Hi guys,

I've read about how consolidation should work but seems I can't get my head around it and seem to find stuff that does not match any understanding I have with it.

I've tried consolidating Forex tick data in 30 minute bars and then tried to make sense of the logic behind it.

Here goes:

1. first algo, just outputs the bid/ask values:

    def Initialize(self):
        
        self.SetBrokerageModel(BrokerageName.OandaBrokerage)
        
        self.SetStartDate(2018, 10, 1)  # Set Start Date
        self.SetEndDate(2018, 10, 1)  # Set Start Date
        self.SetCash(1000)  # Set Strategy Cash
        self.AddForex("EURUSD", Resolution.Tick, Market.Oanda)
       

    def OnData(self, data):
        self.Log("Bid " + str(self.Securities["EURUSD"].BidPrice))
        self.Log("ASK " + str(self.Securities["EURUSD"].AskPrice))

 

2. second algorithm uses a consolidator to generate 30 minute bars:

    def Initialize(self):
        self.SetStartDate(2018, 10, 1)  # Set Start Date
        self.SetEndDate(2018, 10, 1)  # Set Start Date
        self.SetCash(100000)  # Set Strategy Cash
        self.AddForex("EURUSD", Resolution.Tick)
        consolidator = TickQuoteBarConsolidator(timedelta(minutes=30))
        consolidator.DataConsolidated += self.OnDataConsolidated
        self.SubscriptionManager.AddConsolidator("EURUSD", consolidator)


    def OnDataConsolidated(self, sender, bar):
        self.Debug(str(self.Time) + " > New Bar! " + str(bar.Open))

 

When I look at the first line in the logfile for the consolidator algo, I have:

2018-10-0100:30:00 2018-10-01 00:30:00.173000 > New Bar! 1.159645

If I look at the tick bid/ask values for the begining of that minute, I have:

01/10/201800:29:57Bid1.1593701/10/201800:29:57ASK1.1595201/10/201800:30:02Bid1.1593701/10/201800:30:02ASK1.1595101/10/201800:30:02Bid1.1593601/10/201800:30:02ASK1.1595001/10/201800:30:03Bid1.1593701/10/201800:30:03ASK1.15951 

I don't understand where the bar.Open value (1.159645) comes from, looking at the Bid/Ask values for around that time.

When backtesting an algo using minute resolution, I noticed that the price the market orders were filled was usually the last bid/ask of the previous minute so when I was looking at open/close data to generate a trendline (for ex), the market orders were pretty far away from the open/close values.

Hope my question makes sense.

Dan

Author