Hi guys, as some of you know i make EA’s and i`m a good programmer, so i decided to use my programming and math skills to actually make a study about the ATR indicator.ATR is the average true range indicator, and can be found by default in all MT4 platforms.
What it does?
It checks the previous bars and measures the relative size of each bar + considering the volatility to it, that means that a bar that changes color many times will have a higher value on the ATR indicator.And it calculates the average of the previous values to smooth it out.
How useful it is, or is it useful at all?
Since it’s a volatility measurement tool,and since traders are driven by high volatility, since high volatility is actually meaning high volume, therefore no spikes or consolidations which is very very disturbing when you are trading.Be there a trend or a ranging period, volatility is necessary to improve your win rate.Ok but thats only the theoretic side, and its logical, but let’s test this theory that:
HIGH VOLATILITY = HIGH WINRATE = ATR IS USEFUL
But how to test it in practice?
Here is where my awesome programming skills come handy, i made an EA that scalps a secret pattern, i wont tell you which one, but its irrelevant actually, whats important that i`ve set the TAKEPROFIT to 1 pip.
And the exit/stoploss (in case of a losing trade) to the end of the pattern.
Why? Since a small takeprofit is very bad when you trade live, but its very good for research and backtest,since i measured the probability of that candle.You know that a bar until it closes it switches colors many times, it swings up and down like a yoyo, but because the takeprofit is only 1 pip, that means that it locks the winning trade only and exactly when the bar is moving in the direction that i took the trade.
If i lost the trade that means that the bar didnt even moved 1 damn pip in the trade’s direction, therefore it had a direct path in the opposite direction even before i took the trade, and also means that the pattern failed there, for any reason.And since no pattern is 100% accurate, and mine was about on average 90% without applying the ATR indicator, that means that anything that can help my pattern have a higher winrate (not necessarly profit too, but only winrate) will be something useful.
And the ATR did just that.Since i cannot predict my pattern when will fail or not, and the 1 pip takeprofit will increase my patterns winrate to its natural maximum, so anything that can increase its winrate further than its default maximum winrate, must be a good filter, and ATR is a good one.
Let my show you my results:
[B]with no ATR[/B]
[B]ATR period 1[/B]
Bars in test
12483
Bars in test
12481
Ticks modelled
64651170
Ticks modelled
64649925
Modelling quality
n/a
Modelling quality
n/a
Mismatched charts errors
298299
Mismatched charts errors
298303
Initial deposit
10000
Initial deposit
10000
Total net profit
-45.89
Total net profit
-79.88
Gross profit
92.77
Gross profit
10.1
Gross loss
-138.66
Gross loss
-89.98
Profit factor
0.67
Profit factor
0.11
Expected payoff
-0.05
Expected payoff
-0.08
Absolute drawdown
47.2
Absolute drawdown
81.6
Maximal drawdown
48.45 (0.48%)
Maximal drawdown
81.64 (0.82%)
Relative drawdown
0.48% (48.45)
Relative drawdown
0.82% (81.64)
Total trades
1000
Total trades
1000
Short positions (won %)
478 (92.89%)
Short positions (won %)
473 (94.93%)
Long positions (won %)
522 (91.57%)
Long positions (won %)
527 (94.69%)
Profit trades (% of total)
922 (92.20%)
Profit trades (% of total)
948 (94.80%)
Loss trades (% of total)
78 (7.80%)
Loss trades (% of total)
52 (5.20%)
Largest
Largest
profit trade
0.18
profit trade
0.08
loss trade
-6.91
loss trade
-6.91
Average
Average
profit trade
0.1
profit trade
0.01
loss trade
-1.78
loss trade
-1.73
Maximum
Maximum
consecutive wins (profit in money)
41 (4.11)
consecutive wins (profit in money)
72 (0.75)
consecutive losses (loss in money)
3 (-2.33)
consecutive losses (loss in money)
2 (-8.89)
Maximal
Maximal
consecutive profit (count of wins)
4.11 (41)
consecutive profit (count of wins)
0.75 (72)
consecutive loss (count of losses)
-8.89 (2)
consecutive loss (count of losses)
-8.89 (2)
Average
Average
consecutive wins
13
consecutive wins
19
consecutive losses
1
consecutive losses
1
The left one is the “naked” pattern and the right column is the result with the ATR filter, although the net profit is lower, thats only because the 1000 trades that i took was extended in other trading days, since the filter filters a few trades, i had to complete the 1000 trades in each test to have an accurate statistical data.And since both tests were done on 1000 trades and yet the winrate is higher, i can conclude that ATR does help.
Here is the same model but with ATR period set to default (14):
with no ATR
ATR 1
ATR DEFAULT
Bars in test
12483
Bars in test
12481
Bars in test
12482
Ticks modelled
64651170
Ticks modelled
64649925
Ticks modelled
64650556
Modelling quality
n/a
Modelling quality
n/a
Modelling quality
n/a
Mismatched charts errors
298299
Mismatched charts errors
298303
Mismatched charts errors
298300
Initial deposit
10000
Initial deposit
10000
Initial deposit
10000
Total net profit
-45.89
Total net profit
-79.88
Total net profit
-79.52
Gross profit
92.77
Gross profit
10.1
Gross profit
9.86
Gross loss
-138.66
Gross loss
-89.98
Gross loss
-89.38
Profit factor
0.67
Profit factor
0.11
Profit factor
0.11
Expected payoff
-0.05
Expected payoff
-0.08
Expected payoff
-0.08
Absolute drawdown
47.2
Absolute drawdown
81.6
Absolute drawdown
81.5
Maximal drawdown
48.45 (0.48%)
Maximal drawdown
81.64 (0.82%)
Maximal drawdown
81.52 (0.82%)
Relative drawdown
0.48% (48.45)
Relative drawdown
0.82% (81.64)
Relative drawdown
0.82% (81.52)
Total trades
1000
Total trades
1000
Total trades
1000
Short positions (won %)
478 (92.89%)
Short positions (won %)
473 (94.93%)
Short positions (won %)
479 (94.57%)
Long positions (won %)
522 (91.57%)
Long positions (won %)
527 (94.69%)
Long positions (won %)
521 (95.59%)
Profit trades (% of total)
922 (92.20%)
Profit trades (% of total)
948 (94.80%)
Profit trades (% of total)
951 (95.10%)
Loss trades (% of total)
78 (7.80%)
Loss trades (% of total)
52 (5.20%)
Loss trades (% of total)
49 (4.90%)
Largest
Largest
Largest
profit trade
0.18
profit trade
0.08
profit trade
0.04
loss trade
-6.91
loss trade
-6.91
loss trade
-6.54
Average
Average
Average
profit trade
0.1
profit trade
0.01
profit trade
0.01
loss trade
-1.78
loss trade
-1.73
loss trade
-1.82
Maximum
Maximum
Maximum
consecutive wins (profit in money)
41 (4.11)
consecutive wins (profit in money)
72 (0.75)
consecutive wins (profit in money)
72 (0.75)
consecutive losses (loss in money)
3 (-2.33)
consecutive losses (loss in money)
2 (-8.89)
consecutive losses (loss in money)
2 (-8.89)
Maximal
Maximal
Maximal
consecutive profit (count of wins)
4.11 (41)
consecutive profit (count of wins)
0.75 (72)
consecutive profit (count of wins)
0.75 (72)
consecutive loss (count of losses)
-8.89 (2)
consecutive loss (count of losses)
-8.89 (2)
consecutive loss (count of losses)
-8.89 (2)
Average
Average
Average
consecutive wins
13
consecutive wins
19
consecutive wins
20
consecutive losses
1
consecutive losses
1
consecutive losses
1
Now the ATR win a complete victory, over 3% more winrate if the ATR period is set to 14 (default)
Even if you think the net profit is smaller, that doesnt matter because the ATR filters a few trades, then if i run the backtest with and withour the filter, the ATR filter will filter out like 1/3 of the trades and i cant really compare 100 trades to 150 can i? So thats why the net profit is different, but statistically the winrate counts only since the TAKEPROFIT was fixed and the same every time time :41:
So by reading my research results, you should consider adding the ATR indicator to your system, because it works with every system, no matter if its trend following or ranging,swing or even full PRICE ACTION system, it works with all.It can’t harm or decrease the winrate of your system it only can help by filtering 3-4% of the losing trades, no matter what system you use 
Once again the ATR indicator is not a signaling indicator (it signals trade entries), its the opposite, it tells you when NOT to enter a trade because if you do, statistics show that you will have 4% less chance of winning, when the ATR tells you not to.
I used in this study the ATR indicator period set to 14 (default) and trigger set to 0.001 and timeframe H1 (but i guess it works in all timeframes, although not tested).The method is simple, if the ATR line is below 0.001, then you DON’T enter the trade no matter what your system tell you to do,or risk to lose by 4% more.Of course this was just one test here but you can download all my test results in a spreadsheet format below.
I hope I helped you to increase your winrate by 4% with my awesome math and programming skills.It’s a Halloween present for all BabyPips members/readers! Happy Holiday! :43: :43:
ATR RESEARCH.zip (11.8 KB)
