Real Money Automated Trading Journal

gmst · Jul 21, 2012

Quote from frostengine:

I have been developing automated trading systems since about 2003. I have learned a lot and made a lot of mistakes over the years. I have produced numerous "profitable" systems, however due to the large draw downs I do not trade any of them live.

Hey Frost,

Nice journal you have here!

So, you say you have been developing trading systems from 2003 onwards thats like almost 10 years. But are you saying you still after 10 years, dont have any profitable system that is tradeable due to all systems having large DDs? If this is the case, how do you trade currently? Discretionarily? Are you a profitable full time trader or is trading a second income for you? Want to know more about yourself.

In any case, looking forward towards a new journal.

frostengine · Jul 21, 2012

Quote from robbob:

Hi Frost,

What are some of the technologies/languages that you are using for automating your strategies and backtesting? Also I'm curious to know who the brokerage your using.

I've been using the Java API at Interactive Brokers for a while now, but am interested in hearing what others on the forum are using.

Thanks,
-Rob

For this round, I am primarily using NinjaTrader for the live trading and backtesting, combined with a C# application I wrote to perform the automated data mining.

I have been with Interactive Brokers for essentially the past 10 years, but recently opened this $500 account at MB Trading for the purposes of this strategy.

frostengine · Jul 21, 2012

Quote from gmst:

Hey Frost,

Nice journal you have here!

So, you say you have been developing trading systems from 2003 onwards thats like almost 10 years. But are you saying you still after 10 years, dont have any profitable system that is tradeable due to all systems having large DDs? If this is the case, how do you trade currently? Discretionarily? Are you a profitable full time trader or is trading a second income for you? Want to know more about yourself.

In any case, looking forward towards a new journal.

I do not do much day trading these days. I do some longer term swing/discretionary/semi-automated strategies. However, trading is a second income. I am an engineer at a technology company.

I believe the primary piece that has been missing from my strategy development efforts has been a way to go "live" in a low risk environment. Now that I have that, I believe this journal will be a success.

robbob · Jul 21, 2012

Quote from frostengine:

For this round, I am primarily using NinjaTrader for the live trading and backtesting, combined with a C# application I wrote to perform the automated data mining.

I have been with Interactive Brokers for essentially the past 10 years, but recently opened this $500 account at MB Trading for the purposes of this strategy.

Just curious why a new account MB Trading for the strategy?

frostengine · Jul 21, 2012

Quote from robbob:

Just curious why a new account MB Trading for the strategy?

MB Trading allows me to trade 1k lots without a minimum commission. While at IB I had to pay a minimum commission of $2.5 which made it difficult to test run a live strategy like this.

frostengine · Jul 22, 2012

Results
Week 1 results: 164 pips
Account value: $516.40

Strategy Changes
Tonight I will continue with the v1 strategy used last week. The current strategy utilizes range 4 bars. For version 2, I have been searching for patterns on 10 minute bars.

Initially I was planning to integrate the V2 signals into the existing strategy, however I am now thinking of potentially opening a second MB Trading account and allow both V1 and V2 strategies to run independently of one another. I will need to think about this a bit more. I may make a switch towards the middle of the week, to collect live data on how V2 performs before making a decision.

Other Observations
Both V1 and V2 strategies at this point has been developed exclusively against the EURUSD. However, I am running against 4 currency pairs. Looking at historical results, as well as the live trades taken last week, it is apparent this approach is not optimal.

One of the next steps will be to search for patterns in each currency pair and "optimize" both the V1 and V2 strategies for each pair. This should be a substantial improvement over the current strategies. The NZDUSD seems to perform the worse for V1, essentially no better than break even after commission/slippage. As such, I am considering turning NZDUSD off until an optimized strategy is ready.

frostengine · Jul 22, 2012

Deciding whether or not to stay with V1 or move to V2 is proving very difficult. The same reason I MUST trade live to evaluate a strategy, is what is preventing me from adapting quickly.

For example, I have data from 4 total sources, 2 different static sources, MB Trading loading tick data from Ninjatrader, and MB Trading downloading tick data then importing.

With all 4 sources I get VASTLY different backtesting results. These inconsistencies is why I term backtesting as just "interesting".

A good way to illustrate this is as follows:
1. Live account for Thurs and Friday +$6.16 on EURO
2. Backesting loading data THROUGH MBTrading -.10 on EURO
3. Market Replay, +5.50 on EURO

As you can see the backtesting results are not even REMOTELY close to what actually occured in both replay and live. This further proves why you can not trust backtesting.

gmst · Jul 22, 2012

Quote from frostengine:

As you can see the backtesting results are not even REMOTELY close to what actually occured in both replay and live. This further proves why you can not trust backtesting.

No, this just proves that the data on which backtesting is being done is completely screwed up.

I think this experience should be enough to clearly demonstrate that MB Trading is a bad choice. Move to OandA - use their data for backtesting - and your backtested numbers will bear resembalance to actual executions. Or, move to IB - but then you will pay full commissions even if you are trading micro lots. FX Futures are other option - you will get good data in all 3 these cases. You decide - but obviously MB Trading is worthless for developing systematic intraday trading algos - unless you are doing algos on dailys.

Finally, I think OandA is what you are looking for.

frostengine · Jul 22, 2012

The EC2 instance is back online and ready for the week to begin. I have configured EURUSD,AUDUSD,and GBPUSD to run against version 1 strategy. While NZDUSD is configured to run against version 2 strategy.

Both strategies are still optimized only for EURUSD, but I want all 4 pairs running so that I can continue to gather data such as slippage etc.

frostengine · Jul 22, 2012

Quote from gmst:

No, this just proves that the data on which backtesting is being done is completely screwed up.

I think this experience should be enough to clearly demonstrate that MB Trading is a bad choice. Move to OandA - use their data for backtesting - and your backtested numbers will bear resembalance to actual executions. Or, move to IB - but then you will pay full commissions even if you are trading micro lots. FX Futures are other option - you will get good data in all 3 these cases. You decide - but obviously MB Trading is worthless for developing systematic intraday trading algos - unless you are doing algos on dailys.

Finally, I think OandA is what you are looking for.

I do not believe OandA works with NinjaTrader? I also believe, there are fundamental problems with Backtesting. It can help give you an"idea" how a strategy will perform, but is not much more than that.