@TheDude , thanks for your reply. After having some time to look in depth, here are my responses/questions:
By market neutral I meant that the profitability of the trade does not depend on a long/short direction. The trade will be profitable if one product (or group) is relatively stronger or weaker than another product (or group). Is this not accurate to say it's market neutral?
Definitely I think understanding what drives a market is important. But for a short term trade, the driver is often not fundamental or explainable at all, and even for longer term trades to some degree this is true. Sometimes the fundamental driver is clear cut, but in today's markets much of the activity on an intraday basis is not created by anyone with a real fundamental reason. Would you disagree?
(edit: I am looking at the chart at the moment but I don't see it apparently. What day was this?)
Great example--so you're saying that you saw H4 was much weaker than J4, got some info from our main man FT, and then felt that a reversion was imminent, so you bought the H4/J4 spread. Right? By the way, you plot the ratio of H4/J4, or the difference? I would think the ratio would be a more accurate way, but maybe with the same product and such similar prices, a change in percent will roughly model the absolute change in price.
Do you just scour various markets and look for anomalies like this, or do you have a set list of markets you track, or...? In other words, how did you find this inefficiency?
Excellent, thank you again for the reply, this will be helpful.
Last edited by josh; September 19th, 2013 at 12:09 PM.
The following user says Thank You to josh for this post:
Perhaps because there is much more chance that a cleverly-built spread be mean-reverting than a given security. And there are statistical tools to check this capacity of mean-reversion. On this topic, you could refer to last book from Ernst P. Chan.
The following user says Thank You to Nicolas11 for this post:
I really should try and stop posting if I havent read the full thread I should also stop posting when Im overly tiered. Sorry for the confusion....
I trade commodity calendar spreads usually holding for a few days or weeks - so they aren't really market neutral. They will tend to drift in the same direction as the front month if you're looking at the first few spreads or so. I see now though that your looking at ES-NQ which is kind of market neutral in that your trading the over/under performance of one index against the other rather than the direction of the stock market. This trade could offer great opportunities for long term trades around economic cycles. From memory, the tech sector outperforms the market in the early recovery stages. I dont trade inter commodity spreads because I got my fingers badly burnt in a live cattle-hogs trade. The the short went sky high, the long puked! Im scarred from that! Cals are nice and simple. 1-1. No ratios or anything.
I didnt realise you were looking at trading these intraday. To my thinking, any spread or market is going to be driven by news and sentiment primarily. Intraday, its mostly trends caused by people moving in and out of positions and trading round positions. So a daily trend will persist for longer and is easier to get a grasp of. I intraday outrights. Nothing wrong with day trading spreads. Loads of people do.
As for RBOB:
Please register on futures.io to view futures trading content such as post attachment(s), image(s), and screenshot(s).
As you see, it starts breaking out of a range around the 9th, and really gets going on the 16th Margin is $550 init, $500 variation. If you entered on the close of the 9th, you'd have just under $4000 on a 10 lot taking very little heat. The spreads not that wild because your trading the 2 adjacent months. If you traded Jan-April, you'd probably have more, but obviously the margin may be more.
Look at Gasoline. There's still a massive distortion in April. I really like this page for looking at commodity curves. Hogs wheat and Nat gas are all looking a bit funky at the moment. Its also a good sanity check - you want to be short spreads (narrowing) contango markets, long (widening) spreads in backwardation. Same idea as trading with the trend.
Last edited by TheDude; September 19th, 2013 at 03:52 PM.
The following 5 users say Thank You to TheDude for this post:
Check equity index and select ND-nasdaq, it will show synthetic pairs, the ratios recommended by CME and the margin credit offered by them.
Regarding z-score, it s mainly used for pairs that show fairly good mean reversion.
The spread pairs like synthetics (ES:NQ,CL:BRN,Corn:wheat), calenders have a tendency to diverge and form a trend...its these spread pairs for which TA indicators can be applied.
The spread pairs like butterfly, condor, synthetic interexchange( CL vs WBS-ICE or cash vs futures or interexchange corn/wheat spreads) show a tendency towards mean reversion where z-score works well.
I am still training with Peter Hamby (SpreadProfessor), but my understanding is this.
High correlation, High co-integration---->trade for mean reversion
High correlation, Low co-integration-----> trade for diversion/widening of the spread pair
Basically, the opportuniy for a spread trade arises when the correlation is out of place temporarily, so pairs with correlation coefficient of 0.95 and less but more than 0.85 are favourable pairs.
Look how prices are in a fairly even state of carrying charge from Jan to March, then there is a massive jump in the price in April, where the curve goes into backwardation to the end of the curve. Thats what the Scarr Trading page is showing - prices along the curve, not OI.
I've never really used OI much in spread trading, but now you mention it, it could be a great thing to look at combined with COT data perhaps?
The following user says Thank You to TheDude for this post: