Does anybody know where I can find the average transaction costs (in percentage) for the following assets in Oanda? I have assumed it to be 2e-4 when training my model with Close prices. Yet, it seems to be much higher than that.

assets = ['AUDUSD', 'EURUSD', 'GBPUSD', 'NZDUSD', 'USDCAD', 'USDCHF', 'USDJPY', 'USDNOK', 'USDSEK', 'USDSGD']

Furthermore, do you know which one is better regarding transaction costs for these assets costs? FXCM or Oanda? I am currently placing limit orders as follows based on the standard deviation of log returns. Does anybody have an advise on how this can be improved?

        stdarr = np.exp(cc.abs().quantile(0.25).values)
        self.Debug(stdarr)
        for i, asset in enumerate(assets):
            ratio = weights[i] if asset[-3:] == 'USD' else (-1.0*weights[i])
            #self.SetHoldings(asset, ratio)
            self.Transactions.CancelOpenOrders(asset)
            oq = self.CalculateOrderQuantity(asset, ratio)
            lp = self.Securities[asset].AskPrice * stdarr[i]
            if oq > 0: lp = self.Securities[asset].BidPrice / stdarr[i]
            if oq != 0: self.LimitOrder(asset, oq, lp)