Australian (ASX) Stock Market Forum

Trading Spreads

Jump to new
havaiana said:
Another question skc/r4jen

It seems to me like a type of style that could be 'easier' to automate in comparison with other styles
Click to expand...

By the time you finish the Pairs Trading thread you will notice SKC knows ALOT about the companies his trading. A computer does not. He also knows the announcements coming up and the potential effect it may have on his pairs, a computer won't. Many more examples. I suppose to an extent it could be programmed, ie - X% from the mean etc etc but I would imagine your equity curve and trading statistics wouldn't be very smooth with more than usual drawdown. My thoughts anyway :2twocents
 
Hi,
Basically with Spreads the time I hold varies alot depending on the market conditions. I can hold anywhere from a few seconds to multiple days. My philosophy is that I dont want to take a loss so ill average until it comes my way (within reason). Since my primary trade is interest rate futures, the movement of that spread especially trading 3 month or 6 month, isnt as much as other more volatile spreads, so with the right risk management you can average knowing that 99% of the time it will come back. Generally it doesn't come back when there is a significant fundamental shift, ie new policy, interest rate changes etc etc.
But in terms of strategy, I always try keep it simple, its not rocket science to play the ranges, i tend to always try find spreads that mean revert well.
Spreads that act like an outright defeat the purpose.
I do keep a daily blog called day in the life of a bond trader.
 
skc said:
I do however have the same question wrt spreading bonds. It seems like a simple thing to program... (showing my ignorance)
Click to expand...

I am by no means an expert here

Most of the bonds (or more accurately aussie bills, ie the short end futures stuff) is played by the big boys as simple mean reversion. I remember one of the guys showed me a chart of a couple of flys or condors or whatever they were and even during the GFC the spreads between these instruments just didn't get that out of whack. They almost always mean revert (I'm sure there's examples when they don't but I don't know of them).

The big thing with the spreads (atleast in the aussie front 12 strip) is que position. If you can get front of the queue in certain spreads then its hugely advantages. Otherwise you need to 'leg them' which is definitely doable but you take on legging risk by doing so.

In some regards it is an easy concept and I can see what you mean in terms of programming an algo to enter spread trades. They have these (called auto spreaders) but truth be told markets are so tight/efficient in these regards these days that the best spread traders are able to identify where the market is positioned as well as where the buying + selling pressure is in the market and then act accordingly.
 
SpreadShortOpportunity.png

This isn't how or what kind of moves I trade but I keep a chart of it open during the day to keep an eye on the sengs progress relative to the china enterprise futs and thought the example might spark some convo in here. There were a couple successive sessions where the seng was showing strength and the HHI just wasn't holding up. This is one of those scenarios where I see why people like to trade spreads on very liquid markets; fairly low volatility trend all afternoon that allows for the legs to be added to nicely on many occasions.
 
Can anyone tell me what would be wrong with trading this spread once it got to the extremes as a mean-reversion system? As it seems like a good idea to me, but no doubt I'm missing something as I've never spreaded stocks.

bhpvrio.jpg
 
Eh potato potahto :D

Feel free to move it then Mr Admin, although I hope to do this with futures data soon, just easier with stocks atm. :)
 
skc said:
A mean reversion strategy on outright share may be completely valid, but you wouldn't trigger it based on divergence away from the mean of a pair's ratio. If you are simply looking at divergence from mean with the stock itself - then really that's just trading off Bollingar bands or moving average etc. Otherwise you can trade a stock divergence from a sector's mean. But without hedging the resulting equity curve would be very different I'd imagine.

On odd occasions I do use signals from pairs trading to take outright trades, but the position sizing, stop placement etc will be completely different.

Anyway, if you have questions on equity pairs probably best to ask them on the pairs trading thread to keep it all together.



Falling profitability - not really. And if there's any fall in profitability it's due to reduced volatility (compared to say 2009-2011) rather than equity pairs being a crowded trade. Automating equity pairs can certainly be done but there's a lot of discretion in my own trading that's probably not programmable.

I do however have the same question wrt spreading bonds. It seems like a simple thing to program... (showing my ignorance)
Click to expand...

Spreads are very easy to work, target commodities these are far simpler and easier to work out. spreads are extremely rewarding with 1/6 of the margin required and very little risk on entry when your looking for the right setup. Spreads is about understanding.

some of the biggest gains i ever made were spreads :)
 
Today being practice and recovery from the week, i took a look at spreading the DAX off against the FESX. Now for the ratio i just used 2.5 to 1 given that the DAX is a 25 EUR contract and the FESX is a 10 EUR contract. So i was actually using 5:2. I have no idea if this is even right and i'm just going by what i've read.

So, today was an interesting day as the DAX was quite a ways out in front of the FESX in filling the gap as its much more volatile. I have the indicator for pairs but really once you have the ratio about right its just a matter of looking for some plays, i guess.

I'm just trying to figure out how to take profits, as its seems the spread never really gets that far out of whack before it comes back in...

Also, sometimes it looks like it may be better to drop one leg and then take the outright trade if that looks or feels like a better play...

I'm really just thinking out load here so please steer me right if I'm of course...

In an up trending market i'm generally looking to short the DAX and go long the FESX as the Dax corrects itself...its really just an eye ball thing...obviously key levels will play a part here too....
 

Attachments

  • FDAX 12-13 (60 Min) _ FESX 12-13 (60 Min) 10_25_2013.jpg
    FDAX 12-13 (60 Min) _ FESX 12-13 (60 Min) 10_25_2013.jpg
    110.3 KB · Views: 16
  • DAX ESX.jpg
    DAX ESX.jpg
    176.1 KB · Views: 24
After reference Rajen's book again, I'm using an 8:1 ratio, so we'll see how that works. I've also got my charts setup correctly now too...before i had the YT and XT on my correlation charts...:eek:
 
I've been continuing my research into trading spreads and in particular calendar spreads with some seasonal trends. These types of spreads can be traded with a 1:1 ratio as they are intra-market. Inter-Market spreads need careful consideration for the hedge ratio so you get closer to a 1:1 ratio at the time you take the trade.

Here is an example of a play I'm watching.

This is Live Cattle and the strategy here is to sell the February contract and buy the June contract, expecting the spread to reverse and trend down sharply. We first need the spread to give us a signal, a place to enter the trade. We know that we are entering the period in the season where this spread usually takes a dive. Its been pretty strong this year though, but its at the upper side of the channel. If we can make a lower high we will have a place to enter the trade should it turn lower again....

There are a couple of ways to find the seasonal opportunities. One way is to subscribe to a service like Moore Research. The other is to subscribe to a search algorithm service like Seasonalgo.com. Either way you need to be able to mine the seasonal history of the spreads to look for opportunities.

The seasonal trends don't happen every year, but often enough that it can provide a little edge provided that everything else on the trade is agreeing with it.

Also, you must check the COT to make sure the commercial interests are positioned on the correct side of the market. You can see from COT report that the commercials and the funds are at extremes with the funds well long and the commercials fully hedged to lock in their prices...This also tells us that if we do see a move to liquidate longs then it could move pretty swiftly.

There are several Agricultural commentators indicating that they are very long live cattle outright at the moment, so we must wait for the spread to give us signs of a reversal.

We'll need to also check the latest COT report when its issues this coming week.
 

Attachments

  • LC_JUNE.png
    LC_JUNE.png
    27 KB · Views: 7
  • LC_FEB+SPREAD.png
    LC_FEB+SPREAD.png
    30 KB · Views: 8
  • LC_June _ FEb.jpg
    LC_June _ FEb.jpg
    156.3 KB · Views: 10
  • COT LC.GIF
    COT LC.GIF
    31.3 KB · Views: 7
  • LC_Weekly.png
    LC_Weekly.png
    25.9 KB · Views: 6
Larry Williams on Seasonality....



- - - Updated - - -

A great explanation on how spreads work by Jay Richards of JustSpreads...





Enjoy!
 
Last edited by a moderator:
Setting up spread charts in IB's TWS....

[video=youtube_share;GPeScAFUEkU]http://youtu.be/GPeScAFUEkU[/video]
 

This Video does a good job at explaining spread Futures with real examples
 
Last edited by a moderator:
Trading and charting spreads with IB's TWS...this supersedes post 38...

[video=youtube_share;byiKS05v4tk]http://youtu.be/byiKS05v4tk[/video]
 

Similar threads

Richard Dale
  • Article
Resources for Systematic/Algorithmic Trading
Replies
12
Views
2K
Richard Dale
Richard Dale
Fiona96
ATO Tax!
Replies
18
Views
2K
Ibza
Ibza
D
Adventures in AI
2 3
Replies
43
Views
6K
mullokintyre
mullokintyre
ducati916
November DDD
2
Replies
30
Views
2K
ducati916
ducati916
B
BMW's trading journal
Replies
15
Views
3K
BMW
B
Top