Curve AdvisorKeymasterAugust 24, 2015 at 2:16 pmPost count: 612
One of the key things about trading anything based on history, whether it involves datamining, technical analysis, historicals, etc. is that you have to be on the lookout for a “regime change.” That is, a change in the way the markets behave towards various parts of the curve. For example, we got a minor regime change when the Fed removed “considerable time” from their statement December 2014. We got another minor regime change when the Fed removed “patient” from their statement in March 2015. Both of these minor changes have helped certain trends on the curve – mostly the trends I described in the “What happens to the curve during a hiking cycle?” thread (whites and reds staying offered, and the curve flattening further out).
However, you need to be alert to things that could change the current regime. On the hawkish side, a major regime change could be an actual liftoff. On the dovish side, a major regime change could be the presence of conditions that could cause another recession. This is particularly important now, as I think the odds of a recession in the US by the end of next year is about 20%. This is not a small number, considering how strong people think the US economy is. Your own assessment of the risks may be only 5%. The number is not particularly relevant. What is relevant is that you think about what changes are going on in the markets, and how key events could cause structures you’ve been watching to no longer trade in a “predictable” manner, within the confines of the current regime.
Curve AdvisorKeymasterSeptember 30, 2015 at 5:12 pmPost count: 612
For people new to curvature trading, it may be helpful to see what the curve looked like at various points in time. One of the ways to get a “big picture” view of curvature is to look at a static 1 year fly curve. That is, calculate the 1 year flies, centered at each contract and graph the results. Below is a snapshot of the 1 year flies on Sept 25, 2015. The red arrow points to the 1 year fly with ED16 at the center (so the ED12-16-20 fly on 9/25). Recently in the CA, I mentioned that this was not an equilibrium curve – that it looks more like the curves from two different regimes mixed together. In a Fed hiking environment, you would expect to see the year flies be much higher on the left and lower on the right. In a recessionary environment, you would expect to see negative flies on the left (rather than the positive left hump), and the right probably would be a little higher.
Recognizing when the curve is signaling a regime change is critical to curvature trading. This is because while some structures may remain rangebound within a regime, when the regime changes the new range could be much different. This is one of the reasons why I keep stressing not to overly rely on historicals as a basis for trading. Basic historicals have no idea when a regime has changed.
But how would you know what the curve is supposed to look like in different regimes?
Attachments:You must be logged in to view attached files.
Curve AdvisorKeymasterOctober 1, 2015 at 9:18 amPost count: 612
If you have the historical database as described in the “Data Set Methodology” thread, and you know how to write a macro, you can easily make yourself a little “movie” on how the curve moves over different regimes. So you can look at, say the 2004-2006 hiking cycle to see how the curve moved after liftoff. Or you can look at how the curve moved prior to liftoff. Or you can even look at how the curve moves prior to a recession.
All you need to do is
* Create a chart (or charts) that shows the 1 year flies, some indicator of the level of rates (and/or slope) and the date. You will want to fix the axes.
* Have all the data for the charts on a sheet, with the chart(s). This is not necessary, but you may want to do further analysis on a particular date at some point, and it’s nice to be able to access everything on one sheet.
* Write a macro that “transfers” the data from your database to the sheet with the chart. You can do this using lookup or index functions. You can also do this using copy/paste, but this will not be as “smooth.”
* You will probably want to add some features, like start/end dates, date increment (daily, weeekly, etc), and speed adjustment.
Day to day, the curve barely moves, but over larger periods of time, you will see dramatic changes.
Curve AdvisorKeymasterOctober 5, 2015 at 9:02 pmPost count: 612
One of the more interesting things going on right now is how the 1 year flies have gotten smoked the past few months. Five months ago, the ED4-8-12 fly was as high as 35. As of Oct 5, 2015, it is 8. This is highly unusual when the Fed says they will most likely hike within the next three months. Typically, you see the markets pricing in a number of hikes when the long term rate is so far away. Currently, the largest year spread is only 55bps. So the markets only see TWO hikes a year. Here are two common scenarios:
1) The will hike approximately once a quarter, as per their Projections. The long term rate may be in question.
2) We are doomed. The Fed will never hike.
So if you weight the two scenarios above, roughly 50-50, you more or less get the current curve. Year spreads are closer to 50bps, rather than 100bps, and the long term rate is closer to 1.75% than 3.5%. Earlier, in the year, we were pricing a little closer to the Fed projections (although with a terminal rate still lower than the Projections). Obviously, this is a gross simplification of what is going on, but the point I wanted to make was that we are currently in the midst of switching regimes. We are going from the story of the Fed being on a gradual hiking path to the Fed will have to be on hold for a while (possibly after a few hikes). When you look at historicals for trading, you always have to be cognizant when the way something trades could change. You can’t expect the trading ranges in one regime to hold up in another.
bambamParticipantOctober 11, 2015 at 6:35 pmPost count: 24
The funny thing is how global factors have pushed down the expectations on FED moves, while the super independent FED has only recently acknowledge events outside the US.
IF and only IF we were to see an aggressive reversal in oil and inflation was to stop hiding in rent/housing and the fed goes aggressive but quick, then this fly is an amazing buy, correct?
What are you representing on the X axis on that chart?
Curve AdvisorKeymasterOctober 12, 2015 at 5:54 pmPost count: 612
The x axis is the center contract of the year fly.
This fly is a good buy for the “slow and steady” growth scenario – the Fed stays accommodative, and it takes years for them to normalize rates.
If for some reason we had an inflation scare (unlikely), the flies in the whites are reds would be the ones that go berserk, and you should be looking to buy closer to the front of the curve. However, I do not think this is a very high probability scenario.
FWIW, I wrote a blurb in the CA newsletter a month ago, where I discuss why the yoy inflation measures will pop around 1.3% in three monthly data prints early next year. Here is a link to the excerpts: http://www.curveadvisor.com/2015/09/pce-inflation-will-rise-in-q1/
Curve AdvisorKeymasterOctober 29, 2015 at 1:24 pmPost count: 612
We are on the verge of another regime change. That last FOMC statement was unusually hawkish. So depending on how the data progresses the next 1.5 months, we could see a new “post-liftoff” regime. Here are some general thoughts I have on the new regime:
* The markets have been used to false Fed hike promises, so the belly will lead the selloff initially, but eventually, the whites (and reds) will catch up.
* The curve is so flat now that all parts of the curve could lead a selloff. Eventually, you should see the long end of the curve start to flatten. This is because hikes further out will be pulled in.
* Unlike the previous cycle, this will not be a long string of hikes. This is because the Fed is not hiking based on economic strength or high inflation. They will be hiking to remove *some* of the excess accommodation. The hike may cause growth and inflation to slow, so unless the economy picks up steam, the hiking cycle could be short-lived. So I think the Fed may hike 3-4 times, after which the need to hike further will depend on a pick-up in growth and inflation.
As with any other regime change, be wary of relying on historicals – put less weight on them than you normally would.
bambamParticipantOctober 30, 2015 at 10:56 amPost count: 24
“The curve is so flat now that all parts of the curve could lead a selloff. Eventually, you should see the long end of the curve start to flatten. This is because hikes further out will be pulled in.
* Unlike the previous cycle, this will not be a long string of hikes. This is because the Fed is not hiking based on economic strength or high inflation. They will be hiking to remove *some* of the excess accommodation. The hike may cause growth and inflation to slow, so unless the economy picks up steam, the hiking cycle could be short-lived. So I think the Fed may hike 3-4 times, after which the need to hike further will depend on a pick-up in growth and inflation.”
I totally agree with both views, how would analyse value with options/futures based in this view(opinion).
Let’s say, play white vs reds, or reds vs greens? or white vs greens ( I know you like large legs :D).
Curve AdvisorKeymasterOctober 31, 2015 at 2:34 pmPost count: 612
Analyzing “value” is a bit difficult in the front of the curve this environment. Historicals won’t be as useful, because we hadn’t really been in this position before. I suppose you could argue we were here before the June and September meetings, but 1.5 months before those meetings, I didn’t think a hike was even close to “likely.” The Dec is the first meeting where I would put the odds noticeably above 50%, assuming the data comes in-line. With historicals being less important, because we may be changing regimes, what you are left with is a judgement as to how you think the central bank will weigh the odds and/or an opinion as to how you think the data will come out.
One of the core trades we have done is to buy Z-H 3 mo spread. This was not-so-great during the last payroll, but considering the data has been poor, it has been a good trade to buy on a dip. I mentioned in the newsletter last week that I liked buying 0EH put structures on a rally, but this statement was more hawkish than I would have expected, so we never got to do anything. Other ways you can express a view is to buy a spread in the whites vs selling a spread further out the curve. But it’s hard to say in a vacuum. The ideal structure for you will depend on how you see the cumulative distribution of hikes by each meeting date, and how you see the future probability distribution of hikes.
[Note: I will move the last few posts into the “Regime Change” thread next week]
Curve AdvisorKeymasterNovember 1, 2015 at 9:44 pmPost count: 612
Getting back to your question, I think a shorter spread is better than a wider one. I discussed this in this weekend’s CA that the Fed wants to hike not because the economy is showing strength nor because inflation is high. They are hiking mostly to remove some of the excess accommodation. As a result, you would not expect a long “consistent” hiking cycle. They will remove some accommodation, and see where the economy is. After the first few hikes, you would expect the economy to slow (slightly) and for there to be some downward pressure on inflation (all other things being equal). Since growth and inflation are not on any kind of strong/firm trajectory, this should result in fewer hikes after the first 6-12 months, than in a typical hiking cycle. So I think it pays to focus in the next year or so. The greens led the selloff last week, but I’m not so sure this is “correct,” for this environment.
Curve AdvisorKeymasterDecember 14, 2015 at 12:39 pmPost count: 612
This upcoming week should be the most interesting trading week of the year. We will get liftoff in US rates. The markets have been anticipating this move for years. You should probably mark December 16, 2015 on a list of “key dates”, that accompany your historical analysis worksheets. The reason you keep such a list is to identify points where there may have been major (or minor) regime changes – points in time where the way the way the markets trade could vary. As I always say, you never want to use historicals if you will be comparing apples to oranges. Keeping a list will help you identify where the “turning points” could be. Rather than just looking at the “standard” 3mo, 6mo, 1yr time frames, you may want to also create a custom time frame starting with your key date (or a more relevant point around the key date). Doing so could help you identify how a structure will trade in the new regime more quickly. I mentioned a “more relevant point” because it may be that for whatever reason, there is some excess noise around the key date, or the move involved in the key date was transmitted much earlier. Make a note of this date as well.
You must be logged in to reply to this topic.