Real Money Automated Trading Journal

[frostengine]

Starting tonight, I will be recording every trade and the slippage incurred during that trade. This will be helpful when training strategies in the future.

I am also considering taking advantage of MBTrading pay per limit feature. Using limit orders instead of market orders will decrease my slippage and my commission. I need to do more testing to determine if a limit order makes sense for my strategies or not.

 

[gmst]

Quote from frostengine:


I am also considering taking advantage of MBTrading pay per limit feature. Using limit orders instead of market orders will decrease my slippage and my commission. I need to do more testing to determine if a limit order makes sense for my strategies or not.

I will be highly interested if you can share your findings with detailed stats. It will really improve quality of journal from readers point of view. Cheers.

 

[frostengine]

This week I have added USD/JPY to the mix. I will also be recording the slippage on every trade. Slippage will be determined by saving the current price at the time of placing the order, and then comparing it to the actual fill price.

 

[frostengine]

Today's Results:
P/L: $-2.62
Pips: -26
Account: 1012.49

Mostly a quite day. A glitch in my reporting code caused me to only get slippage on 3 trades. Lost 2 ticks on a GBP trade, and lost 2 ticks total on 2 EUR trades. This slippage was only on the entry side, as I am not recording the slippage on the exit.

I made several changes today:
- The 10 min strategies will all continue to use market orders and record slippage
- The 4 range strategies will attempt to perform limit orders tomorrow. I am curious to see what my fill rate will be as well as testing the pay for limits program

 

[frostengine]

Since starting with limit orders on my 4 range strategy, I have been filled on 13 trades and missed on 6. I am going to be looking for ways at improving this by dynamically moving my limit order up as the best bid/ask changes.

Also, I noticed I am running NinjaTrader as full pips when MB trading supports tenth pips. I believe making this adjustment will help tighten my limit order and hopefully improve my fill rate.

While I am still trading live, the actual account balance is not important to me at the moment as I am introducing a lot of new code for improving order executions etc.

Making the switch from full pip to tenth pip has shaved about 1 full pip of profitability off of my strategies. I wonder if the strategies were benefiting from the "rounding" to a full pip?

 

[frostengine]

Last week was a bit like a roller coaster ride:
1. Learned market orders in forex can easily have 3 or 4 pips of slippage
2. Learned changing NinjaTrader to 1/10th pip versus full pip can have a bit of negative impact initially on the strategy
3. Learned limit orders results in about 30% of my trades not getting filled.

Using the info learned last week, my order handling system has grown by several hundred lines of code. This week I will attempt to dynamically adjust limit orders to increase overall fill rate.

Both account 1 and 2 will utilize limit orders going forward. Current Account value: $1003.48

 

[Wide Tailz]

Quote from frostengine:


Using the info learned last week, my order handling system has grown by several hundred lines of code.

:eek:

And here I thought swing trading was the holy grail.

FX day trading seems like a pretty competitive weight class.

 

[clearinghouse]

Quote from Wide Tailz:

:eek:

And here I thought swing trading was the holy grail.

FX day trading seems like a pretty competitive weight class.

Order control, imo, is the among the most important parts of automated trading. A lot of guys think they can just use market orders and get a roughly decent execution and walk from there to riches.

The best example I can think of of trader laziness is when people have to leg into positions on multiple assets and they figure that after leg 1 comes into the picture that they can just always market into the second leg. Well, what happens when your spread is widened because your market just went illiquid in one of your legs?

And that's just one example. The part that sucks about order control is that it's the worst part to test and debug, and sometimes conditions pop up that you didn't know were possible with regards to whatever API you're using.

 

[frostengine]

<b>Today's Ending Balance:</b> $1007.95
<b>P/L:</b> +$4.47
<b>Pips:</b> +45
<b>Trades:</b> 9
<b> Commission:</b>-$.12 (1.2 pips)
<b>Skipped Trades:</b> 3
<b>Entry Slippage:</b> -11.4 pips
<b>Cost Per Trade:</b>1.4 pips
<b>Net PL Per Trade:</b>5 pips

Even though I am using limit orders, I am still experiencing slippage due to the dynamic price adjustment. Its also worth noting that both my exit and entries are limit orders. Some of them I am not getting the "commission" benefit on. I need to look closer at what is happening there.

 

[j2ee]

Quote from frostengine:

<b>Today's Ending Balance:</b> $1007.95
<b>P/L:</b> +$4.47
<b>Pips:</b> +45
<b>Trades:</b> 9
<b> Commission:</b>-$.12 (1.2 pips)
<b>Skipped Trades:</b> 3
<b>Entry Slippage:</b> -11.4 pips
<b>Cost Per Trade:</b>1.4 pips
<b>Net PL Per Trade:</b>5 pips

Even though I am using limit orders, I am still experiencing slippage due to the dynamic price adjustment. Its also worth noting that both my exit and entries are limit orders. Some of them I am not getting the "commission" benefit on. I need to look closer at what is happening there.

I have read all of your words in this post, feel like it is your broker house who gains the advantage, not yourself.