The big story on the curve last week was the collapse of the ED-FF spread.  I had been saying for a while now (as recently as the last CA) that EDM7-FF seemed strangely high to the rest of the ED-FF curve.  That finally capitulated last week, with EDM7 rallying 7.5bps, while FFN7 only rallied 1.5bps.  I updated the chart from the Trade Thoughts section last week to show the dramatic changes in the ED-FF spreads in the past week.  You can see that EDM7-FF declined 6.3bps on the week.  Now EDM7-FF looks a little too low on the curve, but it’s not unreasonable to expect libor-FF to drift a little lower in the near future.  There are a few things we should take away from this:

  • The markets can sit on an “incorrect” position for a long time. It takes a lot to keep something at a non-equilibrium level.  So when you see the mispricing finally give, you can expect that we could go through “fair value.”  We may have seen this Friday…  or we may see more.  “Fair value” in this case is hard to quantify, since it would be reasonable to expect libor to continue to drift lower (as per the strong recent trend).  But I don’t see any unusual activity in futures open interest – the change in OI the past week is only +1.6K contracts.  So I’m a little unsure of where the move is coming from.
  • Some of the move appeared to have been options-related, as there anecdotally appeared to be a lot of liquidations (and hedging) of put structures. Not including Friday (the OI data is 1 day delayed), there was a decrease of 159K puts and an increase of 108K calls in the previous four days.  However, the magnitude for OI on options can be overstated because most people put on multi-legged structures (put flies, put 1x2s, etc), and because the delta on some options can be small.  When we get Friday’s OI data (on Monday) that could be more telling.
  • Just as a helpful suggestion, if you are going to do some options trade that relies on ED-FF being a certain level, keep an eye on libor-FF to make sure it doesn’t get too out-of-line with your assumption. ED-FF will eventually converge to libor-FF.  We didn’t have any options trades on, but EDM7-FF being too wide was probably a good signal to unwind or hedge EDM7 put structures.
  • The ED-FF compression could have also played a factor in the ED rally last week. We can see from the above table that even further out, we saw EDs rally 5bps relative to FFs the past week.  For example, EDH8 rallied 8.5bps on the week, but if 5bps of that was just the spread to FFs, then the Fed probabilities only moved 3.5 bps on the week.  That is not excessive considering the weaker Retail Sales, CPI, and Trumpidity.  But from Thursday to Friday, ED-FF narrowed less than 1bp, so this isn’t the full explanation of “something else.”
  • As a second-order effect, the collapse around EDM7-FF led to a steepening of the ED-FF curve in an environment where everyone is clearly grabbing for yield. I put out a new trade (G16) that tried to take advantage of this, and this was up a very painless 1-1.5bps in less than a day.
  • I have suspected that the relative richness of EDN7 (and to a lesser extent EDU7) has been related to the cheapness of EDM7. Some guy has been just relentlessly been on the EDM7-N7 offer.  I’ll discuss this further in the Flip Trades section.  Possibly related is the fact that there have been large sales of EDM7-U7 spread (and buys of EDU7-Z7 spread) in the past few weeks.  Let’s see if this is the next shoe to drop.
  • Monday’s libor fixing could be interesting. EDK7 has a 69% exposure to the June meeting, which dropped 2.5bps Friday.  Yet EDK7 is only pricing in a 0.1bp rally from Friday’s fixing to Monday’s fixing.  The libor fixings tend to be sticky, and I have said in the past that we have no competitive advantage in trying to guess how the fixings will come in.  That having been said, this is a head-scratcher, as I would think the people who actually fix the libor rates should know where EDK7 should be trading.  But if EDK7 is correct, this means that libor-FF is going to be about 1bp wider than it was the day before.  This also means the front EDs may have rallied 1-2bps too much relative to FFs (as compared to libor-FF).  This would corroborate the “overshoot” through fair value theory above.  If EDK7 is incorrect, that means the libor Chinese Wall does not belong in the same category as Santa Claus, the Easter Bunny and unicorns.  I mean what are the chances that of the “16” banks, the Wall is intact on all 16 banks?  It only takes one leak in the Wall.  The cynic in me would be stunned.  Let’s see what happens.

I discuss some trading takeaways of the above in the Trade sections below.