In a headline dependent market, things can change in a hurry. We saw some of this last week, as we had surprising progress in trade talks with the EU. We could get a number of game-changing headlines in the next few weeks:
- Substantial trade progress (or lack thereof) with any of our key trading partners,
- Any surprising moves by central banks (in particular the BOJ),
- Any surprising economic data,
- Any surprisingly large move in equities, and
- Any serious frictions with Iran/Russia/China.
In addition, we always have the usual terrorism, Trumpidity, etc risks. Usually, I’ll look at the curve and see what I think is rich/cheap. This week, I thought I would look at the curve, and just look at the tail probabilities on the curve. It’s more manageable to look at the curve in 1 year pieces:
- Prelude: the ED-FF spread. Let’s assume for now that the ED-FF spread is a constant 32bps. It’s going to chop around, but I would think the tail on this is towards widening from the current levels. While the markets are very calm about the ED-FF spread, it’s still somewhat possible that we get some kind of EM or credit crisis. I can’t see this spread getting much below the mid 20s (if at all). We are going to start seeing some massive deficits, and gradual increases and the QE taper should continue to put widening pressure on the ED-FF spread once we find a seasonal bottom.
- EDZ8 (aka 2018). EDZ8 settled at 97.325. Using an ED-FF adjustment of 33bps (32bps plus 1bp for the year-end turn), we get a FF rate of 2.345, which happens to equal the average of FFF9 and FFG9 (an approximation of EDZ8). The current FFER is 1.91, so we are pricing in 43.5bps of hikes in EDZ8 (or 42.5bps through 2018). I mentioned last week that I thought 40 bps was a little high, and we are a few bps higher. When considering downside tail risks with EDZ8, there are two: (1) an aforementioned Libor blow-up, and (2) the possibility of the Fed hiking MORE than 2x by the end of the year. I suppose these are possible. Could tariffs lead to an inflation surge that the Fed could overreact to? I suppose. But Junker’s negotiations collapse is going to make a prolonged trade dispute difficult. So you would need the Fed to overreact to a presumably temporary high inflation print, which is fairly low probability. For upside tail risks, it’s possible that the Fed may be done hiking for the year if we get low inflation, growth or a noticeable tightening of conditions (equity meltdown).
- EDZ8-EDZ9 spread (aka 2019). Since we are assuming a level ED-FF curve (and it is fairly level to within a few bps), the markets are pricing in approximately 39bps for 2019. It’s not a crazy stretch to say that it’s possible for the Fed to hike three or more times in 2019. Nor is it a stretch to say it’s possible for the Fed to not hike at all next year, or even ease. So I’ll throw Kong a banana and say the risks to owning EDZ8-Z9 are probably neutral in the current environment. There would be clear upside if we get a trade deal done and/or we get other central banks (BOJ, ECB) to move from their current positioning. And clear downside if the data were to turn south.
- EDZ9-EDZ0 (aka 2020). EDZ9-Z0 spread is 0.5bps. It’s what the odds-makers would call a “pick ‘em”. You are either with the Fed in thinking they could still be hiking in 2020, or with some analysts thinking that we may be in a recession by 2020. My general feeling is that if we don’t steepen this part of the curve by the end of the year (either via trade deals or other central banks normalizing), we could see more recession being priced at the start of next year. Brexit will be just around the corner.
Some of you may be familiar with the series of articles on curvature trading I wrote for the CME. The third article of the series was on a probability interpretation of ED options flies: https://www.cmegroup.com/education/files/curvature-trading-part-three.pdf. I can only cover so much in one issue, so in this week’s Trade Thoughts section, I discuss what the options markets are telling us about #1 above (EDZ8), and where I think there is some value.