A boring double butterfly story
September 3, 2020
–In spite of a 1.5% surge in SPX and 1% in Nasdaq to new highs, longer end yields fell with tens down 2 to 64.8 and 30s down 4.3 to 1.374%. Curve flattened. After Powell’s Fed framework speech last Thursday, 5/30 had posted a new high for the year, ending at 123.4 on Friday. Yesterday it was down to 112 as the market appeared to conclude that the Fed’s mindless buying of assets will never allow inflation to take hold. The dollar rebounded, bitcoin was pummeled.
–Jobless Claims and ISM Services today in front of tomorrow’s payroll data.
–A colleague mentioned a trade to me yesterday (thanks LL, I guess), 5k EDH’23/EDU’23/EDH’24/EDU’24 double butterfly at -1.0. You may want to stop reading right here. A sale of a butterfly is -1/+2/-1. A double is one fly against another, so -1/+2/-1 vs +1/-2/+1, or aggregated, -1/+3/-3/+1. These trades were popular with the locals for exploiting the turn, or for forecasting possible Fed activity. If legs are equidistant, as they are in this case ( a 6-month double) then it’s the middle spread, which is 3x as large, which determines the double. In this case, just imagine that the Fed is on hold, but is guaranteed to raise 25 at every meeting beginning in Dec 2023. Then EDH’23 to EDU’23 might be nearly flat, but EDU’23 vs EDH’24 might go as high as 75 and EDH’24 to EDU’24 also to 75. In this case, the double would be Negative 150 (0 -75*3 + 75). In reality, a back month double is quite stable as the attached chart of ED11-3*ED13+3*ED15-ED17 shows, although there can obviously be roll dynamics over time.
–In any case, there are a couple of odd things about this trade. First, on Tuesday it had settled -2.0. Yesterday, it traded -1 and was small -1 bid into the end of the day. However, it settled -2.5. Paper must have thus been a buyer of the double. Open interest changes, -9k, +4k, +7k, -400. Doesn’t reveal much, although from my chart it appears to be smack-dab in mid-range, so there’s no “edge”. (Now tell him the GOOD part Mortimer). I would also note EDH’22 was down 9k in OI, so since EDH’23 was down 9k, it looks like someone exited EDH2/EDH3 one-year. From my vantage point, it’s worth selling the dbl at -1, figuring on a settle at -2 and then legging out over time. Of course, why go to all that effort when you can just buy Nasdaq?
Settles:
EDH1 9979.0
EDU1 9980.0
EDH2 9979.5
EDU2 9977.5
EDH3 9973.5
EDU3 9967.5
EDH4 9959.0
EDU4 9950.5
EDH5 9940.0
so
H1 6-m dbl 0.00
H2 6-m dbl 0.00
H3 6-m dbl -2.50
