Tagged: hedge or not
Curve AdvisorKeymasterApril 28, 2015 at 12:00 pmPost count: 612
Here are some general thoughts on when to add a “directional hedge,” and when to just put on the structure (where the hedge = 0).
* The main factor in deciding what hedge to use (if at all), depends on your directional view. If you are using the trade to express that directional view, then you should lean towards not having the hedge. If you already have the directional view on in large size, you probably want to lean towards having the hedge.
* Make sure you are looking at a comparable historical period to determine the “beta” of the structure to a directional move in rates. For simplicity, I generally assume you are using linear regression, but you can use other statistical methods to determine your “beta.” You will find that the 3mo, 6mo, 1 year betas can vary significantly (the time period does not have to be a round number). Many times, this may be because of a regime change in the way rates are trading, and caution should be taken.
* Sometimes, you may find yourself with a “beta range,” instead of just one number for the beta. I generally incorporate my directional and future view. I know this is not very precise…
* Even if you are the master of data manipulation, at the end of the day, there will always be an “art” component to choosing the beta. Don’t be overly attached to the number your model spits out. The model is based on the past. You are trying to look into the future. As they always say, past performance is not a guarantee of future success.
I may post some examples in the future as I run across them.
DINESH SINGLAParticipantApril 29, 2015 at 1:56 amPost count: 4
I generally don’t have the view on the directions of the markets ( mostly my assessment is wrong). So i generally look for the trade which looks good relatively.
After the NFP reading in march, short sterling strip sold off heavily ( more than it should) and flies bid up aggressively. And after the reversal of 15-20 bps the 5th generic fly was still at +8 bps. I wanted to sell that fly with outrights short so that i could hedge my short flies on further sell off.
But I was not convinced with the ratio which i get by linear regression. I just did it with guesswork.
Can you suggest me a better way to calculate ratio? Or any other suggestion for the thought process.
Curve AdvisorKeymasterApril 29, 2015 at 4:09 pmPost count: 612
I don’t follow the UK very closely, so I can’t comment regarding the specifics (especially since I don’t know what the trade is). But in general, if you are going to try and figure out the hedge, I would suggest the following:
* Pick an appropriate comparison period. A lot of times, we pick a round number like “1 year”, but that may not be the optimal time frame, depending on what happened in the previous year.
* Plot the data (you can look at it multiple ways if you want), to make sure a “linear” regression correctly captures the “directional” relationship of your structure. I like looking at a scatter plot with the structure price on the y-axis and your directional hedge on the x axis. Make sure the plot does not show a “regime change” (where the plots look like there are two different environments plotted). When there are multiple trading regimes, your beta will be less meaningful.
* When choosing the contract for your directional hedge, pick one that makes sense with your view, and/or with respect to the structure. If you have no view, most people will pick the center contract of a fly to use for the hedge.
* Another way to incorporate some kind of direction hedge is to consider doing a trade with differently weighted spread legs. For example, .75x ED5-9 spread vs 1x ED9-13 spread.
I will go over an example in the next post.
You must be logged in to reply to this topic.