I am starting to scalp the crack spread with the exchange contract RBECLE (CQG). Any scalpers/traders here.. i know we can swing trade this but thats not what i am looking at.. i am having a few issues a) My charts not matching the DOM (using CQG).. - i usually have a volume chart for my other trades but does not work at all..any suggestions? - Usually bid/ask is pretty thick but this animal seems very strange.. bids/offers all over at all different depths and not continuous.. so cant really tell order flow easily b) Bid/ask is sometimes 2 to 3 ticks.. sometimes 5 ticks wide.. so are we just placing limit orders? intead of hitting bid/offer? c) Any technicals more suited to this type of scalping?? i use cumulative delta for my other trading but this does not seem to fit Note: I am not a member and thus commissions do eat up but still looking to hold to the next VWAP band so not just a few ticks but more like 15-20 thoughts/opinions are welcome..
Hey, I don't know from nothing but, I would note that when I woke IB's TWS this morning, it reminded me to change out the CL heading from August into September... that might help you right there. (To make sure both platforms are on the same target -- what you've described sounds like what you'd see between a dying front month, and the follow-on/interloper.)
I have tried RBECLEQ8 and U8 and Sept for this crack spread seems lower volume... so i dont think thats it.. but thanks for your suggestion.. this spread is more of a strange beast.. will post a chart and DOM tomorrow when market is in full flow for people to see
I don't have CQG. so can't trace this RBECLE. Let me know which RB gasoline contract and which CL contract you are talking about
Here's the ones from the ice exchange, just replace the codes. (QP = Gasoil. QO = Brent Crude oil.) a) The CQG Codes: The 4:3 Crack spread ratio on CQG is: QP-7.458*QO The 1:1 Crack spread ratio is: QP-10*QO You are heavier on the Oil, hence why the multiplier on QO is higher than the 4:3 (near dv01) ratio. I can see what you're trying to do. Trying to be a s1ck kunt with volume reading on a super highly competitive commodity spread will give you more pain than glory. If you're referring to a Volume Profile chart, I don't think they can be done for spreads on CQG. Also, there is no price-action reading game in these spreads. They are over-run by HFTs. Our brains (reaction time 30ms-100ms) can't compete on their nanosecond levels. The exchanges love to suck off HFTs. They ruined everything. Don't play that game you will just hurt your account and your confidence. It's clear what you want to do... You want to trade a spread (less volatility than an outright) but you want to use the price-action game to do it. This is a dead-end game because it's extremely inefficient. There are HFT/algos which are in every outright product you're spreading and they're beating you individually in each contract. Now that you know this, read; 1. If you want to trade a spread (like the 1:1 crack), you need to do some chart-testing and analysis yourself, or even look for leads/lags in other indices, other commodities... to see if you can find a pattern. Unfortunately I can't give away any edge here because you need to research it yourself. 2. If you want to trade volume... price-action... "bids and offers"... you can't be doing a spread. You'll need the Jigsaw platform or at least the Market-Profile / Volume Profile skills on CQG, and trade outright Oil (CLE) from that. Oil is much more thin than say, the Bund or T-notes, so you might find this game to be hard to learn at the start. Also, at $10 a tick you can ruin yourself pretty easily. b) The bid/ask is of course 2-3 ticks wide. This is because for every Gasoil tick, the Oil needs to move ~3 oil ticks. The CQG ladder showing you the spread is set to "Bid/Ask" by defaults. That means, the 'fills' you think you want to get, require you to get filled Limit on the Gasoil (thicker contract) and then hit market on the QO (thinner). The 3-tick wide gap means you are crossing the spread in both legs. And good luck doing that, I'll say. The exchanges suck off HFT d1ckheads so you'll be left last in queue just about every time, even if you want to do a nice and clean bid-the-spread move. The exchanges (CME, ICE etc.) all love sucking the nobs off HFT/algo guys and never forget this! If you don't believe me, see how much you get legged on your CQG autospreader by trying to get conservative prices in the crack. I strongly do not recommend leaving a quote in to do this without a sufficient Volume Lean (conservative amount). But even then, when Nymex session kicks into gear, it will be like shooting crack into your eyeball with how much Oil shoots around. I've been legged 5-6 oil ticks using a conservative volume lean in that crack spread, with a "market v market" execution. Unfortunately, the exchanges let HFT's cheat you and they always will because they make money off volume... it's not even good enough to have a co-located autospreader server located in London, next to the exchange.. If you charts are not matching the DOM then you might need to manually put up the correct months. e.g. QP?2-10*QO?1 will give you 2nd-contract QP and 1st-month QO. c) Technicals for scalping the Crack-Spread... Try putting a moving average, or Keltner, or Bollinger band... or just try a VWAP Evenlope. Experiment with time-frame. Once again... I wish I could give you answers but we'll need someone with CQG back-testing skills to give us the optimum standard-deviation to fade for the crack-spread if we want to get anywhere. It's not enough to go off "gut feel" anymore (this isn't 2015 easy money days) Plot it out and you'll see a sharp and distinct difference in the spread comparing London hours to Post-Nymex session open hours. And watch out for Oil inventories. And watch out for random Saudi guys making comments on TradeTheNews... I think my Oil leg went like 100 ticks offside against me because of some random comment a few months ago and the Gasoil couldn't catch up. I had to average and hold for days. Even after being on-top of this game, you will probably have a pitchfork sharpened because you are going to learn how much the exchanges love sucking off the HFT's. You will get legged everywhere so make a conservative volume lean and if you're doing bigger size, use icebergs of 1-lot minimum increments.
great writeup. I don't have ICE data. I have the CME nymex data. Which are the popular crack spreads from NYMEX? By popular, I mean in terms of liquidity / volume.
Unless you have access to OTC order flows I would politely suggest you reconsider. Swing trading would be a more productive approach. Even taking two or three well considered trades intraday would be a better risk/reward skew than literally "scalping" that market. I have considerable experience in that market, and you can't consistently buy bids and sell offers without getting your ass run over. I wish you good fortune with your endeavors.
Thank to all for comments.. and thanks to s0mmi for such a detailed post.. im gonna articulate a bit more before i ask for any more comments a) I am not looking at scalping but more several day trades (thanks bone.. as soon as i read it, i realized my mistake) b) I am looking at looking at the pattern and trade it at certain levels (a refinement would be to see what influence external factors have on it like movement of crude/rbob.. though at this stage its not apparent as i have not backtested it at all) c) Technical analysis wise: i just noticed the VWAP bands (i guess same as VWAP envelop seems to as good an indication as any (see charts below) d) Thanks for clarification on the multiple points regarding bid/ask ticks and HFTs, s0mmi - still if I am looking to play the extremes or breakouts at certain VWAP levels, there seems to be a decent risk/reward (even after loosing a few ticks) Note: I dont have access to ICE data but this is Nymex RBECLE U8 as the exchange spread (which i am assuming has already taken the calculation into consideration) anyway more later once i get some positive results (or terrible results) today and tomorrow in SIM First chart is a 2min chart with VWAP bands, next is the same with Heikin Ashi (but i know this looks nice but not real) and lastly a Flex Renko from Sierra (using CQG data) which looks great but even more unreal (i will post thumbnails going forward) Regular 2min 2min Heiken Ashi 321 Fex Renko
But as you have already found out, the market price action is much more 'chippy' than the way it charts.