A single quality indicator for the strategy - page 4
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Different types of strategies require different evaluation. For example, if we compare a counter-trend averaging tool and a trend-following Expert Advisor, such indicator as "profitable/lossmaking trades" may be similar, but for averaging tool it will mean the number of profitable trades without taking into account loss-making trades in this position (which does not give any useful information), while for a simple (without shares) trend-following tool this indicator will be very important to estimate the frequency of successful entries (which helps in loss trade filtering).
The same situation is with the analysis of the balance curve - for different types of strategies the good ERR will be different.
The recovery factor should also be used carefully, only after normalization of deals, profit/slippage of which is extreme, the method of such normalization can be different.
So, if I understand correctly, figuratively speaking, you want to describe all the diseases of all the patients in the hospital, with one parameter, such as the average temperature in the hospital and then based on this parameter to prescribe treatment to all?!
Easy. The severity of the disease.
A remarkable evaluation parameter! - One patient with a compound hip fracture, another with acute appendicitis, a third with an acute heart attack, a fourth with late-stage cancer, a fifth with a fractured skull - all are critically ill, but to determine the severity of each patient not only requires sets of highly specialised parameters, but also narrow specialists who can assess the severity.
It is the same with the TC assessment parameters. Of course one can estimate TC by the number 'y' - woo-hoo-hoo what a cool system!!! But in order to accurately assess the degree of its steepness, you must not only know a lot of criteria, but also thoroughly know the principles of its construction.
Actually the author, in my opinion, slightly misunderstands the essence of building a trading strategy. Before building a TS it's necessary to create a base for it, namely to get statistics of price behavior from predictive factors - predictors.
Profit in trading is always obtained from the 'right' difference between the price of opening and closing a position. It means that by the moment the position is opened the forecast that with such and such probability the price will move in the necessary direction should be available. I.e. there is a set of predictors the combination of whose values indicates the probable direction of price movement.
As predictors there may be anything - indicator signals, fractals being formed, signals of classifiers about market state change, neuronet signals about forming shapes, signals from other currency pairs or financial instruments correlating with the one being traded, etc. - anything that might help to predict the price behavior more or less exactly.
Then on the basis of those predictions statistics is compiled as to how much, in what direction and in what percent of cases price moves after signals from predictors. After that all these statistics are checked on a flat, on a trend, on a highly volatile and low volatile flat, on a strong and weak trend, and on the basis of these statistics very creative and individual process of building a trading system starts.
A remarkable evaluation parameter! - One patient with a compound hip fracture, another with acute appendicitis, a third with an acute heart attack, a fourth with advanced cancer, a fifth with a fractured skull - they are all critically ill, only to determine the severity of each of them requires not only a set of highly specialised parameters, but also narrow specialists who can assess the severity.
...
Only read the first paragraph, just sorry, sick of stupid arguments in isolation from reality.
All those listed in the quote have a common denominator - the amount of attention required, the number of personnel involved and the minimal delay in assistance.
I propose to come up with/develop a single coefficient showing the quality of the strategy, which will take into account its many characteristics (profit, drawdown, number of trades,...). In MT5 it is possible to use this.
This kind of task can be solved graphically:
Profit and drawdown are indicated in stop loss. The three values used in the functions shown in the graph do not contain such important information about the strategy as "growth stability", in part only the drawdown.
The result of these functions can simply be multiplied, resulting in a single number, which can be used to judge the overall quality of the strategy.
Different types of strategies require different evaluation. For example, if we compare a counter-trend averaging tool and a trend-following Expert Advisor, such indicator as "profitable/lossmaking trades" may be similar, but for averaging tool it will mean the number of profitable trades without taking into account loss-making trades in this position (which does not give any useful information), while for a simple (without shares) trend-following tool this indicator will be very important to estimate the frequency of successful entries (which helps in loss trade filtering).
The same situation is with the analysis of the balance curve - for different types of strategies the good ERR will be different.
The recovery factor should also be used carefully, only after normalization of deals, profit/slippage of which is extreme, the method of such normalization can be different.
What is the strategy indicator? Is it used to determine profitability or just to see how well the strategy works?
It's a great idea. I don't think anyone can come up with a better one.
I'm beginning to think so too, just looking at the problem from the outside may reveal some features that I might not have seen.
The overall score will also depend on the complexity of the model.
I'll tell you something scary, guys. You've got the concept switched around.
You are confusing OPTIMISATION CRITERIA and system performance evaluation. It's a little simpler than that. The efficiency of a system is the end result of what it is designed for.
In our case, it is maximum profit at minimum risk when working in an unfamiliar environment. And if the conditions of the task require that a small or even zeroamount of transactions is handled correctly, then there is no sense to talk about the risks and only pennies remain.
We simply give to the strategy a single fixed lot and run it through wolfing forward. And then we simply sum up the net profit of each forward.
And when you will buy a yacht later, nobody will tell you - werefuse from your money, because it was received with too small sharpe ratio! The only question on everyone's mind will be HOW much money you have.
My experience convinced me that if an EA makes more profit than others under the same conditions, then it is usually with sharps and drawdowns alright, and the number of deals is not going over the top and all is well with charts.
Although, it would be nice to automatically estimate the nature of the curve, but not of the back, but of the forward one. At least three.
And there is certainly no point in taking some kind of "complexity" into account as it is very difficult to account for such abstract qualitative characteristics quantitatively.