Search results

If you want to run 1 yr worth of ES futures tick data through your strategy, you are talking about a few billion ticks/bids/asks. When you try to pass each of those billions of data points through your strategy, the choice of your language matters. I have a C++ implementation (btw: my...

I am no expert here; just sharing some thoughts/issues that I have come across when I looked into this area; some of what I write here is the advice that I got from some deep learning experts who happen to be my colleagues at the company I work for. One particular thing that I have noticed when...

Thanks for the link... very informative; here's a list of interesting deep learning literature (not related to trading though): https://github.com/robertsdionne/neural-network-papers

https://mygtc.gputechconf.com/form/session-listing&doSearch=true&queryInput=&topic_selector=Finance check out "S6589 - Algorithmic Trading Strategy Performance Improvement Using Deep Learning" Its not a consistant 70%+ through out the day... their probability increases after trading starts...

I attended Nvidia's GPU conference here in the SF bay area a couple of months ago. There were atleast 2 presentations related to deep learning & trading. https://mygtc.gputechconf.com/form/session-listing&doSearch=true&queryInput=&topic_selector=Finance Interestingly both are from Japan. Once...

I too looked at deep learning sometime back, but decided to get my normal hand written strategies into production before venturing into deep learning based indicators. Do you see any advantage in opting for deep learning based algorithms compared to having something that we know how it works...

You can set market if touched and trade successfully as long as you make sure that there are no gaps in the limit order book. I have seen scenarios in some markets where the price touches the entry price, but the best counter quote is a few ticks on the other side. As long as you avoid such...

I am assuming that the 400ms accounts for the network/broker latencies; time between your system places the order to the time at which the exchange places it in the order book? I use a small number, but, I guess where you are going with this. What matching algorithm are you using for your fake...

Accepted You are right... if one wants to respond within 1us, we have 1000ns/0.2 (4 ghz processor) = ~5000 instructions... of course we can argue about hw optimizations (l1/l2 cache misses etc), ... how much "useful" functionality can one pack within 5000 assembly instructions apart from the...

Read my post again; I never said that HFTs don't/can't change their algos and I never stated that you can't change algos in FPGAs. What I stated is this: if you want to design your own ASICs using highly optimized lookup tables in order to get 850GHz+ (bitcoin folks achieved this), it is highly...

You can use point specifics ASICs like our Bitcoin folks and get to 100x faster than intel xeones, but the problem with that approach is that any minor tweak (say change a + to a -) in the algorithm will cost millions of dollars + months of wait time... the algos will not be adaptable which is...

You can't incorporate intelligence into a system that has to respond under 1-2us... which is pretty much the realm of HFT. A simple division takes a minimum of 20-30ns (approx/avg) and so you must very methodical/through about the operations you want to use to make your system respond within...

This is probably the only quote I have come across so far that validates what I felt when I was initially designing the architecture of my trading system. Several times I ended up questioning myself if spending quality time in getting the design right is more important than developing something...

Lets face this: you CANNOT compete with HFTs in speed; they are faster than you! ...however, they are not as smart as you and so theres your edge. Smart trading is the way to go against HFTs. HFTs are pretty dumb; it is not possible to incorporate compute intensive AI into a system that is...

What round trip times are you using to estimate your PNL for 350 E Cermack vs theomne.net? As I see news events can trigger liquidity crunch and I suspect that HFT guys turn off their algos... if they do so then it should be for a good reason. You might still get away if your algorithm can...

no of best asks/bids + trades is approx 60 entries/sec for ES (@11:12EST)... now, if I include all asks/bids to this, I am getting around 300 entries/sec...

Looks like I am confused with the term "NBBO" :) Today, I need three pieces of data: trades, the lowest ask/highest bid at that time and aggregated sum of the 5 consecutive asks & bids... I get all this in the datafeed from my broker. I can visualize the strategies that can filter out the...

I was referring to storing the non-nbbo data for future backtesting (if it is useful in the first place) in the above question...

Hi Everyone, I have been working on candle stick based strategies and I am now looking at extending them to ticks/quotes. I can envision strategies that depend on NBBO, T&S and aggregated quotes on both sides; a lot of valuable market information can be derived from these three parameters... I...

I have been using Rithmic since the past few months; they are quite responsive. I exchanged a number of api support related emails with Rithmic. Their response to my queries was never delayed by more than a day. RAPI in general is pretty well designed for whatever it does. Rithmic releases the...