Just a stupid question! What happens if I do the opposite? | Page 2 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Stupid question 1 reply

Stupid question 10 replies

Stupid Question about News Trading! 2 replies

very stupid question but grrrrrrrrr!!!!!!!!! 2 replies

Attachments: Just a stupid question! What happens if I do the opposite?
Exit Attachments

Just a stupid question! What happens if I do the opposite?

  • #21
  • Jun 22, 2014 5:00am Jun 22, 2014 5:00am
hmmm well last week, I entered a trade long with a TP of 25 pips with a 10 pip stop. The moment I entered the trade, it went into profit of about 3 pips then reversed to hit my stop.

Now had I entered short instead of long then it would have been a winner????? Is that right ???
  • #22
  • Jun 22, 2014 5:11am Jun 22, 2014 5:11am
for example 6 red and 4 black with the total of 10 shot....so what is the problem?if i bet on red "in this session" i was for sure winner.
Only correct thanks to the hindsight. You couldn't know in advance there will be more red or more black.

spread in this report is 0.
A backtest without taking the spread into account?!

i want to ask only one thing why i have a different amount of total orders?
Proximus gave the answer provided your strategy only accepts one trade at a time: the duration varies. Some of the signals won't be taken (skipped) in this case. Otherwise I don't know.
No greed. No fear. Just maths.
  • #23
  • Jun 22, 2014 5:26am Jun 22, 2014 5:26am
Quoting Nepul
i will try to explain better what i wanna say. look this report, the strategy is to buy when rsi is < 5 stop loss 5 pip take profit 5 pip http://gyazo.com/ffba507e7126438aaf81ebe5fde9a159 as you can see we have 220 total operation ok let's reverse everything, now the ea should make the same operation but this time sell when rsi <5 same interval of time, same timeframe ,same tp e sl, i should see same operation(220) but : http://gyazo.com/6349b4884d1ad4302a930d109c3961f9 total operation : 188 ( ?!?!?!?) why? i think is not because...
What else is happening in the ea? How open trades at a time? In that time period you ran the test, it could be that if only one trade open at a time, then even though rsi < 5 the the stop loss by spike was hit more often. Since rsi is < 5 another trade was open and that happened more times.

For selling when rsi < 5 there were less trade cuz the bar didnt touch either stop loss or take profit. ranging? So, more bars passed before other trades opened. In that time period you tested look at how the bars are made.

Not all bars are created equal (obviously) .............another reason for there being no 'complete opposite'.
  • #24
  • Jun 22, 2014 6:03am Jun 22, 2014 6:03am
I don't know how many components of a trading system you are looking at but since you mentioned "losing equity line", I'm going to assume just one. Jumping to the conclusion of just doing the opposite of your original premise is not a good idea unless you can find the logic in the move. Besides the equity line, you have to look at expectancy and in order to do that you need to know the Average Win, Percentage of Winners, Average Loss, Percentage of Losers then apply the equation (Average Win * Percentage of Winners) - Average Loss - Percentage of Losers).

You also need to see where the system is failing. Take a look at your entry strategy and exit strategy, determine the slippage. Take a look at your Maximum Favorable Excursion (the maximum profit that the trade had before the trade closed) and your Maximum Adverse Excursion ( the maximum loss that the trade had before the trade closed). This may tell you if you have a bad entry or a bad stop strategy.

As far as expectancy is concerned, your results need to be greater than 1.00. The higher the better. Don't worry about having a high percentage of wins because most often your average win is small. You can be 90% right, but if your average win is $10 and your average loss on your 1 winner is $90 then you can wipe out your profits with 1 trade. I tend to be 50/50 but with a $2.00 winner to $1.00 loser ratio. This gives me a smooth 45 degree equity curve.

The key to trading is to accept that you will have losers, but the average loser has to be smaller than your average win. Big wins just happen, but big losers are unforgiving. Once you get your expectancy to where you want it, just trade bigger size. You can't design a system just to find the "big winners".

This is a pretty simplified way of using money management techniques but it is used by the professionals. BTW, if you go to a systems website, you'll see that some show as many as 18,000 active trading systems posted on this site by traders. This tells me that it's the money management techniques that you employ that make the difference.

If you buy or use anyone's system, ask the designer what is expectancy is. If he doesn't know then it's either bad or he was too lazy to compute it. System designers who just tell you how may they got right are only telling you a small part of the equation. But unfortunately, novice traders only look at this data and that's why systems are constantly sold that way. Also if you learn how all the indicators work then you can back engineer anyone's system. I'm not saying copy someone else's work, I'm saying if you learn how it works, then you will learn where it fails. This may help you improve it.

Finally, know what kind of trader you are: trend, counter-trend, systematic, discretionary, a blend. Learn about little tricks that payoff for you like when you trade counter-trend, set a profit-target, but when you trade the trend, use trailing stops. These little things help improve your bottom line.
  • #25
  • Jun 22, 2014 6:03am Jun 22, 2014 6:03am
Quoting cspip
{quote} What else is happening in the ea? How open trades at a time? In that time period you ran the test, it could be that if only one trade open at a time, then even though rsi < 5 the the stop loss by spike was hit more often. Since rsi is < 5 another trade was open and that happened more times. For selling when rsi < 5 there were less trade cuz the bar didnt touch either stop loss or take profit. ranging? So, more bars passed before other trades opened. In that time period you tested look at how the bars are made. Not all bars are created equal...
yes , there's some error in the test of the ea, but i dont understand why, sometimes it open more order while it just should invert sell and buy.
By the way the idea it's not right if we refer in a exactly interval of time.
Let's explain it in easy word and tell me what i wrong.

STRATEGY :
ONLY BUY WITH TP = 20 AND SL = 10
RESULT:
TOTAL OF 10 OPERATIONS
-10
+20
-10
-10
+20
-10
-10
-10
-10
-10
= -30
STRATEGY REVERSE
NOW WE WILL MAKE THE SAME OPERATION BUT THIS TIME WE SELL
ONLY SELL WITH TP = 10 AND SL 20
TOTAL OPERATIONS ALWAYS 10 !!
RESULT:
+10
-20
+10
+10
-20
+10
+10
+10
+10
+10
= +30

only for this time i didn't use the spread, cause i just want explain what i wanna said and asking this, this is only an example.
why it's not possible to reverse this losing strategy in a winning strategy?
  • #26
  • Jun 22, 2014 6:10am Jun 22, 2014 6:10am
Quoting Nepul
{quote} yes , there's some error in the test of the ea, but i dont understand why, sometimes it open more order while it just should invert sell and buy. By the way the idea it's not right if we refer in a exactly interval of time. Let's explain it in easy word and tell me what i wrong. STRATEGY : ONLY BUY WITH TP = 20 AND SL = 10 RESULT: TOTAL OF 10 OPERATIONS -10 +20 -10 -10 +20 -10 -10 -10 -10 -10 = -30 STRATEGY REVERSE NOW WE WILL MAKE THE SAME OPERATION BUT THIS TIME WE SELL ONLY SELL WITH TP = 10 AND SL 20 TOTAL OPERATIONS ALWAYS 10 !! RESULT:...
That's only theoretical. In real trading, it will never be the case.

If you want to be a trader, you should concentrate on the essential : trading ... Obviously that's not what you are doing.
MT4/MT5 EAs/Indicators/Alerts coder
  • #27
  • Jun 22, 2014 6:20am Jun 22, 2014 6:20am
There seems to be a lot going on in this thread. In my opinion, once you enter and exit a trade. It's done. It has its own stats (i.e. Average Win, Average Loss). Going the other way makes it a different trade and its stats will be unique. If you do that frequently then all of the results should be taken into consideration and that becomes your track records, your system or whatever you want to call it.

If you backtest a system that shows you are consistently up $200 before it turns into a losing trade and your stop loss is for $50.00 then when you do the opposite, you will have to withstand $200 of loss in order to capture a $50.00 winner. So you 1 loss can wipe out 4 winners. Before you consider just flipping a system and doing the opposite, you have to know the MAE and the MFE of the system.

Getting stopped out and going the other way is not reversing the system and catching a winner, it may just be that you're not good at identifying the trend and went the wrong way the first time. It shows me that you're good at taking fast losses when you are wrong and fixing the situation.

The first trade of the day for me is usually a loss but I don't let it bother me. Sometimes I have to get the feel for the day, is it trending, non-trending, momentum, no momentum. I am a blend of charts and discretion, that's the 90% part, the 10% is the art of trading. If you trade with enough capital and with a clear head, it opens you up to getting a "feel" for the market. This suits me. It may not be right for everyone. That's why we call it a market.

If you are too emotional, you block out all the good information the market is revealing and attracting all the stuff you want to see. Learn to block that out and go with the flow. Eventually all you want to be is smarter than the other guy and pick off the dumb money available everyday. Eventually, you'll learn the traps that are set each day to expose and eliminate the weak or undercapitalized.
  • #28
  • Jun 22, 2014 6:43am Jun 22, 2014 6:43am
I saw a post here where the trader was using the RSI with a stop loss. Go back to the original RSI equation. It uses CLOSES so using a stop loss based on a previous high or low is financial suicide. It has nothing to do with equation. The other thing. It was developed as an indicator not a trading tool. The general rule is over 70 is overbought and under 30 is oversold.

But the author of the indicator, Welles Wilder, also said that a move over 70 is a sign of strength and a move under 30 is a sign of weakness. Traders who think the lower it is, the bigger the expected rally, may not understand how the signal actually works.

Before you use an of these "short-cut" indicators, learn to chart the market. Make the charts by hand so you can actually see what is going on in those 14 days the RSI calculation is looking at. The logic behind the signal says that a market that is in a downtrend tend to close near the low of the session and lower for the session. If a market is getting ready to turn up after a prolonged down move then you will start to see it close higher or near the high of the session.

Too much emphasis is placed on the indicator and not enough on the price action. If the indicator is based on the close and your stop is based on the price, chances are your stop will be hit because the stop is not related to the indicator. Example: If the RSI is falling and under 5 as the poster stated and he buys with a stop under the previous low, the market can trade as low as it wants to and still close higher. This would mean your stop is hit on the day the RSI turned.
  • #29
  • Jun 22, 2014 7:15am Jun 22, 2014 7:15am
Quoting Hyerczyk
There seems to be a lot going on in this thread. In my opinion, once you enter and exit a trade. It's done. It has its own stats (i.e. Average Win, Average Loss). Going the other way makes it a different trade and its stats will be unique. If you do that frequently then all of the results should be taken into consideration and that becomes your track records, your system or whatever you want to call it. If you backtest a system that shows you are consistently up $200 before it turns into a losing trade and your stop loss is for $50.00 then when...
You're right i will have a 200$ drawdrawn ( that before was my gain) but we dont care, we need that it's losing in the long period.
For now i still can say that if i have a losing strategy, for losing i want to mean that it lose in the backtest for years and years, i can reverse it in winner..
i really thing it's not possible cause is too easy to be true , but i just need a reason!can somebody help me to know why is not really possible ? getting the example of the #post 25 ??
Before i explained a losing strategy with rsi but it was just an example...we can use bollinger or others, i dont care...
  • #30
  • Jun 22, 2014 7:22am Jun 22, 2014 7:22am
Quoting Nepul
i really thing it's not possible cause is too easy to be true , but i just need a reason!can somebody help me to know why is not really possible ?
The answer is in the hundreds of threads opened by dreamers since many years ... You will find easily these threads with a simple search ...

Don't get me wrong. I have nothing against you ... Just that you are obviously new to trading and that isn't the right way for a good start ...
MT4/MT5 EAs/Indicators/Alerts coder
  • #31
  • Jun 22, 2014 10:13am Jun 22, 2014 10:13am
Quoting Nepul
hi traders ! in this period i'm trying to explore and understand better all the world around the expert advisors. So i'm testing a lot of ea and i'm noticing that the most i'm testing are with a losing equity line, so i wondering what if i do the opposite? I know maybe it's a really stupid question, but if i open a buy order with a 20 pips take profit and 10 stop loss and this order go to stop loss (-10) it mean that if i sold (and not bought) with 20 pips stop loss and 10 take profit i was in profit ( +10). So look this photo: http://gyazo.com/68d2d206e6592f8ce0b5d847aaa8fe8e...
this is a perfect losing strategy , but what if i made the perfect opposite in every single trade??
What if, in this strategy,it was selling but i will buy and vice versa?

The problem is you will get the same losing system, with long or short trades, because the stops and take profit are pretty close together and PA will gobble you up with both directions................
The More You Learn, The More You Earn.....
  • #32
  • Jun 22, 2014 12:10pm Jun 22, 2014 12:10pm
Quoting FerruFx
{quote} The answer is in the hundreds of threads opened by dreamers since many years ... You will find easily these threads with a simple search ... Don't get me wrong. I have nothing against you ... Just that you are obviously new to trading and that isn't the right way for a good start ...
well, thanks you very much FerruFx for your advice, by the way it's evident you are too expert to talk with somebody who "you think" is starting now in the trading world, so please dont lose your time to write in this thread if you believe it's all already written...
bye bye
  • #33
  • Jun 22, 2014 12:18pm Jun 22, 2014 12:18pm
Nepul, go ahead and trade your opposite. Only then you will know what happens for sure. :nerd:

Keep us posted.
  • #34
  • Jun 22, 2014 12:27pm Jun 22, 2014 12:27pm
Yes go ahead and do what you want. I will be looking for your money with a big red bow on it.
  • #35
  • Jun 22, 2014 12:49pm Jun 22, 2014 12:49pm
So, now i made a very easy ea and its reverse
this ea only buy when rsi = 5
the other ea is the same but it sell when rsi =5
i use 15 timeframe...SL e TP 50
useless to say that is not a losing strategy or winning, but i just need to make a very simple reverse.
at the end it work well , but if you test it for a long period it turn back some errors that i dont understand.

period from 2013/01/01 to 2013/01/11
normal ea report :
http://gyazo.com/b4832089a3eed51aca267717b3b01927 (LOSING)

reversed ea( rsi_ugualecontravalue_ea) report :
http://gyazo.com/5927126ff08f877b926c570583b41a22 (WINNER)

as you can see in both of the ea we have the same total operation : 8 .

period from 2013/01/01 to 2013/03/30
normal ea RSI_ugualvalue_EA) report:
http://gyazo.com/041676a2e9399755170ab9554f381bb4

reversed ea report :
http://gyazo.com/3db8a4a35d5f828af71e28792c2521d5

here, if the interval of period to test is bigger backtest is not more good, it turn back some error, in fact the amount of the total operation is not the same.
Attached File(s)
File Type: mq4 _RSI_ugualecontraValue_EA.mq4   4 KB | 289 downloads
File Type: mq4 RSI_ugualeValue_EA.mq4   4 KB | 268 downloads
  • #36
  • Jun 22, 2014 2:23pm Jun 22, 2014 2:23pm
As it was explained earlier, when your RSI hits 5, price likely broke out into down trend. By going counter to your original logic you capture a portion of the trend. Reversals likely happen at higher readings of the indicator, but you rarely capture them when trying to buy at 5.
  • #37
  • Jun 22, 2014 4:38pm Jun 22, 2014 4:38pm
Quoting Nepul
ok, PipMeUp, i'm sorry, but i'm still not understanding...spread in this report is 0 . Please read # Post 4 , i posted 2 link of two report, first one ea only buy , in the second i reversed everything so ea only sell , with the opposite tp and sl . i want to ask only one thing why i have a different amount of total orders? the signal is the same! so i should have the same order the only difference should be buy and not sell, or sell and not buy. can you understand what i'm saying?
Looking at the reports you posted there is a spread of 1.3 pips. You might be using Metatrader for backtesting and setting spread to 0 but then Metatrader will use current market rate for the spread (Don't ask me why).

Remember, in a strategy relying on TP and SL you aren't just losing the spread, the spread is putting you closer to the SL from the trade entry and increasing the probability of a loss. Hence you're getting 35:65 in 1 case and 40:60 in the other. The difference between the positive and negative case is likely due to market conditions during the test.

Try setting your SL to be the same as TP plus the spread, the result will be closer to 50:50 but of course the average loss will be greater than the average profit.
  • #38
  • Jun 22, 2014 8:16pm Jun 22, 2014 8:16pm
Quoting Nepul
So, now i made a very easy ea and its reverse this ea only buy when rsi = 5 the other ea is the same but it sell when rsi =5 i use 15 timeframe...SL e TP 50 useless to say that is not a losing strategy or winning, but i just need to make a very simple reverse. at the end it work well , but if you test it for a long period it turn back some errors that i dont understand. period from 2013/01/01 to 2013/01/11 normal ea report : http://gyazo.com/b4832089a3eed51aca267717b3b01927 (LOSING) reversed ea( rsi_ugualecontravalue_ea) report : http://gyazo.com/5927126ff08f877b926c570583b41a22...
It seems like you're just curve-fitting the signal just to get what you want to see.
  • #39
  • Aug 10, 2014 10:39am Aug 10, 2014 10:39am
If you reverse a losing strategy then its not necessarily you will make money , however if you reverse a losing signal then definitely you will make money minus costs , so as long as these costs and commissions dont add up then you will make money , example :
Quoting hanover
Suppose I play roulette, and I bet continuously on black for a session. At the end my net P/L is a loss of $500. Does that mean if I bet on red continuously, I have a winning roulette strategy? If you can see the fallacy in that logic, then you'll also understand why inverting a 'losing' forex strategy doesn't work, either. If you don't understand, read Fooled By Randomness by Taleb. Like FerruFX says, this question gets asked on this forum at least once every...
If i took the opposite of your bets at the same table at the same time then i will definitely have made money minus costs - and the cost here is when it lands on the table zero - .
  • #40
  • Aug 10, 2014 10:57am Aug 10, 2014 10:57am
Quoting tar
{quote} If i took the opposite of your bets at the same table at the same time then i will definitely have made money minus costs - and the cost here is when it lands on the table zero - .
This is of course wrong. You would make money if you knew in advance that hanover will lose. But if that particular day you oppose his bets he is lucky and he wins you lose the same amount as he won. Each time he wins you lose. Each time he loses you win. Altogether you make no money. In the roulette example the probability he loses is 50% whatever the number of times he bets.
No greed. No fear. Just maths.
0 traders viewing now