edge init

the thread has died so I will attempt to rekindle it.

I have been reading a lot about pair trading and spread trading recently - they are basically the same thing. I love the way you can take out the market (beta) component. Why would you want to do this? Well there are several reasons but one of them is you may have a view on something other than the price. Often professional traders will trade these as they have a view on something created by specialist knowledge from somewhere. Other times the spreads show some price action that is easier to trade.

Lets take a very simple example of a heating oil calender spread (Aug 2012 - Dec 2012). I just knocked up a quick chart in Esignal using the custom spread function. This is a synthetic spread (i.e. I made it up) not an exchange traded spread. Hence to trade this I would have trade each position manually (leg in and out).

Now take a look at the chart. You can see that once or twice a day the spread comes away from the average by some considerable distance and then mean reverts nicely. At 12.00 today London time the spread got really out of whack.

As you can see this is not like trading an individual stock, FX, future, bond where you can be taken out by any manner of 'noise'.

The amount of future contracts you can spread is almost infinite. I would say this is nothing short of awesome.

Also you would think you need masses of money to trade these spreads like the pros. Not so, you can get Esignal for $134 p/m plus exchange fees that you are looking at, use this to chart your spreads. You dont need any expensive spreading execution software you can just execute through Ninja, brokers like AMP/Ninja have very small account minimums and margins for a lot of futures contracts are c.$500.
 

Attachments

  • HOZ2HOQ2 spread.png
    HOZ2HOQ2 spread.png
    56.4 KB · Views: 178
Last edited:
ChocolateDigestive said:
the thread has died so I will attempt to rekindle it.
Click to expand...

(y)(y)

Futures margins are lower on spread trades too because of generally lower volatility. Seasonal spreads, also called calendar spreads, are fairly regular in the grains and soy markets and take advantage of time decay and the seasonal nature of those commodities.

I've always had some interest in spreads but never really looked into trying it.

In my experience pairs trading usually refers to equities in which you pair up 2 stocks. Long a stock from a strong sector, short a stock from a weak sector. You can profit even if both stocks move in the same direction as long as the strong stock is outperforming the market and the weak stock is lagging.

Peter
 
wackypete2 said:
(y)(y)

Futures margins are lower on spread trades too because of generally lower volatility. Seasonal spreads, also called calendar spreads, are fairly regular in the grains and soy markets and take advantage of time decay and the seasonal nature of those commodities.

I've always had some interest in spreads but never really looked into trying it.

In my experience pairs trading usually refers to equities in which you pair up 2 stocks. Long a stock from a strong sector, short a stock from a weak sector. You can profit even if both stocks move in the same direction as long as the strong stock is outperforming the market and the weak stock is lagging.

Peter
Click to expand...

Hi Pete

Yes pair trading normally refers to equities however the concept is the same as spread trading. yep on exchange traded spreads the margins are tiny.

Here is an intraday pair chart of a couple of small finance companies. You can get some great price action. You are unlikely to be whipsawed out due to some general market event. A lot of pairs mean revert which means they can trend nicely on smaller time frames (one mans trend is another mans mean revert and so on).

There is a free tool at this website which allow you to check the co-integration between stocks over different time periods.

Pair Tool | Catalyst Corner

you need to join register at the site to use the pair tool - its free.

when opportunities like this are around I wonder why many people like to take a punt on a single instrument.
 

Attachments

  • ntrsfitb.png
    ntrsfitb.png
    47.9 KB · Views: 171
ChocolateDigestive said:
the thread has died so I will attempt to rekindle it.

I have been reading a lot about pair trading and spread trading recently - they are basically the same thing. I love the way you can take out the market (beta) component. Why would you want to do this? Well there are several reasons but one of them is you may have a view on something other than the price. Often professional traders will trade these as they have a view on something created by specialist knowledge from somewhere. Other times the spreads show some price action that is easier to trade.

Lets take a very simple example of a heating oil calender spread (Aug 2012 - Dec 2012). I just knocked up a quick chart in Esignal using the custom spread function. This is a synthetic spread (i.e. I made it up) not an exchange traded spread. Hence to trade this I would have trade each position manually (leg in and out).

Now take a look at the chart. You can see that once or twice a day the spread comes away from the average by some considerable distance and then mean reverts nicely. At 12.00 today London time the spread got really out of whack.

As you can see this is not like trading an individual stock, FX, future, bond where you can be taken out by any manner of 'noise'.

The amount of future contracts you can spread is almost infinite. I would say this is nothing short of awesome.

Also you would think you need masses of money to trade these spreads like the pros. Not so, you can get Esignal for $134 p/m plus exchange fees that you are looking at, use this to chart your spreads. You dont need any expensive spreading execution software you can just execute through Ninja, brokers like AMP/Ninja have very small account minimums and margins for a lot of futures contracts are c.$500.
Click to expand...

Try it - I think you will find that there are as many hurdles here as with any other type of trading.

Fact is, no matter how you trade, you need to make a bet on a future outcome. Spread trading, whilst it has the allure of being non-direction, is still a punt & you are still competing with other traders for $$$.

With statistical arbitrage your competition is going to be software. It is software that keeps the S&P index & futures in line. That is both on your side and against you.

It's on your side because there is a mechanism to pull things back into like. It's against you because they will be a lot faster than you. They will also be a lot more savvy in calculating what an acceptable deviation is to arb & when to stand aside and do nothing.

It'll be interesting to see how you get on....
 
Well done for getting the thread going again CD. Mixed feelings on the logic though. For one thing the decreased volatility will lead most to leverage up to make an acceptable return which ultimately puts risk levels back where they were for single instrument trading. Secondly, the adage about the market staying irrational longer than you can stay solvent applies here too: spreads can continue widening for no apparent reason longer than you can bear the loss. But... Ultimately this is now how I trade in my day job and if nothing else, the discipline of knowing that the market is not going to bail you out if you get the stock selection bit wrong has helped me achieve better results: but as Dionysus says, just avoid arbing spreads where your competition is big banks' and hedge funds' supercomputers.
 
Agree with DT & Jack here.
Stat arb is quite often HFT turf so impractical for most.
Longer term trades although not directional still carry risk,
especially with stock spreads.
If the correlation is just sector related and the stock you are short
continues and breaks the sector correlation altogether for instance.

E-mini - Wikipedia, the free encyclopedia
http://www.futuresindustry.org/downloads/Audio/Companion/Three-812.pdf
Salient points of above PDF highlighted below:

Just my view, but I think if you are going to try longer term spread trading,
I would stick to U.S. indices.
Different ES contracts or ES against synthetic ES (made up of YM, NQ and ER2)
Or find another sythetic combo:

More examples:
5 Year US Treasury Future – 10 Year US
Treasury Future.

2 Year US Treasury Future – 5 Year US
Treasury Future as a Package against the
10Year US Treasury Future – 30 Year US
Treasury Future Package.

Other pairs could be:
Dax – EuroStoxx
CAC –FTSE

Not so keen on the above Euro indices, as I personally think any correlation is just sentiment based.
Sector and industry weighting are potential correlation breaks.

In a nutshell its free money for the HFT stat arb outfits.
For anyone else there are still risks.
Examples from above pdf:

ES up 5.0 points, YM down 50 on CAT
earnings worries.

GE – HON Take over Arbitrage Spread was
trading at 0.60 and blew out to over 12.00
when the Deal fell through.

LTCM Yield Convergence trade blew up
when Russia Defaulted.

2 Year TED Spread went from 35 bps to 75
bps when plane hit WTC.

It can be a good method, even excellent with the right tools and resources.
It can also be an over complex, over leveraged disaster waiting to happen.
Don't forget, the man with the large cojones was spread trading...:LOL:
I would say his name but he might come back :eek:
 
Last edited:
Liquid validity said:
In a nutshell its free money for the HFT stat arb outfits.
For anyone else there are still risks.
Click to expand...

Indeed - but only to an extent. With 'free money' comes competition - which is what will ultimately kill HFTs.

Let's say someone is finding a 10 unit deviation to arb and doing nicely. Someone is going to come along and arb 9. It naturally gets to the point where people compete and to do that they will enter on narrower and narrower spreads. There comes a point where you kill the golden goose or rather, the costs outweigh the profits.

Fees are also an issue in that you pay 2x, 3x the fees for a spread trade because you are playing both sides. You have double the bid/offer spread to overcome too.

Interesting game though...
 
DT - mostly agree, even HFT stat arb is basically a competition, survival of the fittest.
I'm not sure I think that competition will kill HFT, rather rationalise the practice.
I'd say the newer smaller entrants would suffer most, unless they had a tech edge,
which lets face it with exchange co-location thats a level playing field, at least for HFT...

I still think legislation is the only think that will kill HFT outright.
Either directly outlawing the practise or indirectly by increasing costs - transaction tax.
I don't see either of those being a possibility in the short term at least.
Long term who knows, it would affect all of us though...
Anyway that topic is giving me a sense of deja vu from another thread :D
So I won't derail the topic from spreads to HFT anymore.
 
You have to pick your battles. It would be unwise to go against a HFT supercomputer.

DionysusToast said:
Try it - I think you will find that there are as many hurdles here as with any other type of trading.

Fact is, no matter how you trade, you need to make a bet on a future outcome. Spread trading, whilst it has the allure of being non-direction, is still a punt & you are still competing with other traders for $$$.

With statistical arbitrage your competition is going to be software. It is software that keeps the S&P index & futures in line. That is both on your side and against you.

It's on your side because there is a mechanism to pull things back into like. It's against you because they will be a lot faster than you. They will also be a lot more savvy in calculating what an acceptable deviation is to arb & when to stand aside and do nothing.

It'll be interesting to see how you get on....
Click to expand...
 
DionysusToast said:
Indeed - but only to an extent. With 'free money' comes competition - which is what will ultimately kill HFTs.

Let's say someone is finding a 10 unit deviation to arb and doing nicely. Someone is going to come along and arb 9. It naturally gets to the point where people compete and to do that they will enter on narrower and narrower spreads. There comes a point where you kill the golden goose or rather, the costs outweigh the profits.

Fees are also an issue in that you pay 2x, 3x the fees for a spread trade because you are playing both sides. You have double the bid/offer spread to overcome too.

Interesting game though...
Click to expand...

Free money is when you run an ETF on behalf of a 'client' and you are responsible for rolling the contracts whilst at the same time your prop desk division is spanking the spread on the pre-roll. This libor scandal ain't nothing. Lol

If you chart calendar spreads on roll day you will see what I mean.
 
ChocolateDigestive said:
Free money is when you run an ETF on behalf of a 'client' and you are responsible for rolling the contracts whilst at the same time your prop desk division is spanking the spread on the pre-roll. This libor scandal ain't nothing. Lol

If you chart calendar spreads on roll day you will see what I mean.
Click to expand...

lol....

But that's OK - because they went to work on a different bus and are therefore trading independently with no knowlegde of what the other is doing....

we are a big company with different divisions... etc etc etc...
 
Everyone is talking about HFTs killing spreads but I'm not sure it's the case. When I was looking at this stuff, the only thing that made sense to me was to take a view on the opportunity cost aspect of the pricing mechanism (anyone remember my contango phase lol) in a similar way to how the option traders take a view on the greeks. As far as I could see, that's the only way to viably assess any mean reversion as a change in spread could represent a fundamental shift and your goose would be cooked if you're in a trade there. Imagine you were long BP in an equities pair trade before deep water for example.
In regards to the argument that HFTs eat up all teh opportunities for stat arb, I really don't think that's the case. HFT's are directionally unbiased and are in a out of the market in a flash so why would they wipe out trending opportunities? If anything HFT's feeding off of each other could provide more opportunities imho.
Anyways I'm not really making any significant contribution here as I'm quite sick and I'm just rambling on so I'll go back to lurkyness.
 
Last edited:
scose-no-doubt said:
Everyone is talking about HFTs killing spreads but I'm not sure it's the case. When I was looking at this stuff, the only thing that made sense to me was to take a view on the opportunity cost aspect of the pricing mechanism (anyone remember my contango phase lol) in a similar way to how the option traders take a view on the greeks. As far as I could see, that's the only way to viably assess any mean reversion as a change in spread could represent a fundamental shift and your goose would be cooked if you're in a trade there. Imagine you were long BP in an equities pair trade before deep water for example.
In regards to the argument that HFTs eat up all teh opportunities for stat arb, I really don't think that's the case. HFT's are directionally unbiased and are in a out of the market in a flash so why would they wipe out trending opportunities? If anything HFT's feeding off of each other could provide more opportunities imho.
Anyways I;m not really making any significant
Click to expand...

HFT's is not killing spread trades Scose, they are talking bo77ox. In fact the opposite is true, spreads are great for cancelling out the noise of HFT. if you get me.

Scose - I reckon you should make commodity spreads your niche. You can leverage your brains into that, its also a great way to combine fundamentals with price study. Don't listen to vendors, the extra commish and spread can be avoided to using exchange traded spreads. let those liquidity monkeys take a punt on the ES.
 
Last edited:
No one said HFT is killing spreads.
The conversation went thus:
HFT is the most effective way to spread trade.
Topic then turned to the longevity of HFT (unrelated to spreads).

HFT stat arb won't wipe out any other arb opportunities.
The point raised was that longer term arb (i.e. not HFT)
is more susceptible to margin risk, or correlation breaking completely.
Hope that clears up the confusion :)
 
You must log in or register to reply here.

Similar threads

S
The Famous Edge
Replies
4
Views
2K
Paul71
P
bbmac
your ' trading Edge '
2 3
Replies
43
Views
10K
andrewbee
A
S
Trading and Blackjack
2 3
Replies
46
Views
13K
DeltaGuy
DeltaGuy
timsk
The 'Edge'
2 3
Replies
59
Views
21K
ymonly
Y
K
What's your 'edge'?
6 7 8
Replies
143
Views
32K
foroom lluzers
F
Top