Curve AdvisorKeymasterMarch 12, 2016 at 11:03 pmPost count: 612
Some of you may like to buy something and sell it out for a very small profit. I suppose the common names for this practice is “jobbing,” “scalping,” or “flipping.” Some reasons you may do this are: you have limited capital, or you like to keep busy, or it helps you to get a feel for how the markets are trading, or you’re just really good at picking up little bits of P&L there and there. There have been many months when I’ve traded over a million futures contracts. MANUALLY! This wasn’t some kind of algo I was using – either I or one of my brokers would enter each trade into the system. So I know a thing to two about buying something to flip out a little higher.
In this thread, I wanted to discuss some things you can consider doing, if you want to manually flip trades. Of course, the main prerequisite is a very low cost structure. This is because what you need in a structure you flip is some sort of either “boundedness” or some sort of mean reversion. It’s foolish to attempt to flip a trade that is directional – you would be basically attempting to buy (or sell) something that has a lot of downside to flip for a tiny gain. Most structures that you can try and flip are going to involve multiple contracts, and to buy and sell all those contracts will require a low cost structure to show any kind of meaningful gain.
Curve AdvisorKeymasterMarch 15, 2016 at 9:17 pmPost count: 612
I suppose there are many different types of “jobbing.” What I primarily focused on was jobbing around a core position I believed in. So I put something on, and tried to add more at levels below and take off at levels above. This has a number of benefits:
* I have on a position that I like – something that I think is good risk/reward. If you are going to job something, you need a structure where if you fail to flip quickly, you don’t mind holding onto it. If you can’t inventory something, it’s very difficult to trade.
* If I add more at lower levels (assume I am long), it’s a bonus, since it is my core view and I don’t mind adding. Presumably you bought a trade that was near a “bottom” so any point below should be a welcome addition.
* If I take some off at higher levels, I’m locking in profit on my purchases and the overall value of my holdings goes higher.
* With interest rates, you can always find related trades to do on a slightly different part of the curve – and around those parts of the curve, you can find cheap structures and rich structures.
Curve AdvisorKeymasterMarch 18, 2016 at 9:30 amPost count: 612
There are two types of flipping (note that I just made up these names):
“True” flipping – where you buy something on a dip (or sell something on a pop), with the intention of selling higher (or buying lower), when the curve “normalizes.” However, in order to do this, you need a rich/cheap view of the curve. You want to focus on buying in the areas of the curve that are cheap and/or selling on the parts of the curve that are rich. You should always have a contingency plan in case you can not flip, and the buy or (sell) starts moving the wrong way.
Example – Let’s say you see value in being long the EDU8-U9 spread and EDU8-U9-U0 fly, and possibly some of the surrounding spreads and flies. You may want a base position (or if you want to minimize inventory, you can start with no position). Your intention is to buy U8-U9 spread on a dip. If you buy on the bid, you plan to work the offer so that you can “flip” the spread trade. However, if it does not look like the offer will trade, you can always sell U9-U0 as a back-up to convert the spread to a fly. In this way, you have multiple ways of making money: (1) you can flip U8-U9, (2) you may just like holding U8-U9 since you saw value in the spread, or (3) you can convert to a (safer) fly structure, since if you saw value in the fly (it should increase at some point).
Next: “Value” flipping
Curve AdvisorKeymasterMarch 21, 2016 at 5:03 pmPost count: 612
“Value” flipping is where you trade around a particular position, taking profit on the pieces that are high and buying the pieces that are low. You should take a look at the Decomposing Flies Thread (http://www.curveadvisor.com/forums/topic/basics-decomposing-butterflies/). This thread describes how you can decompose your fly positions into smaller fly buckets. So assuming that your brokerage costs are low enough and your capital structure allows, you can trade some of the smaller structures around your main structure, as value emerges.
Example – Let’s say you see value in being long the U8-U9-U0 fly and you have a core position in that trade. Let’s say you bought the position at 4bps. Assuming the flies are relatively flat in this portion of the fly curve (there isn’t a severe sloping in the fly curve), the year fly can be decomposed into 4 six month flies or 16 three month flies. The approximate value of the 6mo flies are 1bp each (since you bought the year fly at 4). Note that you need to look at the turn-adjusted prices and you should not accumulate too large a position in any Z contract (as that may indicate an incorrect turn). So if you can buy a 6mo fly in this part of the curve for 0.5bps, you are actually picking up 0.5bps of value on the curve. Conversely, if you can sell a 6mo fly in this part of the curve for 1.5bps, you are picking up 0.5bps of value as well.
Curve AdvisorKeymasterMarch 22, 2016 at 8:40 amPost count: 612
Example (continued). If the shape of the 1 year fly is not flat, but is sloping (for example the U7-U8-U9 fly is 8 and the U9-U0-U1 fly is 2), then you need to make some adjustments to the levels that you work on the 6 mo flies. A good level on selling U8-H9-U9 6mo fly will no longer be 1.5 – you probably want to sell something closer to 2 to pick up some value. This is because the average 6mo fly around U8 (the center of the U7-U8-U9 fly) is going to be around 2 (the 8 price of the 1 year fly divided by the 4 6mo flies that are in it). Conversely, buying U9-H0-U0 6mo fly at 0.5 may not be attractive any more because of the downward sloping fly curve. The 6mo flies around U0 will be around 0.5, so you may need to buy lower to get more value. You probably want to bid 0.
In effect, what we are doing is taking the 1 year fly, and looking at the pieces of it. You can look in 6mo increments, or even 3mo fly increments. At various times, where will be more or less demand on various parts of the curve, and in volatile markets, you can take advantage of being a liquidity provider.
Curve AdvisorKeymasterMarch 23, 2016 at 4:48 pmPost count: 612
You can also “value” flip calendar spreads. For example, say you see value in the EDU8-U9 year spread at 24. You can buy this year spread there. But keep in mind you can always find portions of it trading richer or cheaper. So If U8-H9 6mo spread looks cheap, you can buy more of that, and if H9-M9 3mo spread looks too high, you can sell some of that. The U8-U9 year spread can be thought of as a combination of U8-Z8 3mo spread plus Z8-H9 spread plus H9-M9 spread plus M9-U9 spread. The sum of these four 3mo spreads will equal the value of the year spread. So each 3mo spread will be about 6 on average (24 divided vy 4). You can also think of it as U8-H9 6mo spread plus H9-U9 spread. There are many combinations of 3mo, 6mo and 9mo spreads to construct the year spread.
And similar to the fly example, keep in mind that there will be an increasing (or decreasing) pattern to the year spreads. Recently, the year spreads closer to the front of the curve is higher, and the year spreads further out are lower. So when you decompose the year spread into 3mo (or 6mo) spreads, you probably want to assign a higher value to the 3mo spreads closer to the front of the year spread. So in our example, the U8-Z8 spread will be higher than the 6 average (especially after you add the turn-adjustment), and the M9-U9 spread will be lower than 6.
Curve AdvisorKeymasterMarch 28, 2016 at 10:38 amPost count: 612
Perhaps I got a little ahead of myself. I should have discussed cost structure. I am blown away by the wide range of transactions costs (brokerage, clearing and exchange fees). Some people pay less than 30 cents and others pay more than $2 per contract (one way). So it is difficult for me to say what types of trades are worthwhile for your particular situation. For example, it may be very worthwhile to flip a butterfly for 0.5bps if you only pay 30 cents, as your transactions costs would be less than 0.1bps. So you can net 0.4bps. But it would be absurd to do so if your cost per contract was $2, as your total cost would be 0.64bps, and you are only looking to make 0.5 bps.
Let’s say your cost was $1 per contract. This would be $8 per round trip on a butterfly trade. This means if you try to flip the trade for 1bp, you would be making $17 on the trade ($25 for the 1 bp gross profit less $8 in transactions costs). If you have to scratch the trade, you lose $8. If you have to take a loss of 1 bp on the trade, you lose $33. So you need to make sure there is enough cushion in your expected view, for this to generate a sufficient return. Even if you thought the trade was 2:1 to make money, you are only barely showing a profit. So you need the trade to be much higher probability than with a lower cost structure.
DCRParticipantApril 15, 2016 at 8:07 amPost count: 1
This is probably one of the most useful websites for STIRs traders, accessible to traders at all levels!! I take my hat off to you Joseph!
One topic that I think would be a great extension to this one (jobbing) would be a thread discussing how to determine WHY a structure is RELATIVELY ‘rich’ or ‘cheap’ to the curve. For example, you may see that a 1yr Cal has overextended due to say a sharp bond led sell-off (or whatever). How do you look at all of the cals/flys/boxes along that 1yr Cal to determine which structure offers the best opportunity in terms of being ‘rich’ or ‘cheap’?
Curve AdvisorKeymasterApril 15, 2016 at 3:03 pmPost count: 612
I’m glad you like the site. I do plan on adding more content, so it should be even better.
I do have the “rich or cheap” thread (http://www.curveadvisor.com/forums/topic/basics-how-rich-or-cheap/), and I still need to add to the “vehicle selection” thread (http://www.curveadvisor.com/forums/topic/single-contract-spread-single-fly-or-double-fly/). Let me look those over early next week and I’ll try to focus on them. It’s really hard to speak about these things in a vacuum. A lot of trading opportunities depend on the little details of the moment. And so what may be a good structure at one point in time, may not be so good at another, even if the price levels are the same.
Perhaps what may be better is if you see something in the markets you think is interesting, write a post about it. If you are super-paranoid about people stealing your trade, you can always get some on first. 🙂 And then we can discuss it in more detail. It can even be a (recently) past “trade” or view. I’ll start a new thread called the “2016 Trade Discussion” thread. Participation will probably be low for a while, but when I have finished some of the other basic features, I will finally do some marketing. I’m not sure if more participation will be good or bad, but we shall see!
Curve AdvisorKeymasterAugust 18, 2016 at 7:35 pmPost count: 612
READER QUESTION: Do you stop out of these portfolio/job trades if they go against you? Or, are all of these things something you would either want to keep inventory of or have another way to get out of?
Before you start jobbing (or do any kind of trading), you need to have a plan. Presumably you’ve identified some factors that make you want to do a particular trade. And rather than hold for a larger gain, you decided that flipping the structure for a small gain made more sense. So once you’ve identified the value, you need to determine the sizing that is appropriate. You probably want to weigh various factors like how much the trade fits your view, and how well it meshes with your portfolio. You also want to consider factors like what the historicals/technicals look like, what the market positioning is, what the data calendar looks like (and how you expect the markets to react), etc.
After all of these thorough considerations, there will be times when a jobbing trade goes against you. Be like a chess player and think a few moves ahead. Think about the various possible scenarios. Obviously, if a trade moves in your favor, you take profit and the flip is complete. What do you do when it moves against you? Do you add? In what increment(s)? How low could it go? Something like equities or FX basically has no upper limit, so these would not be good jobbing candidates. At some point, is the structure you chose going to require a stop? Or does it fit your view so well that you are comfortable holding a large position? Plan all these things in advance. What you DON’T want to have happen is that you keep scaling in and be uncomfortable with sizing because you have no plan. Don’t be afraid to reassess and admit when you are wrong about something. It may be you started off thinking you would hold, but market conditions changed unexpectedly and you now feel differently about a trade.
But one thing you should consider are alternate exit strategies. If you really like a particular structure, chances are there is value there – especially relative to the rest of the curve. Let’s say you really like buying EDH8-H9-H0 year fly @ -3. And to your surprise, you find it at -5. If you saw value in it at -3, there is probably value at -5. So stopping out may not be ideal. But perhaps you can do something with it, like selling the H9-H0-H1 fly (which would convert to the H8 1yr double fly) or you can add a slope hedge (like selling a small amount of U7-H8 or another spread you think is more appropriate), as just two examples. Or sometimes adding some kind of (crisis) hedge may be appropriate to remove some of the downside on the trade. It can be hard to say what the curve will look like if H8 1yr fly sells off 2 more bps, but have a few things in mind.
So it’s hard to say in a vacuum whether the best alternative is to stop out, hold, convert or hedge. But you have a lot of alternatives you should consider before doing the trade in the first place.
Curve AdvisorKeymasterAugust 19, 2016 at 10:29 amPost count: 612
READER QUESTION: How do you determine whether the book has too flat vs. steepening bias, mechanically through some kind of scenario tester, or just based on intuition in terms of how trades ought to perform?
One of the things you should be doing as you scale in or out is to look at the effect on your overall portfolio. You generally want to keep your portfolio in some kind of “balanced range.” I’m not saying your book needs to be balanced. And the “range” I am talking about can be quite wide. What you want to avoid though is having an undesired excessive exposure to a particular market move. There are times when you may want a large position as the structure you are scaling into moves against you.
You should look at the effect of all of your scaling trades in three main ways: (1) directional, (2) slope, and (3) curvature. You are almost never going to scale into a directional trade (unless it was at some boundary, like the zero bound). However, many trades you may do will have some directional component. Just make sure you are not adding excessively to a directional view you may not want. Many times, if you scale multiple structures, they could all be leaning the same (directional) way. If this is the case, pick the ones that are most attractive to you, or adjust your sizing accordingly.
There are a few things you can do to see how a structure adds to the directional/slope/curvature risk in your portfolio:
• Historical regression. Have the level of rates be the x variable in a regression. The beta of the regression will be indicative of the directionality of the structure in a vacuum. Keep in mind that history is not always a good indicator of the future.
• Value-at-Risk. If you have access to a VaR calculator, you can see how your historical risk changes as you add more of a structure(s). If you can see the actual 5-10 VaR days, you can see if most/all of your VaR days occur on a rally/selloff or steepening/flattening. Keep in mind that VaR tends to overstate risk for trades near a “boundary.” It may be helpful to adjust or even remove certain trades from a VaR analysis if you believe it is stable and self-contained.
• Daily P&L. You can take a look at your daily P&L for confirmation of your risk. If you P&L moves excessively as you add (or not enough), you can adjust accordingly.
bambamParticipantAugust 21, 2016 at 5:03 pmPost count: 24
Hey, nice to you posting again.
It is also important be aware of news/events in order to evaluate liquidity.
Getting something on the book and having to puke it because liquidity is drying up can be costly.
Legging into the trade is key as well, but you must have a plan b on what you can hit if its going against you. Sitting a on the bid on the butterfly ladder will never get you filled(at least on BAX).
Curve AdvisorKeymasterAugust 25, 2016 at 6:39 pmPost count: 612
Thanks for contributing the tips.
I have a posting schedule: http://www.curveadvisor.com/forums/topic/announcements-2/#post-3221 I took a little time off here and there for the summer, but now that the kids are back in school and people are back from vacation, I will probably post a little more.
Curve AdvisorKeymasterAugust 25, 2016 at 6:52 pmPost count: 612
When you start jobbing multiple structures, you need to take a step back and look at the effect to the overall risk in your book. You generally don’t want to be jobbing a number of structures that all lean the same way. For example, if you decide to job EDH7-U7 spread and EDH7 1 year fly and EDM7 1 year fly and H7 1 year double fly and FFK7-Q7 spread, you are basically jobbing five versions of the same thing! So what you will end up with as the market moves “against” you is you will start having a massive directional, slope and/or curvature exposure that you may not want. If you want that exposure, it’s no problem. But more often than not, you will end up with a position that you do not want. Plan ahead for when you get filled.
It’s okay to job a few structures with overlapping risk profiles. There could be times when one version may trade easier than another. However, you want to avoid doing the same trade in too many ways. Take a look at how the curve has moved and select the variations that you think provides the best risk/reward. Just do the best ones. This view may change over time. If it does, you can change your focus.
You may at times look to do a “balancing trade.” This could also be a jobbing candidate, but look at a trade that you feel should work out if the other trades move against you. Having a somewhat balanced book of jobbing trades (i.e. a not-so-correlated bunch of jobbing trades) will help you keep going, even in “difficult” markets.
Curve AdvisorKeymasterSeptember 1, 2016 at 7:46 pmPost count: 612
Have a back-up plan. For example, let’s say you think it will be difficult for Z8-Z9-Z0 year fly to get above 0, for reasons X, Y and Z. So you sell some at -0.5. You sell some at 0. And then it goes 0 bid. What do you do?
Well, this is something you probably should have considered before you put the position on. Sometimes, you may just want to “back up the truck” and keep adding, because everything is right, and you feel very strongly about it. Other times, you will realize there was an error in judgement, and you want to stop out. However, most of the time, you will find yourself somewhere in the middle – you still like the position but not enough to keep adding indefinitely. So what do you do?
I think I mentioned this before in the Forum, but my family used to have a high-end cabinetry and interior renovation business (appeared in many magazines). My father used to say that the true skill in a craftsman was knowing how to fix an problem when one occurs. It’s hard to generalize, but there are a few things you can consider.
[to be continued]
You must be logged in to reply to this topic.