Scott’s Journal

[ScottMcNab]

Julian Cohen wrote:

Nick Radge wrote:

Quote:

Isn’t min comm with IB US $1 ?

If that’s the paper trade account it’s probably set on the wholesale commission rate.

Quote:

My understanding is that if the order if completely filled then the MOC order is sent immediately. The time setting relates to the time for which orders that only received partial fills remain …after the time (that the user sets in the preferences), then the remaining partially filled orders are cancelled and MOC orders are then sent for the partial fills

So ask the developer to look at my suggestion. Won’t work otherwise.

Intra day data is not really an option – very expensive and coding becomes significantly more complex.

Scott the developer has sent me exactly the option as Nick suggested today so it is coded and ready to test, which I will be doing in the next few days

Be interested to hear how it goes….cost of the API’s starting to add up so I better make some real money before I look to any more

[TrentRothall]

Scott McNab wrote:

Trent Rothall wrote:

Hi Scott, how is slippage looking with the MOC orders?

It seems to vary quite a bit from AB but I suspect that is still due to it being a paper acct. I hope to finally begin some live trading soon once I have paper tested this one for a few weeks.

Yeah ok, will be interesting to see how it compares.

Thanks

[TrentRothall]

The US might be different to the ASX but looking in time and sales on IB for some Austrailan stocks in the last few minutes a number of different prices can be traded, so it might not just be a paper trading issue.

[ScottMcNab]

Trent Rothall wrote:

The US might be different to the ASX but looking in time and sales on IB for some Austrailan stocks in the last few minutes a number of different prices can be traded, so it might not just be a paper trading issue.

Cant imagine would be any different at the close to the open or am I missing something Trent ?
Cheers

[ScottMcNab]

System took 10 trades overnight while Amibroker took 11…INTU had buylimit order at 113.17 which I missed…why wont the damn thing miss the losing trades!

[TrentRothall]

Scott McNab wrote:

Cant imagine would be any different at the close to the open or am I missing something Trent ?
Cheers

Yeah you might be right there actually…

A range of small transactions in that window!

[ScottMcNab]

Maurice Petterlin wrote:

You guys are on another planet in comparison to what I’m doing. I hope you make a shed load of money with this strategy. What are the backtester stats like?

The CAR isn’t improved over the MRV systems I have where the exit is close>ref(c,-1) etc but the maxDD is reduced. As Said and Trent and others have mentioned the potential impact of slippage live trading remains to be seen.

[ScottMcNab]

Trent Rothall wrote:

Hi Scott, how is slippage looking with the MOC orders?

I just compiled this quickly from last night’s trades

[SaidBitar]

Scott McNab wrote:

System took 10 trades overnight while Amibroker took 11…INTU had buylimit order at 113.17 which I missed…why wont the damn thing miss the losing trades!

Yes sometimes I have the same feeling as it is doing it to pissed me off

[ScottMcNab]

Looked over last nights closing fills with MOC system again..all seems to even out

[ScottMcNab]

Quick question please….when I do a backtest with no margin (25 positions at 4%) the max trade % drawdown is -20…..and yet when I use same system on margin (25 positions at 8%) (accountmargin changed to 50) the max trade % drawdown is still only -20…I was expecting it to go to -40….indicating that this figure is calculated without margin ? The other metrics change (CAR doubles and max DD of system doubles) indicating that the backtest is using margin. In anticipating worst trade drawdown on margin should I double this figure ?
Thanks in advance

[SaidBitar]

Maximum Trade DD is calculated based on the entry, Low and Exit. so the position size is not part of the equation.

and actually it is saying -20% in both cases 4% position size and 8% position size but the value in $ is totally different since 20% of 4% is half of 20% of 8%

so no need to double anything

[ScottMcNab]

Missed another 2 trades on Friday…PTEN and STRZA….so out of the total $1600 profit Amibroker backtester shows since started paper trading this system the paper account has missed $600 worth from trades that were part of spread trade/special conditions etc….getting a little edgy but cant imagine how any of the systems using buy limit orders would avoid this issue ? Interestingly, in these cases the price that Amibroker backtest uses to enter the trade is quite low so the phantom trades are quite profitable ones (don’t understand enough about spread trades or other special condition trades to know why this seems to occur).

[Nick Radge]

Quote:

don’t understand enough about spread trades or other special condition trades to know why this seems to occur

These trades are ‘off market’ trades meaning they’re a part of a basket or combo trade (options).

in Australia these aren’t actually registered as a trade in the data so they’re not seen. Not the case in the US and I agree that the exchanges could handle them somewhat better.

Next 1000 trades…

[ScottMcNab]

Thanks Nick….system still profitable on ASX so feel a lot better (only change on asx is reduction of volume filter from 1mill to half a mill)

[Author]

Posts