How many lots to trade

R

razf1

I have a backtested system for RUT and would like to trade the Russell emini.


I used a mock trading amount of $1,000,000 and fixed position size of same per transaction.


Here are the statistics for the system since 1995:

Net Profit $2,020,884.96
Profit per Bar $3,402.16

Number of Trades 117
Avg Profit/Loss $17,272.52
Avg Profit/Loss % 1.73%
Avg Bars Held 5.08

Winning Trades 55
Winning % 47.01%
Gross Profit $3,669,349.75
Avg Profit $66,715.45
Avg Profit % 6.67%
Avg Bars Held 8.24
Max Consecutive 7

Losing Trades 62
Losing % 52.99%
Gross Loss ($1,648,464.79)
Avg Loss ($26,588.14)
Avg Loss % -2.66%
Avg Bars Held 2.27
Max Consecutive 5

Max Drawdown ($328,171.63)
Max Drawdown Date 3/7/2005


I want to switch from IWM to the emini.

For every 10k I have how many lots should I buy based on the drawdown identified ( -32% )?
 
Even at one contract per $10K of equity, your leverage is about 6:1. That's already very risky. You should ALWAYS be concerned about how not to lose your capital rather than how fast you're going to profit.
 
Quote from razf1:

I have a backtested system for RUT and would like to trade the Russell emini.


I used a mock trading amount of $1,000,000 and fixed position size of same per transaction.


Losing Trades 62
Losing % 52.99%

Avg Loss % -2.66%

Max Consecutive 5


I want to switch from IWM to the emini.

For every 10k I have how many lots should I buy based on the drawdown identified ( -32% )?
More...
==========
For max % gain /thrills and ;
max account blow up , could trade 5 or more contracts per 10k:D Not recommended however.

However since you mentioned 10k and ''1,000,000mock trading'';
most any professional systems i have seen start with 1 [one ]contract,
So RAZF-one;
it really depends on whether you want to do it as a business or;
want max gain -max chance for account balance of zero or negative:cool:
 
(1) Do you even have $10,000 in your checking account? (2) What are the system results for the past 1, 2, and 3 years? Those results will psychologically "overshadow" what happened during the earlier years of the ten year history. (3) Any system that incorporates the late 1990's is attempting to emphasize a 10-year track record instead of poorer recent results.
 
nazzdack,

I used 10 years data since this is a weekly system. Are you saying going back 5 years is enough data?
 
Has the e-mini Russell even been around for 10 years?

Were you using futures data or cash data?

If so, was the liquidity sufficient to trade then?

Did you test on tick data or EOPeriod?

Have other related markets like ES undergone change in the last 10 years?

Do you think the market conditions then are the same as they are now?

Are the results from that far back really a useful indication of future performance in todays market?

How do they impact the overall results of this backtesting?

Will the market be the same in 5 years time?

Will you change the method, or keep it as it is in 5 years time?

Have you thought of these issues before?

If so, what conclusions did you draw that support the implementation of this system?
 
Streamline,

You bring up some good points. Here are my thoughts:


Quote from StreamlineTrade:

Has the e-mini Russell even been around for 5 years?

Irrelevant to my system. I am backtesting on RUT. Which instrument I use to actually trade is (IWM, ER2, etc. )should not impact my returns drastically. This is a weekly system.

Were you using futures data or cash data?

RUT data from Yahoo. I do not have historical futures data.

If so, was the liquidity sufficient to trade then?

Have other related markets like ES undergone change in the last 5 years?

As markets are constantly evolving, I would expect RUT to undergo changes. However, I am trading basket of indices based on my trend following system. While some might fall out of favor others should continue to outperform.

Do you think the market conditions then are the same as they are now?

I dont look at market conditions. If there is a trend I should benefit. If not, I need to find a better trending index to replace the non-trending ones.

Are the results from that far back really a useful indication of future performance in todays market?

I believe they are since it should show the profitability of the system during different types of markets. Also since this is a weekly system I want to ensure I am backtesting with sufficient amount of data.

Will you change the method, or keep it as it is in 5 years time?

They system will stay put..which indicies are in it is a different story.

Have you thought of these issues before?

If so, what conclusions did you draw that support the implementation of this system?
More...

Again, I am not questioning the validity of my system but rather position sizing. I believe position sizing is more important to me at this point than which EMA I use to maximize profit.

Thanks for your thoughts.
 
I think personal comfort is the important part in position sizing or any risk based decision. Only you know the answer to that one.

Generally I would start off light, and increase with confidence.

I'm no expert though :)
 
You must log in or register to reply here.
Back
Top