Login Page - Create Account

Support Board


Date/Time: Tue, 11 Mar 2025 17:46:40 +0000



Post From: Denali data feed for algorithmic trading?

[2022-04-01 01:41:03]
1+1=10 - Posts: 270
Hi User814333,

I'm a trader that dabbles in algorithmic futures trading.

Is the data cleaned in any way by the server (i.e. zero/negative prices removed, duplicate data removed, etc). I don’t know how much of this comes from the exchange vs the data providers, but just thought I’d ask.

In futures prices can be negative -- definitely keep that in mind in your algo. An example of this was March 2020 crude oil: https://www.cnn.com/2021/04/20/investing/oil-prices-negative/index.html

Regarding errors or duplicate data in a real-time futures feed it strikes me as excessively unlikely. If you're worried the most comprehensive response would be from the CME's data team themselves. You can send your question here: https://www.cmegroup.com/market-data.html#contact

4. Is there anything that would prevent the clients from storing the data they receive in a local data base? For this I’m assuming a custom ASCIL study would be needed, but I’m just wondering if it’s possible. The idea would be to have low latency access to recent data for algorithmic purposes.

Don't charts in SC that display "live" market activity store the chart data locally?

Yes, SC stores the data in local files. Yes, from a technical perspective an ACSIL study, which allows custom C++, can write to a database. However, the CME charges significant amounts for data use & distribution. You can see the fee schedule here: https://www.cmegroup.com/market-data/distributor/files/mdla-cme-schedule-5-apr-2020.pdf

For example, real-time data distribution, the first heading, for each CME Group Exchange (CME, CBOT, NYMEX, COMEX) is $24k annually, so $96k in total!

If you wanted to play it by the book and know whether your use would be permitted, use the aforementioned CME Data contact form.

Also, regarding the best tips for latency with SC I recommend you check out the 3rd post here and perhaps what follows if you're interested: Backtest results are radically different from live trading results