How long before I blow up? And other ramblings ...

[Black Swan]

Thanks, I was thinking of getting into that for shorter term trades given that most of my discretional trades are longer term in nature these days.

I guess you have to have quite a lot of programming competence to give this a go

I'd love to be in a position to automate my trading method, it's my 2010 project. IMHO it's where so many EA buyers go wrong, they want an auto bot to make money immediately, as opposed to wanting their own proven strat. automated...

 

[TraderCat]

Thanks, I was thinking of getting into that for shorter term trades given that most of my discretional trades are longer term in nature these days.

I guess you have to have quite a lot of programming competence to give this a go

If you wanted to incorporate the latest Complex Event Processing, AI based, Ultra Low Latency, High Frequency Superbots, with lots of technical doo-dads and other such stuff out of the reach of the average retail trader, then yes, programming competence comes in handy. Otherwise, not so much. Good oddsmaking skills are more helpful.

A good poker player probably has better odds of trading successfully than a good programmer since it seems risk management is the hardest part of all this stuff to get right. The rest is just mechanics.

imho.

 

[Black Swan]

If you wanted to incorporate the latest Complex Event Processing, AI based, Ultra Low Latency, High Frequency Superbots, with lots of technical doo-dads and other such stuff out of the reach of the average retail trader, then yes, programming competence comes in handy. Otherwise, not so much. Good oddsmaking skills are more helpful.

A good poker player probably has better odds of trading successfully than a good programmer since it seems risk management is the hardest part of all this stuff to get right. The rest is just mechanics.

imho.

Good call, keep posting bud, good to have an EA based journal...

 

[TraderCat]

2010.01.01 - 2010.01.08: Sloppy start

Posting summary results for the week as weekly stats give better picture of overall performance.

All in all, not a good start. Took on too much position risk (i.e. size) to achieve results.

As usual for me, discretionary GC (gold) and CL (oil) trades were a waste of capital.

 

[TraderCat]

2010.01.10 - 2010.01.16: Risky Business

Still taking too much risk as evidenced by the equity swings. Time to reign in the bots a little. More work on risk management needed ...

 

[TraderCat]

2009.12.13 - 2010.01.15: Performance Results to Date

The good, the bad and the ugly ...

 

[TraderCat]

Performance Results By Instrument

Breaking down the results by instrument is an eye opener. Clearly the fully discretionary trades (currencies, oil, gold) are making a negative contribution to the P&L.

 

[DionysusToast]

Cool thread...

I didn't see in your first post - can you tell me what your starting capital was ?

 

[TraderCat]

I didn't see in your first post - can you tell me what your starting capital was ?

A little over $500K USD nominal buying power.

Periodically swept.

 

[TraderCat]

2010.01.24 - 2010.01.30: Recalibrating

Flat now.

Too much overall risk. Again.

Volatility really picked up and I found myself behind the curve all week, trying to recalibrate the bots on the fly while manually putting on trades to attempt to hedge the equity swings.

As one can see from the results, It wasn't always successful.

Note the significant drawdowns required to gain the returns.

Stayed out of market on Friday session to reassess.

 

[TraderCat]

2009.12.13 - 2010.03.13: 90 Day Performance Results

Loosened up risk constraints to take advantage of increased market volality.

Decided not to sweep account to provide larger capital cushion for correspondingly large draw-downs (real & imagined).

Capital allocation moving towards Kelly criterion (i.e. maximize geometric return, Risk <= Quantified edge per trade series) as edge becomes quantifiable.

Edge defined here simply as (win% * $AvgWin/$AvgLoss).

This increases volatility of returns but has potential for significant advantage (over the long run) when compared to fixed ratio capital allocation or other alternate strategies due to the potential for exponential capital growth (with the caveat that exponential growth is limited for any finite resource).

This is grossly oversimplifying, but one can research the works of Ed Thorp et al. for further details on this type of capital allocation approach.

 

[TraderCat]

Trading Odds Spreadsheet

Without further polish or explanation, here is an Excel spreadsheet for playing with some basic odds for various win % and risk/reward ratios and getting a sense of what the resultant equity curves may look like given those odds.