulging intermediate Treasury yield curve.
  • My theory was spot-on, but I miscalculated the convexity of long-dated bonds.

In the month of QE2 2010, I put on a 4-part pair trade to try and profit from a future bulge in the middle of the yield curve.  I had viewed intermediate Treasuries as the forgotten children of bond issuance: Quantitative Easing was multiplying demand for short maturities, and a low-rate/low-return environment made the extra yield on long-dated bonds attractive.  A donut-hole of demand had been created, into which intermediate Treasuries fell.  Therefore, yields in the middle of the curve were to climb higher at a faster pace than those at the bookends, which were buoyed by demand.

Thus, I went long 1-3 year Treasuries (SHY) & 20+ year Treasuries (TLT).  I shorted 3-7 year (IEI) & 7-10 year (IEF) to capture their relative underperformance.  The trade has underperformed so far:

What appears to only be 73 bps in cumulative loss is actually more like 97 when you add up the total returns (including interest payments).

I write you today to first update the status of this trade.  Done.  I also write to remind myself that no matter how fundamentally sound a thesis like this may be, it doesn’t matter if the Market doesn’t make it happen.  As you’ll see below, I was right about this trade, a bulge did develop in the middle of the curve.  But, I’m still at a loss in the open pair, which means the Market didn’t cooperate (and I perhaps enacted the idea through the wrong vehicles, ETFs).

This morning, Doug Short posted this graphical representation of what happened in the yield curve:

Treasury Bond Yields- long & short dated yields climb far less then intermediates.

Again, that’s exactly what I had expected to happen, but the duration risk of my longer dated maturity exposure (TLT) made the price fluctuation of 20+ year Treasuries far too sensitive, overrunning my profits from shorting the middle of the curve.

I’m reminded of my prior advice to John Hussman, as Josh Brown explained, “Cardinal Sin Number 1 in the investment game: Conflating his views of what should happen with what could happen.”  I’m maintaining the trade for now, with the expectation that demand will redouble on the long end.  But, I’ll have a quick trigger in unwinding this, because QE2 ends in June, and I have no desire to maintain long positions at the front end of the curve throughout that period.

I hate being wrong, and I hate losing trades.  However, I’ve committed little capital to this, and it’s really been a Market Neutral-ish Beta buoy from the outset.  I know I’m 80% accurate in my long-only trades, 65% accurate in my shorts, and 75% accurate in my relative strength pairs.  That means that I rear absolute return, as long as when I’m wrong, the damage is minimal.  This here is trading with momentum, it’s long/short, so as long as I’ve closed it before QE2 unwinds, I expect risk/reward between -2% downside and +4% upside total return.


  1. […] closed that convoluted 4-pronged pair trade in the Treasury yield curve; total return ~+0.76%, including dividends.  These pair trades are all […]


Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s