huge spike in Dow

Forums › ProRealTime English forum › ProRealTime platform support › huge spike in Dow

Viewing 15 posts - 16 through 30 (of 37 total)
1 2 3
  • 08/25/2018 at 10:32 AM #79038
    robertogozzi

    It happened yesterday night on DAX at closing time.

    The first pic, 5 minutes, is from my Real account, the second one from my Demo account.

    My real account reports a closing price of 12405.0, while my demo one reports 12551.9!

    1 user thanked author for this post.
    avatar GraHal
    08/27/2018 at 9:42 PM #79201
    GraHal

    Spike happened again didn’t it at 09:15??

    I can’t see a spike on my PC Platform but I can see a triggered limit value that is 80 points higher than levels I can see on my Chart!

    08/27/2018 at 9:43 PM #79202
    verygrubby

    Yep, 250 points again

    08/27/2018 at 11:16 PM #79207
    verygrubby

    ..by the way, there is also another spike daily at 22:00 too. The second spike usually sorts itself out at 23:00.

     

    Amazed this hasn’t been fixed yet. How hard is it to just duplicate the live price feed for the demo feed?

     

    Good luck to all those with demo strategies holding open positions after 9pm!

    08/28/2018 at 12:04 PM #79242
    GraHal

    I just did a Live Chat with IG and asked when the spike  would be sorted and got the answer below. Seems a bit weird as I thought it was Live data on a Forward Test Platform??

    Hi, this was resolved just now for IG platform so please wait until tomorrow as data is sent to PRT each night

    1 user thanked author for this post.
    avatar robertogozzi
    08/28/2018 at 4:09 PM #79252
    verygrubby

    Perhaps they mean the historical spikes will be cleaned up tonight….and the live prices should be fine from now on…

    1 user thanked author for this post.
    avatar GraHal
    08/28/2018 at 9:19 PM #79264
    GraHal

    Oh well it happened again, maybe tomorrow night we may be rid of this stupid spike!?

    The IG guy maybe meant software changes are uploaded overnight and so the fix will be an overnight upload … let’s flippin hope so!!!!!   It all seems such nonsense, it’s like a bad dream!!!!

     

    08/28/2018 at 9:39 PM #79265
    Vonasi
    GraHal wrote:

    it’s like a bad dream!!!!

    Not a bad dream just a bad lot of demo data. How they expect people to decide to trust IG with real money if this is the demo data is beyond me. We are all interested in auto trading and data analysis via indicators and screeners which we base our decisions to trade on and this is the data they give us and then expect us to deposit and risk large sums of money. Beggars belief – let’s hope this is fixed and not replaced soon by another demo bug that is low priority to IG as it is apparently only demo and not important.

    1 user thanked author for this post.
    avatar GraHal
    08/29/2018 at 9:20 AM #79280
    Leo

    I wonder why demo and real data are differences…

    Does IG make his own test how to manipulate data?

    …yes, I do believe in that conspiracy theory that the price move to take my stop loss and then after it triggers it moves contrary or the price move to my contrary and move opposite if I remove the stops loss. Or many many times that my stop profit is not trigger by half a pip but my trailing stop does

     

    08/29/2018 at 9:34 AM #79283
    David

    I don’t know about demo but live prices are based or are supposed to match Bloomberg pricing.

    In FX I find data gaps form time to time which as far as I understand is a data feed issue from IG to PRT – of course that’s if you’re an IG client.

    08/29/2018 at 9:47 AM #79285
    GraHal
    Leo wrote:

    Does IG make his own test how to manipulate data?

    Same thought crossed my mind … why else are Demo and Live not exactly the same price levels ??  Anybody know if it states anywhere on IG website that Demo and Live are different prices??

    If different (and clearly they are different!) it makes a mockery of all our Forward Testing??

    This latest issue has made me even more wary. We deserve an explanation from IG, but doubt we ever will get one?

    Only safeguard I can think of is to not have any System trade (entry or exit) between 20:55 to 23:05??

    08/29/2018 at 10:04 AM #79286
    David

    As part of my testing I compare live results versus the backtest.

    Eg. When the system is running, I run the backtest to see if the entries and exit are matching up.

    Spikes are spikes but if they erroneous then that’s an issue that needs to be addressed. One of the good features with the IG platform is that you can look at the bid/mid/ask price. With PRT it’s only the Mid.

    RE: 20:55 to 23:05 I think this is around the Rollover period for many instruments. The spreads are nasty – avoid trading during the rollover period.

    I know traders that complain about the differences in spikes from different brokers but this is the reality in trading the stream of liquidity, speed of systems and internet, amount of positions on the brokers books differs so there’s always a difference in pricing between brokers.

    Moreover, this is why too much optimisation leads to failure – loose pants might not be a perfect fit, but they will fit many.

    08/29/2018 at 10:13 AM #79287
    GraHal
    David wrote:

    When the system is running, I run the backtest to see if the entries and exit are matching up.

    If you had to put a % difference on results of above exercise what would you say please?

    The spikes we are annoyed about on this Topic  is the massive 200 point + spike at 09:15 and 23:00 on Demo only (not on Live Platform) … surely this is not acceptable every night for 6 nights due to rollover period etc etc (and anyway same spikes have not showed on Live so they must be due some weird Demo Data Issue)?

    08/29/2018 at 11:57 AM #79292
    verygrubby

    Two +250 point spikes at 21:15 & 22:00 each day is clearly an error, but the really worrying thing is the fact that this doesn’t appear in PRT charts or backtesting results. It’s as if PRT demo uses IG’s LIVE price feed for charting, but the IG DEMO price feed for the actual strategy execution. Which, as everyone has said, makes backtesting/demo walk forward impossible to use with any confidence. I mean, how many other (smaller) discrepancies have there been that have gone unnoticed, but have caused differences between live demo / demo backtest?

    08/29/2018 at 12:28 PM #79294
    GraHal
    verygrubby wrote:

    I mean, how many other (smaller) discrepancies have there been that have gone unnoticed, but have caused differences between live demo / demo backtest?

    Exactly what I’ve been thinking for days!
    Is there any way we can set up some monitoring system between Live and Demo so that it flags up each day’s worth of price level discrepancies??

    verygrubby wrote:

    It’s as if PRT demo uses IG’s LIVE price feed for charting, but the IG DEMO price feed for the actual strategy execution.

    We have to monitor this possibility even if only as we are doing on this Thread (visual check) and report each and every occurrence … however small?

    But you do see spikes on the IG App Grubby so doesn’t that rule out … PRT demo uses IG’s LIVE price feed for charting (as Live had no spikes??) ??

    I’ve never used Demo on my App (only Live) I’ll have to get it sorted.

     

  • Author
    Posts
Viewing 15 posts - 16 through 30 (of 37 total)
1 2 3

Find exclusive trading pro-tools on ProRealCode Marketplace

Similar topics:
stratégie dow m1 midnight candle

Create your free account now and post your request to benefit from the help of the community
Register or Login