Fully Automated Stocks Trading

[ValeryN]

What is your opinion of QuantQuote? They are not cheap, but certainly not $13 mil. Supposedly has delisted stocks info as well.

Last time I checked the cost is $20K for tick data for all instruments. $9K if you just want 1 min bars instead.

I'm still considering them, but have read mixed opinions. Still, being able to test intraday without survivorship bias is tempting.

Post here about your experience is you end up getting it.

To me - looks just like a dozen of others like that. No opinion.

In general - the more data you deal with the more time you send on QA and I just don't have the time. So I stick to what proved to be a high quality for reasonable price.

If someone had a great success with intraday data from a particular vendor and can comment on quality going back at least 10 years - please mention it, I bet many people will appreciate it.

PS. Heard very good things about NanexCore from other traders. But it is an overkill for many. They offer historical tapes back to 2004.

Val

 

[Same Lazy Element]

I'm curious what everyone else does if you're trading a systematic strategy. If possible - with a context on how you came up with the strategy. Pure data mining without having any idea why it works or you formalized some long-term discretionary observation etc.

My approach (granted, we trade different products and I am much more driven by capacity) is to scale the strategy down 50% if I hit a drawdown that's bigger than max DD and gradually scale it down to de minimis size if i don't see a recovery. Now, all my strategies are based on specific flows or behaviors of the market players so I have a fair amount of confidence if something is going to stick around or stop working.

Last time I checked the cost is $20K for tick data for all instruments. $9K if you just want 1 min bars instead.

There are plenty of solid providers in that range. In fact, TRTH will probably something similar per year. Algoseek 1min TaQ bars is probably the most affordable option if you don't care about lower time frames.

 

[Same Lazy Element]

If someone had a great success with intraday data from a particular vendor and can comment on quality going back at least 10 years - please mention it, I bet many people will appreciate it.

It's a tricky question. If you are trading a small set of instruments and have a quality data feed, the smartest thing is to record your own. A bit more work but also you get a lot of auxiliary data etc.

Otherwise, it depends on your budget (and 20k for someone who's running a 500k portfolio should not be acceptable, IMHO). I had reasonable experience with TRTH in an institutional settle, as it's an industry benchmark. I have played with Algoseek - they do offer TaQ bars which is nice and they give you a reasonable set of corporate action data to boot.

Tickdatamarket.com is OK for liquid futures and tickdata.com is OK for both stocks and futures (though they don't have corporate actions which i find annoying). Neither have special symbols (e.g. BTIC for ES) or calendar spreads.

 

[ValeryN]

Would greatly appreciate if someone can share their experience with Alpaca if any. Posted detailed question here

Feel free to PM me if seems more appropriate.

They seem to have started Beta testing for Canadians. Hence looking into it now. Otherwise we don't have commission free equity brokers with APIs, as far as I know.

Val

 

[mhparker]

I have whole-market as-traded tick data for 2004-2016 in the form of Nanex NxCore tape files. I will sometimes use those to spot-check strategy ideas modeled with daily bars. For my purposes (portfolio-level short-term stock trading, similar to what Val is doing) daily granularity turns out to be sufficient. The extra precision from tick-level order fill price and sequence modeling gets lost in the noise of the much larger factor of the future being different from the past (non-stationarity). In other words, the tick-level differences average out over time. A lot of effort and money can be wasted chasing false precision in backtesting. Keep it simple.

 

[ValeryN]

Today's IB outage lead to my systems making extra 1,659.93$

Too bad I don't have anything on my shopping wish list.. Maybe some fancy paddle board?

It strikes me how so many people complain about loosing everything due to outages, for instance - Robinhood had massive ones this year and there lots of stories.

But no one really comes out saying - I made a fortune due to my brokers' outage. My theory is - it is not due to a lack of people who ended up with a fortune as a result, it is just when something unusual happens humans tend to look for the reasons that makes them look or feel smarter. Being a rich person as a result of broker's outage doesn't look too smart.

Unless we're talking about @mhparker. Marsten is way too realistic and honest

What was a biggest one time $-figure anyone here made as a result of a mistake/bug/not planned side effect?

Mine was probably somewhere around 5-10k USD.

Val

 

[horizon]

How does the outage create an extra gain? Would you please explain it more details?

 

[Same Lazy Element]

What was a biggest one time $-figure anyone here made as a result of a mistake/bug/not planned side effect?

Anyone who's been in finance for a long enough time would have stories like these, as well as fuckups that cost you money.

The biggest one was definitely a result of a pure error, I've made 7 "units" during the flash crash because of a booking error. There was a bunch of far OTM options that I've been meaning to sell and asked my junior trader to move them to the market making book as liquidate-only. Instead, he apparently moved them to the book that held the rates and div hedges. We only discovered the booking after they became "not so far OTM". Pure luck, if not for flash crash, we would have bled out like two hundred grand.

 

[ValeryN]

How does the outage create an extra gain? Would you please explain it more details?

In this case it was due to a positive slippage. Some stocks, between open and the end of outage have crossed the point where they became a "bargain" and went further, making them even better deal. So by the time outage was over the entries happened close to market prices at the time, effectively creating a positive slippage comparing to the "model".

Noteworthy - in those cases an outcome of the trade is irrelevant. As I can quantify the impact vs model immediately.

Val

 

[ValeryN]

Anyone who's been in finance for a long enough time would have stories like these, as well as fuckups that cost you money.

The biggest one was definitely a result of a pure error, I've made 7 "units" during the flash crash because of a booking error. There was a bunch of far OTM options that I've been meaning to sell and asked my junior trader to move them to the market making book as liquidate-only. Instead, he apparently moved them to the book that held the rates and div hedges. We only discovered the booking after they became "not so far OTM". Pure luck, if not for flash crash, we would have bled out like two hundred grand.

Thanks for sharing!

Did you get a bonus for that one?

Val