1) “It’s hard to get in and out.”

  • There is a Globex market for ED butterflies – just as there is one for outright contracts and calendar spreads.  Many butterfly markets are 1/2bp wide (because the spread markets flow into the fly markets), and with all the high frequency electronic trading, if you are willing to cross the bid-offer, you can get filled in reasonable size faster than you would think.
  • When trading butterflies, do only your “core” view with butterflies – those trades that you will not want to get in and out quickly.  Do the rest in other instruments.  The curvature trade will typically provide better risk/reward, so consider diversifying the trades in your view.  For example, if you have a core bullish view, express 50% in curvature trades and express the other 50% in higher liquidity instruments.
  • Many butterfly structures are non-directional, or can be reweighted to be made non-directional, so there will be less need to trade quickly.
  • I will typically avoid trades that need to be exited quickly on upcoming economic indicators.
  • That’s why you need me!  I can suggest ways of putting on a similar view, and I can suggest alternate ways of neutralizing a position (direction, slope, curvature) rather quickly.


2) “I can use flies to put on a directional view, the directional move can materialize, but the fly can move differently.”  Yes.

  • That’s because while sometimes a butterfly could be highly correlated with direction, the two measure related but different things.  An ED contract price is a function of the Fed rate at a particular point in time, and an ED butterfly price is a function of Fed liftoff / landing probabilities over a time range (in the current rate environment).  Most people prefer to think of the former, but there are times you may want to think about the latter (because it provides a different perspective).
  • There is usually a historic alpha present in butterfly trades I recommend.  So even if the directional view does not materialize, you can make money (or lose less money) on curvature.  This benefit usually outweighs having some tracking “noise” to direction.


3) “I can’t be bothered to look at something that is only 3-5bps.”  4bps on 1,000 butterflies is $100,000.  I used to regularly trade 10+K butterflies.  You get a dozen of those a year and then you are talking about a decent year.  I was not “bothered.”


4) “All you do is look for historical highs and lows.”  No.  There are a number of other things to consider:

  • relevant historical data – does the past period reflect the current environment?
  • makes sense fundamentally – what is going on right now?
  • makes sense based on the current curve shape
  • rolls in your favor
  • has error skew in your favor

The historicals are a good way to screen for good trades, but the better trades take into account more of the above factors.


5) With relative value trades, you can lose a lot of money if there is a “six sigma” event.  You can lose a lot of money on a lot of trades if there is a six sigma event.

  • That having been said, the way you define a “sigma” is usually based on historicals.  As previously mentioned, there is more to identifying a trade than just the historicals.  You need to constantly ask if the historicals are an accurate reflection of the current environment, because a “six sigma” event historically may only be a one sigma event based on the current situation.  Almost all big losses in curvature trading are not from “six sigma,” but poor foresight or planning.
  • You only lose a lot if you choose extremely poor sizing.  There have been plenty of times when I have told clients to put some on, and refrain from adding until I tell them it is time to add.  I won’t get it right every time, but I’m pretty good at making sure you have room.


6) “Having a lot more contracts in my book means there is a lot more risk.”  When you do a butterfly trade, instead of an outright directional trade, you may have as many as 20+x the number of outstanding contracts on the trade.  For example, if a year bullish butterfly centered in the blues had a beta of 0.2, you would need to sell 500 butterflies (and have a 2,000 total contract position) to have the same equivalent bullish exposure as being long 100 of a blue contract.  Even though your positioning is 20x vs 1x, on a Value at Risk basis, you may only have 20-50% more “risk.”  But because you can actually MAKE money on the no-move or selloff scenarios, I would argue that the fly is noticeably less “risk” than the outright contract.  Butterflies are reasonably well-behaved, so don’t be afraid to up-size.


7) “There is a lot of P&L noise and thus is not friendly to Sharpe-ratios.”  There will be minor daily MTM swings.  But those swings are what can cause the mispricings and value on the curve.

  • As a stand-alone book, it performs fine.  I’ve had Sharpe ratios over 2 over an entire year.
  • But as part of your diversified portfolio of trades, it should perform even better, because many of these trades will have little correlation to the rest of your book.  The marginal impact to VaR may be negligible if you do in modest size (relative to the rest of your book), especially since many trades can be non-directional.


8) “I can’t watch it every day.”  You don’t have to!

  • It doesn’t move much.  If it was a directional trade, you need to watch it like a hawk.  But curvature trades are generally range-bound.
  • In fact, I recommend you don’t watch it every day.  Sometimes it can take weeks or even months for something to really kick in.  It’s like watching grass grow – slow and steady, but nothing worth watching.
  • We’ll watch it.  Your broker will keep a record of what you have on.  I send out regular emails updating clients on trades (especially when something noteworthy happens), including how any recent or upcoming data would affect the trade, and any changes in my assessment of good entry, exit and add levels.


9) “I can’t be bothered to do all the maintenance.”

  • It may be more work entering fills if you do not get automated pricing.  But that’s what the junior guy on the desk is for.
  • “I can’t see what my non-fly positions are.”  Get a second book, or add a sheet on Excel that nets out your fly position.  Again, that’s what the junior guy on the desk is for!


10) “That’s a lot of brokerage.”  It depends.

  • We are talking about a fractional amount of basis points.  A roundtrip (in and out) on one butterfly is less than a quarter basis point.
  • Unlike other newsletters, there is a reasonable probability that we will restrict Curve Advisor readership once we reach sufficient volume – I don’t want overcrowding in trades.  We have already started tiering the distribution so that active clients get the information much sooner.  Don’t get left out.