A few thoughts on bonds

Between the bank holiday in the UK and the Sohn conference in New York, it was unsurprisingly a very quiet start to the week yesterday.  Macro Man availed himself of the soporific trading conditions and the brilliant sunshine to climb some of the local hills on his bicycle, and returned after a couple of hours to find the SPX virtually unchanged to the decimal place from the time of his departure.   Says it all, really.

One thing that grabbed his attention was an article from Bloomberg entitled "Can't Find Enough Treasuries to Buy?  Here's Why."  His first reaction to the story was "this would have been a lot more useful in January", though that's perhaps a trifle unfair on the journalist, whose job is to explain rather than forecast.   His second reaction, however, was "well, if that's not the beginning of the end of the bond rally, it may well be the end of the beginning."  Although this was hardly the cover of The Economist, it is still worth noting when market developments become sufficiently entrenched to attract journalistic attention.

That having been said, it still looks to Macro Man like 30's are going to 3%.  Other than a bit of yield support around 3.30%, there isn't much technically to prevent long bonds from taking a peek at a 2 handle.   The latest CFTC data has duration-weighted bond futures positioning still short, and frankly if bonds couldn't meaningfully sell off last Friday, then it probably isn't going to be data that turns this sucker around.

Would Macro Man go long bonds here?   Not necessarily.   Although one could argue that a 30 year long could use a yield stop around 3.55% (the cluster of lows in Feb-March as well as the mid-April yield high), thus offering a decent risk/reward, he generally eschews trading the long end save at clear price or economic extremes.  His rationale is that the policy rate anchor is pretty tenuous, and markets can choose to interpret the same set of factors in radically different ways depending on flow, positioning, and even regulatory factors.  In other words, it can more or less go wherever the hell it wants, and he doesn't feel like he has an edge.

Curve trades are normally a bit better, because they strip at least some modicum of the (oft-inexplicable) market beta out of the equation and provide some degree of protection in getting the fundamentals wrong.   Of course, curve trades are generally directional in and of themselves, so it's not like all risk is defrayed.

Although markets have been short US bonds in aggregate, hidden in that data is a reasonably sized 5-30 flattener.   One would have to think that these trades were established with a bearish tilt in mind, expecting Fed normalization to lead the belly higher in yield.   Curiously, the trades appear to have worked splendidly...as a bullish trade, given long-end outperformance.   Still, happy days, right?

Well, maybe.   Remember, these trades have some of the characteristics of bearish trades, including, unfortunately, the negative carry/rolldown.   Macro Man recalls attending a traders' meeting once when the senior figure asked why no one seemed to be able to make money running short rates.  This environment is a perfect example, in more ways than one.   Not only do bond bear markets typically "enjoy" vicious short-covering spikes (2014, anyone?), but you really have to trade them on a hit and run basis due to the carry.

Consider the 5-30 flattener mentioned above.    The curve peaked at 2.53% in mid-November and has subsequently flattened to 1.73% in close to a straight line.  Imagine you did a 5-30 flattener in swap space, 3 months forward, at the ding-dong highs in the curve.   After that 3 months, you must have had it off, right?  Ummm, no.   In fact, you broke even, as the forward was some 30 bps lower than the spot flattener, and there was....30 bps of flattening in that 3 months.   Ouch!  Indeed, the maximum profit that you would have made from this trade during the 30 year rally was 18 bps, and that's if you did it exactly 3 months ago.  For most of the year, the trade would have been offsides or flat.  (The chart below depicts the spot 5-30 curve with the 3m forward from 3 months previous.)

And that, in a nutshell, is reason #3456343 why 2014 has been such a scuffle.   This flattener has been one of the most pronounced trends in the macro space in 2014, but even with virtually no pullbacks it's been difficult to monetize.  The good news is that the carry isn't quite as prohibitive now....the bad news is that the spot curve is as flatter than at any point since 2009.  Good luck if there's ever a pullback!

So it's damned if you do, damned if you don't.  Suddenly, the prospect of a another round of hill climbs doesn't sound so painful....