Charts Patterns are not correct

Discussion in 'Technical Analysis' started by ligerny, Mar 7, 2011.

  1. ligerny


    I've been looking on various platforms(Esignal, Thomson, thinkorswim, openecry) and i've noticed while looking at futures charts that the candlestick formations look different with each software. I've missed good oppurtunities because on one charting software it doesn't show a reversal bar, on another software it does.

    Why is that? Is the market data bad? On what software can i get an accurate chart/data to do analysis on?
  2. data, time stamping, your ISP, you name it.
  3. dave4532


    Welcome to the club "Illusion". Close the door when you leave please.
  4. ligerny


    Well illusions are costly :)

    But anyone know why there is such a difference, i mean i can understand data delays, but the feed is coming from the exchange, why should the formations of the candlestick be different every where...
  5. wrbtrader


    Yeah...this is a common problem involving Japanese Candlestick patterns. A few ticks different in a particular interval can easily change the entire pattern when comparing one data vendor to another data vendor.

    You can't fix it but you can use two different data vendors to cross check each other. Thus, you'll only take trades when both data providers have the same candlestick pattern or very similar. Another solution is verify the pattern in two highly correlated trading instruments. For example, if you get a bullish white hammer pattern in Eurex DAX futures you want to see the same in Eurex DJStoxx futures.

    The last option is expensive...that is you use only the higher end data vendors (bloomberg, CQG et cetera) because their data is supposedly the most accurate.

    Why the Difference?

    All data vendors aren't getting their data from the same source. Simply, the more middle man that exist...more chances of lost price ticks by the time it reaches you. Throw in your own lost data packets and you and your buddy next to you will easily see different patterns for the same trading instrument if using different data vendors, different ISPs and different software.

    With that said, if you understand the price action prior to the appearance of your candlestick pattern won't matter if your data vendor is accurate or not when you see a white hammer line for one vendor and a doji line for another data vendor for the exact same interval...another reason why many traders have different back test results for candlestick patterns.

    Yet, data vendors are a lot better today than they were 15 years ago because in the old days they would update or fix there bad quotes after the close. Thus, for example, that bearish engulfing pattern you saw at 12noon that failed and you lost money when prices went higher...

    After the close there would be no bearish engulfing pattern but it's been changed to consecutive doji lines. :mad:

  6. Depending on the software you use, the sausage may be cut by your own computer clock. If that is indeed the case, you may wish to look for an atomic clock synchronizer. Freeware, of course.
  7. ligerny


    Thank you that does make alot of sense, sadly i would think paying $200 for a data package would be premium enough for accurate data, and i guess asking people which software/data is more accurate would be like asking them to point out the best looking girl
  8. BSAM


    This is something akin to taking your taxes to ten different CPAs. You'll get ten different results.

    $200 for data? If you can't make money using IB data (I don't know if that would happen to be your broker), then you can't trade anyway.
  9. ligerny


    I was using that figure as an example, traditionally i've used elliot waves in combination to catch reversal's, but i would like to be able to trade shorter time frames just on candlesticks, unfortunatly the data discrepency can be quite costly..
  10. wrbtrader


    If it's too costly or occurring frequently...

    Time to scale up to a higher tier data provider especially if Japanese Candlestick patterns are an important aspect of your trading plan.

    Thus, the cost of paying for good data may be much cheaper than the cost of those losses due to bad data.

    #10     Mar 7, 2011