[From time to time, I will post trade-related commentary from the web site.  The EDM7-FF spread collapsed the following week and this trade thought has resulted in one profitable trade and the second order effect of the ED-FF curve steepening has given us a an initial profit on a second trade.]

One of the more interesting developments this year has been how the ED-FF spread has collapsed.  Take for example EDM7 vs FF.  It was over 35bps at one point this year, and it settled over 15bps lower on Friday.  I suppose considering we’ve had a yield-grab rally, it makes sense that the yield grab extended to receiving in swap spreads.

More recently, I think there has been an interesting development with the term structure of ED-FF.  EDM7 seems to be strangely offered to the Fed Funds curve.  The three most reasonable explanations for this effect are:

  • Some aggressive options-related play because EDM7 settles after the FOMC meeting. However, I don’t see anything that looks particularly strange in options open interest.
  • Some sort of funding issue which happens to coincide with mid-June or late-June.  However, EDK7 also goes over mid-June and quarter-end, and EDK7-FF has not been affected.
  • It’s related to the high probability hike in June. I think this could be the cause.  While libor-FF spiked in December after that hike, libor-FF fell in March after that hike.  Also, the spike in December occurred mostly after the day of the FOMC meeting.  I think if libor-FF continues to sit around here, we could take profit before the meeting.  So I think there is value here.

This has resulted in a few observations I previously made via email:

  • EDM7 may reach “full 25bp pricing” before FFs. If you assume that libor-FF spread remains unchanged, EDM7 may settle as high as665 on a hike (= 100 – 91bps FFER – 25bp hike – 17.5bps libor-FF spread).  And it could be higher if you think libor-FF continues to decrease.  So if you need a relative or absolute long, this could be attractive.
  • EDK7-M7 spread settled at 10bps. That spread contains 32% of the June meeting and 38% of the July meeting.  If the June meeting was fully priced, this spread would only be 8bps.  So the rest of the spread is the term structure of ED-FF over the 1 month period.  Of course, the June meeting is not fully priced, so the difference in term structure is worth closer to 3.4bps.  That’s a lot of roll for 1 month.  The main problem with fading this is that EDK7 settles in 1 week.  Afterwards, we have to hedge the EDM7 with something else (presumably FFs) to express your view.  But I am just mentioning this, in case it fits your view.
  • EDM7-N7 spread settled at 3.5bps. However, this was a high settle – it is better offered, and there have been large trades @ 3 the past few days.  The EDM7-N7 spread contains 30% of the July and 30% of the Sept meeting.  This spread has the benefit of rolling favorably down the ED-FF curve.  Once EDK7 rolls off, you would expect EDM7-FF to roll down to cash libor – FF spread faster than EDN7-FF.

The EDK7-M7 is awkward because EDK rolls off in a week.  However, I do think there could be some value in a EDM7-N7 position.  This is a new Flip Trade.  This is also related to Trade G14 (Buy 2x EDM7 vs Sell EDU7).  I just think that EDM7 and EDN7 are so close together that they should mostly move together (and to a lesser extent EDM7 and EDU7).  EDM7 is more likely to roll quicker to the cash libor settings, and the positive term structure of ED-FF should put downward pressure on EDN7 (since the ED-FF spreads further out are wider).  So these trades make sense to me.

[1] Assumes a 0.5bp year-end turn.