Day 19: +250$

No fill for ES/TF, it should have been 150$ more with market orders.
I'll change the robot for next week, market orders only (and a SMS sent after each trade, just for fun).
I'm removing the () in my first week p&l, that was a "bug" of day 4, but bugs or human errors can happens, so I should not ignore them.


Week 1: +977.5$
Week 2: +130$
Week 3: +1180$
Week 4: +97.5 $
From the beginning: +2244.5$ before com

Big Mike 

or can also add another time frame in initialize Add(PeriodType.Minute, 30); and then submit the limit order to that time frame so it will last 30 minutes before being canceled or etc. Second you can use limit orders but set it to never cancel. Third you can use market orders. Mike

I use limit orders, but add or subtract an offset to the price, this way if desired, it is practically a market order but with the limit safety.

First I use a variable so it can be optimized..

private int priceoffset = 1; // Default setting for Priceoffset

If buying and wanting price to come to me:
EnterLongLimit(DefaultQuantity, GetCurrentBid() - priceoffset * TickSize, "");
or if aggressive:
EnterLongLimit(DefaultQuantity, GetCurrentAsk() + priceoffset * TickSize, "");


or selling conservative:
EnterShortLimit(DefaultQuantity, GetCurrentAsk() + priceoffset * TickSize, "");
aggressive:
EnterShortLimit(DefaultQuantity, GetCurrentBid() - priceoffset * TickSize, "");

or variations thereof.

dsherman 

I use limit orders, but add or subtract an offset to the price, this way if desired, it is practically a market order but with the limit safety. First I use a variable so it can be optimized.. private int priceoffset = 1; // Default setting for Priceoffset If buying and wanting price to come to me: EnterLongLimit(DefaultQuantity, GetCurrentBid() - priceoffset * TickSize, ""); or if aggressive: EnterLongLimit(DefaultQuantity, GetCurrentAsk() + priceoffset * TickSize, ""); or selling conservative: EnterShortLimit(DefaultQuantity, GetCurrentAsk() + priceoffset * TickSize, ""); aggressive: EnterShortLimit(DefaultQuantity, GetCurrentBid() - priceoffset * TickSize, ""); or variations thereof.

Keep in mind -- for backtesting -- that Market orders will be filled @ the ask instantly, but Limit orders @ the Ask will require price to fire @ the bid before being filled @ the ask if you use the Normal fill type (which you should) instead of Liberal.

Mike

Day 20: -295$


Week 1: +977.5$
Week 2: +130$
Week 3: +1180$
Week 4: +97.5 $ Week 5: -295 $From the beginning: +1949.5$ before com

Day 21: -535$


Week 1: +977.5$
Week 2: +130$
Week 3: +1180$
Week 4: +97.5 $
Week 5: -830 $From the beginning: +1414.5$ before com

Day 22: -825$


Week 1: +977.5$
Week 2: +130$
Week 3: +1180$
Week 4: +97.5 $
Week 5: -1655 $From the beginning: +589.5$ before com

OK, so this week so far has seen some relatively large drawdown. How are you feeling? Are you starting to doubt your strategy or did you have some occasional drawdowns like this allowable in your strategy? At want point do you throw in the towel and decide that the strategy doesn't work?

My BOT had 2 losses today so far right at the open. One small-ish, the other not as small. So I'm not very happy about that. But I have to consider that the probabilities of success are in my favor, until they aren't, so there's always a seed of doubt that I have to listen to or battle with.

I used to think that auto-trading removes the psychological challenges of trading. It does, as far as immediate pre-, intra-, and post-trade psychology goes. But seeing big wins and losses that your auto-strategy makes has its own psychological challenges as well to consider.

shodson 

OK, so this week so far has seen some relatively large drawdown. How are you feeling? Are you starting to doubt your strategy or did you have some occasional drawdowns like this allowable in your strategy? At want point do you throw in the towel and decide that the strategy doesn't work? My BOT had 2 losses today so far right at the open. One small-ish, the other not as small. So I'm not very happy about that. But I have to consider that the probabilities of success are in my favor, until they aren't, so there's always a seed of doubt that I have to listen to or battle with. I used to think that auto-trading removes the psychological challenges of trading. It does, as far as immediate pre-, intra-, and post-trade psychology goes. But seeing big wins and losses that your auto-strategy makes has its own psychological challenges as well to consider.

You can always doubt, but the criteria which will decide if the strat is fine or not are not already met.
Before launching the bot, the failure/success criteria were already known and wrote: at least a run period of 8 weeks, no more than 4 loss in a row, avg win > 1.3 * avg loss, an emergency <stop-strat> button after an overall loss > X $, and few other "rules".
Once these criteria are not met (their are some AND and some OR...), doubt or not, it's too late, I'll have to click on the <start> button.

If auto-trading remove a part of the psychological challlenges, it's a small part: you build the strategy (have I imagined all possible cases ?), you launch it (hmm, was today a good day for this strat ?), you decide to stop it when it looks to bad (and if I let it go 1 or 2 days ?).
So yes, it's easier to launch a strat and then go watch TV, than keep your eyes on you growing losses, but not so easy...

Day 23: +1390$

A small change today, I'm now playing with 2 NQ contracts per broker, instead of 1 before. And always the same other contracts (ES/TF/EMD).
The YM may be back on the list later, I think I've found what it had bad results, compared to the other contracts.

Week 1: +977.5$
Week 2: +130$
Week 3: +1180$
Week 4: +97.5 $
Week 5: -265 $
From the beginning: +1979.5$ before com

Nice!.. glad you had a great day today Sam.