The first night I had no entries trigger (which was confirmed as correct with a backtest) and the second night I had 3 entries trigger.

One of the entries worked correctly and the position was closed at the close of the market. The other 2 entries closed immediately after opening. I’m not sure why this happened as everything appeared to be entered correctly. Will enter on my live account tonight anyway to see if it was just a problem on Sim.

I also spoke with a Tradestation programmer this morning who is confident he can program me something that will enter orders from the Amibroker output (rather than me manually entering every order) and get around the Tradestation limitations on order numbers (as each buy limit order reduces available capital). Similar to the Chartist IB API in a way I expect. Fingers crossed anyway.

Well 30 positions triggered last night however I have to own up to a very embarrassing mistake.

When I re-created the the order template from sim account to my live account, I forgot to change the order that closes the position to a sell!

2 seconds before the market closes a sell order is meant to be sent to close my positions. The default when you create an order is buy and I forgot to change this to sell so two seconds before the market closed all my open positions wanted to buy more!

Now I have been sitting here in the after hours market closing out positions where I can to at least get my margin under the overnight margin requirements (which I have). 15 positions to go with 15 more orders waiting to be picked up.

Lesson learned, double check your orders closely!

Got spanked by the market last night but all in all I am pretty happy as (after my order entry f up on Friday) everything worked as it should. All orders closed just prior to the market close.

Spent a lot of my long weekend combining daily system results in a spreadsheet and noticing that a good touch up from the market every now and then is not uncommon so I am not bothered in the slightest with the results from one day.

Another week and the MOC system ran pretty much as expected with the following observations:

– As previously noted, there is some slippage compared to the Market on Close order prices in the backtest and the price I can get 2 seconds prior to market close. Sometimes it is in my favour but overall against me. Nothing I can really do about this until my API type solution is ready.

– I have had one instance where a position was entered live where the Norgate Data in AmiBroker shows that price did not reach that level, so it wasn’t in the backtest results.

– I had one instance where there was a price spike down at the open. My order was at the low and didn’t get filled but it did in the back test.

The other drama I had this week was a computer glitch that took me hours to resolve. My computer just kept scrolling down or sideways whenever I opened something. Finally got it resolved, had entered 35 out of 40 orders and ran out of available margin. Realised that in my haste I had forgot to adjust my capital in the exploration and the quantities were all wrong! Finally adjusted all the orders and entered the remaining 5 orders 20 mins before the market opened. After all that the market went straight up from the open and nothing triggered for the session.

After many hours over the last couple of weeks I finished my spreadsheet combining the two systems over the last 17 years to get an idea of the combined gains, losses, drawdowns, number of orders etc on a daily, monthly and yearly basis.

It is not often that the system is fully invested. It is a mean reversion system that enters on a buy limit. It has me thinking is there an opportunity to develop a momentum type moc system that takes advantage of days where price just runs from the open and there are very few entries for the mean reversion system. Something that I will look at.

The other thing that I am thinking about is when to put the rest of my capital into this system. I probably have about 65% in at the moment. I have done a fair bit of testing and run it for 3 weeks now. I feel comfortable with the system but I know that I can be optimistic about things and I’m trying to take a balanced approach to ensuring everything is working as it should and not just getting excited about trading if that makes sense.

Hi Chris, it is still being programmed at the moment but it is expected to cost a few thousand dollars. The features so far are those above.

Happy to discuss it more with you when it is finished if you like and put you in touch with the programmer.