Menu
Sister Sites

Options

Bookmark Thread

First Page First Unread Last Page Last Post

Print Thread

Similar Threads

Importance Of Daily Closing Price In Forex 3 replies

Importance of trading mentality 6 replies

The importance of taking profits through shorter term trading 25 replies

importance of fundamentals/technicals in intraday trading 0 replies

The Importance of Trading Time Frame Assessment 0 replies

Attachments: Importance Forex Trading Performance Stats
Exit Attachments

Importance Forex Trading Performance Stats

  • Post #1
  • Quote
  • First Post: Nov 1, 2016 2:31am Nov 1, 2016 2:31am
Hi guys,

What is the most important performance stats for Forex Trading? Especially for automated or algorithmic trading.
I have googled but results are not for forex, mostly for stock.

Below are my stats: (PPT is Profit per Trade, Profit in pips)
What i'm missing?

Attached Image (click to enlarge)
Click to Enlarge Name: Screenshot1.png Size: 57 KB


Thank you
When you lose, I'm the loudest who laugh at your face )
Quoting Zelo
Hi guys, What is the most important performance stats for Forex Trading? Especially for automated or algorithmic trading. I have googled but results are not for forex, mostly for stock. Below are my stats: (PPT is Profit per Trade, Profit in pips) What i'm missing? {image} Thank you

1. Sample size
2. Out-of-sample test
No Highs? No Lows? Must be Bose!
Quoting Tardigrade
{quote} 1. Sample size 2. Out-of-sample test
1. Sample size is more than 60,000 bars on M5 from February to October.
2. Out-of-sample test is a good point, i will test it on 2015 and 2014.

And what else? More please!
When you lose, I'm the loudest who laugh at your face )
I have added number of winning/losing streaks. Maybe average running time of trade next? And StdDev?
When you lose, I'm the loudest who laugh at your face )
I have implemented Avg duration of trades. Maybe next i will separate long and short to have another view of my model.
When you lose, I'm the loudest who laugh at your face )
Quoting Zelo
{quote} 1. Sample size is more than 60,000 bars on M5 from February to October. 2. Out-of-sample test is a good point, i will test it on 2015 and 2014. And what else? More please!
An alien just landed near a basketball court. A father was teaching a 5-year-old little girl to shoot some hoops. Magnificently, this little girl scored 5 in a row from the baseline; without prior knowledge of basketball, the alien concluded that this little creature must be very skillful at what she does.

Question: Was the little girl a future basketball star, or was she just got lucky?

Trade count is 100ish, not enough sample to determine the robustness of those systems.

About relevant statistic to system testing, I recommend Kaufman's "A Guide to Creating a Successful Algorithmic Trading Strategy". The book is not very organized but Kaufman does provide some insights into the integral process of trading system development.

Cheers.
No Highs? No Lows? Must be Bose!
Quoting Tardigrade
{quote} Question: Was the little girl a future basketball star, or was she just got lucky?

She just got lucky, of course. But there is a chance that girl learn how to shoot from 2 years old.
What here:
[Little Big Shots - Four-Year-Old Piano Prodigy]
Inserted Video


Quoting Tardigrade
Trade count is 100ish, not enough sample to determine the robustness of those systems.

Yes, i know. But i haven't told you the cool thing yet (and I never will ).

Thank you for your reply. But i am looking for trading stats are for forex. Because forex is different than other markets.
When you lose, I'm the loudest who laugh at your face )
There is a book called "The Evaluation and Optimization of Trading Strategies". It may help someone. You can read it from google.
When you lose, I'm the loudest who laugh at your face )
I also interested in this, got many brach/version on several strategy. My approach now is to split on several account and run several branch that i think better. One of the branch was on my te(third one), the one in my sig on live testing stage. Too many acc, too many version, my head spins.
Quoting Passerby
I also interested in this, got many brach/version on several strategy. My approach now is to split on several account and run several branch that i think better. One of the branch was on my te(third one), the one in my sig on live testing stage. Too many acc, too many version, my head spins.
Why not use only one account run multiple strategies? You can separate strategy by using comment, later analyze each strategy using that comment. It is my current approach now.
When you lose, I'm the loudest who laugh at your face )
Quoting Zelo
{quote} Why not use only one account run multiple strategies? You can separate strategy by using comment, later analyze each strategy using that comment. It is my current approach now.
Thats will be too long on my excel spreadsheet, my ea took 700+ position a week. I also separate pairs and trading hour in my stats. 10 pairs a branch, 700+ trade a week per branch, with only 5 branch it ll be a very long sheet. Anyway i use dedicated server, not vps. For many accounts, dedicated server is much cheaper per acc.

And, if one of the ea fail i wont lost all the money as it was put in several acc. This is real money, as i do test ea on live acc(few grand per acc).
Quoting Passerby
{quote} Thats will be too long on my excel spreadsheet, my ea took 700+ position a week. I also separate pairs and trading hour in my stats. 10 pairs a branch, 700+ trade a week per branch, with only 5 branch it ll be a very long sheet. Anyway i use dedicated server, not vps. For many accounts, dedicated server is much cheaper per acc. And, if one of the ea fail i wont lost all the money as it was put in several acc. This is real money, as i do test ea on live acc(few grand per acc).
damn.thats crazy.would love to see a video of that.
Quoting ericnyamu
{quote} damn.thats crazy.would love to see a video of that.
Huh? How long that ll be? Its a scalper, its looks like this:
Attached Image (click to enlarge)
Click to Enlarge Name: IMG_20161106_214507.jpg Size: 789 KB

Got myself a new server last week. One server could handle 13 meta without missing trades @15-25ms latency. Load 15 on a server once and i start getting missing trade, solved after i move 2 terminals to a new server. I think its NIC limitation, happens even on higher spec server.
Forex is not all about technical analysis. Fundamental analysis is equally important. One of this is the US nonfarm payroll data. It is very important for the inflation target of the US Federal Reserve.
  • Post #15
  • Quote
  • Last Post: Nov 7, 2016 7:52pm Nov 7, 2016 7:52pm
FOUND IT:

https://github.com/QuantConnect/Lean...eStatistics.cs

Hope it will help someone like me in the future.
When you lose, I'm the loudest who laugh at your face )
0 traders viewing now