Was just reading Kernfusions post from early August and thinking about trading without market data. I'm not sure how this wasn't brought up more (though I did see a mention in early 2020), but we can use snapshot request to get the real-time quote, which as far as I understand is all that's needed for Rob's algo. It's $0.03 per request. If your exec algo does a lot of requests to get the quote, this might add up. Mine does one request only, and I have to pay professional fees, so this pays off big time. Keep in mind this is not available for some exchanges, details here https://ibkr.info/article/2830. I guess if an exchange is not listed, snapshot requests are not available, so no ICE stuff, but CBOT/CME/COMEX/NYMEX work. So if we combine that with the fact that we can get data history without the market data subscription if we use Code: ib.reqMarketDataType(3) , this can be a viable way to trade without market data.
Does anyone else have problems with Lumber mini contract? I entered buy order for LBR Nov 2022, but I got this warning: "YOUR ORDER IS NOT ACCEPTED. This product is not available for trading." However, if I enter buy order for old (soon to be replaced) Random Length Lumber LB Nov 2022, everything works fine without any warning.
You sure it's LBRX, not LBRU? If that's all clear, then maybe your broker has not yet coded LBR into their system yet? Give 'em a call.
Hey everyone (especially Rob), I've been thinking about including both the full-sized and the mini/micro contracts (e.g., ES/MES, MCL/CL, QG/NG, etc.) in my system and letting the tracking error minimization system (what Rob calls Dynamic Optimization) decide which to trade based on granularity and costs. The biggest issue I can think of is having two instruments in my correlation matrix with 100% correlations, which could mess things up, especially if you do any Fisher Transformations, which I do. Any thoughts? A great idea? Or really stupid? Thanks.
and, today seems the day for questions: Re Dynamic Optimisation: I am trying to work out why GOLD_micro isn't represented in the optimised portfolio. It has one of the largest combined forecasts at -14.55, and its notional position size of -2.411 is larger than almost anything else. I thought maybe costs, but it's cost is relatively low, ranking 160 out of 175 on the sorted SR costs list. Not urgent, but a response might help my understanding (or mis-understanding) of DO. Thanks KH
My MGC position is -1 at the moment (unrounded is -1.17), and I can see it was -4 on Sept 11., my current combined forecast is -16.7. Here's the position and forecast plots (the red line is combined forecast), but who knows, maybe your instrument composition makes things different somehow..
I believe Rob currently has a position in Silver. Since that is highly correlated to Gold, it probably covers both positions. (I hope Rob will correct me if I'm wrong.)
Thanks for your response. The figures I quoted re Gold are from Rob's GitHub reports. I don't run DO mainly because I don't understand it, Carry is in the same boat. I do have a gold position, it is my largest position calculated using Ch 15 calcs. I was just wondering why Gold, given how strong the signals are, isn't in Rob's optimised portfolio. It was pointed out to me that the omission of Gold may be because of correlation to the current Silver position, but, even so, Gold has a much stronger signal than Silver and so should (in my mind) have a higher preference in the optimised portfolio. KH
Gold has a stronger signal now than Silver, but what were the signals when Rob opened up the Silver position? It's also possible that Silver is more highly correlated to other instruments (like Copper) and that too is influencing the position.