Discussion of article "Evaluation of Trade Systems - the Effectiveness of Entering, Exiting and Trades in General" - page 2
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
There is one more efficiency indicator - Deal efficiency = ProfitPips / MaxProfitPips, where
ProfitPips - profit of the deal.
MaxProfitPips - the maximum possible profit (>= MaxPrice - MinPrice), which could be obtained during the deal's existence interval.
Due to changes in builds that have occurred since the time of writing the article, for correct operation it is necessary to correct the C_PosStat class in the Bulaschev_Statistic.mqh file:
class C_PosStat in the file Bulaschev_Statistic.mqh
namely in line 324, in the class constructor explicitly assign 0 to the variable All_count_trades.
After editing, line 324 will look like this:
C_PosStat(){PosTotal=0; All_count_trades=0;};There is one more efficiency indicator - Deal efficiency = ProfitPips / MaxProfitPips, where
ProfitPips - profit of the deal.
MaxProfitPips is the maximum possible profit (>= MaxPrice - MinPrice), which could be obtained during the deal existence interval.
But it seems to me that the concept of a trade as an "elementary quantum" of trading (which should be evaluated) is not quite correct at all. And the notion of a trade is almost no better. If we look at trading as a continuous process, then only the position at the current moment and the evaluation of the necessity to change it (resulting from a probabilistic forecast) are important. From this logic, new approaches to developing optimisation criteria may emerge.
// There are a lot of thick nuances, which I am not going to deny.
// But I find the transaction-trade approach to be a hole in its logic. Its casino ears are clearly sticking out of it.
But it seems to me that the concept of a deal as an "elementary quantum" of trade (which must be evaluated) is not quite correct at all. And the notion of a trade is almost no better. If we look at trading as a continuous process, then only the position at the current moment and the evaluation of the necessity to change it (resulting from a probabilistic forecast) are important. From this logic, new approaches to developing optimisation criteria may emerge.
// There are a lot of thick nuances, which I am not going to deny.
// But I find the transaction-trade approach to be a hole in its logic. Casino ears are clearly sticking out of it.
Nice to hear you touched upon the topic of trading approach in general: Balance[k] = Sum(Profit[k]) and Equity[i] = F(Price[i]).
Some of the optimisation criteria are just related to the second (off-trade) logic: profit, recovery factor, max drawdown, etc.
However, one way or another, each symbol has its own ceiling of profitability - MaxEquity (efficiency = 100%). And it (if some nuances are overlooked) is calculated on history on the basis of the post-trade logic.
So justify it and give me an article.
"I don't have time to write articles, I have to trade." (Anti-solver.
;)
However, one way or another, each symbol has its own ceiling of profitability - MaxEquity (efficiency = 100%). And it (if some nuances are omitted) is calculated on the history on the basis of the post-item logic.
But for practical trading it is still more useful to evaluate the "forecasting quality" of the system, no? For example, such criteria would be additive (i.e. it is quite correct to sum them up) on arrays (committees) of trading systems.
And don't "poke me", or I'll start too. ;)
We need to compare two series: MaxEquity[i] and Equity[i]. Both series are known to us. The question is only in the methods of comparing the series. These can be, for example, comparisons of the first derivatives to assess co-directionality and other characteristics.