The question was an unintentional IQ test. You failed, too. Acquaint yourself with the Ignore button if you don't understand what I write.
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
Bars created from broker real-time data feed differ from bars created from eSignal
- Thread starter abattia
- Start date
I would not say that "somebody" intentionally corrects data. The data are just too much, dispersed on different exchanges (for example there already NYSE registered stocks trading not only on NYSE, NYSE being only 25% of their trading volume).
Currently there is no way any single entity to have a "representative" intra-day data.
The limitations to get intra-day data are too many. One of it - not all sources are giving you enough historical data; a lot of bid/ask data is just noise - I posted here on ET about this - it is new consequence of the new market structure and HFT.
The point is to get your data from a place most probably "correct" and close to the environment you want to trade in and focus on "select" stocks only, not all of them. On stocks least likely a subject of manipulation - this leaves you not many, but at least they are more reliable and predictable.
Currently there is no way any single entity to have a "representative" intra-day data.
The limitations to get intra-day data are too many. One of it - not all sources are giving you enough historical data; a lot of bid/ask data is just noise - I posted here on ET about this - it is new consequence of the new market structure and HFT.
The point is to get your data from a place most probably "correct" and close to the environment you want to trade in and focus on "select" stocks only, not all of them. On stocks least likely a subject of manipulation - this leaves you not many, but at least they are more reliable and predictable.
Quote from abattia:
I have noticed that the intraday bars created from real-time data obtained via my brokerâs data feed often differ from those re-created next day from eSignal data.
As a result, a "back simulation" of a given strategy over âhistoricalâ eSignal data will often show different trades from those that actually happened.
As far as possible, I would like to eliminate this difference between the two data sets. What can I do?
Does there exist a set of âabsoluteâ data that is produced (and is available) in real-time?
Is eSignal data âabsoluteâ in this sense?
Or will it always be the case that intraday data re-loaded the next day will differ from the data that was available in real-time?
What broker are you using, what instruments are you talking about, and what type of data are you working with?
For example, I'm pretty certain that IB compresses/aggregates tick data, meaning that volume-bar and tick charts created from IB data will look different from your Esignal charts, as will time & sales. In this case you should trust the unfiltered Esignal data. I don't know about other brokers but realize that brokers are in the brokerage business, not the data provision business. I'm sure the instantaneous quotes are going to be reliable, but when you start getting into charting, historical data, feeds for third-party applications etc. it gets more dubious.
I'm not sure if 'market fragmentation' actually leads to meaningful data differences on stocks and FX, but it seems possible if you're working with very high-frequency data.
Finally, it could just be a temporary problem either with your broker's feed or the Esignal data.
Quote from Specterx:
What broker are you using, what instruments are you talking about, and what type of data are you working with?
Many thanks.
I am using MBT, autotrading stocks (with NinjaTrader), and making use of various intraday timeframes from the 10s of seconds up to 30 minutes.
I trade off the MBT data feed, but backtest/optimize off the eSignal feed. I am pretty sure I could be trading off the eSignal data, and will look further into this ...
Thanks again.
Quote from Arthur Deco:
... Any strategy you develop that is sensitive to such differences won't work anyway...
You could well be correct; I don't know.
However, to take this line of thought further, the differences in datafeeds lead to bars with different OHLCs.
On this basis, any strategy sensitive to the bars' OHLCs would be sensitive to differences in datafeeds, and therefore wouldn't work (according to your statement above).
Therefore, any strategy sensitive to bars' OHLCs wouldn't work?
Quote from Specterx:
I don't know about other brokers but realize that brokers are in the brokerage business, not the data provision business.
Are implying that brokers are manipulating data? Do you have such observation or it is just an educated guess? Or you are referring to something else?