How Do You Know??

C

CreditViolet

Guest
#1
There are so many indicators nowdays that a TA beginner will think that lining up all the indicators and getting into a trade will guarantee a profit.
But sure he will find out that he cant do that.The reasons are obvious

My question is how would u choose to filter out ur trades?Whats the crieteria?Why would u choose a 18 period stochastic over a 14 one or ADX or Money Flow Index.Sure you can try them one by one losing money and perhaps u will find one of them that works perfectly.But how do u know that there is not anything better to use.

Answer Simple-backtest.Test ur historical data applying the strategies and see how they did in the past.This is such a neglected concept mainly because to backtest u have to learn programming.All trading languages are quite easy compare to full blown programming languages and if u r serious about trading u have to know at least one of them to underatand u r strategies better

I just thought that this concept was being neglected here so this post was just to bring this to all ur attention.I just have around 2 years of experience in the markets and it was lots of backtesting that has given me the understanding about things that work and those are just hogwash.

One of the reasons the myth of elliott wave survives is because u cant bactest them easily.U add new data u get new waves.More the data different the wave count.So they have got the power of hindsight which lot of systems have not.

Questions like which period of stochastic or macd can be easily solved if u just know a little programming and backtest.Try it and u will never be the same.
 
#2
What if one knows nothing about programming? Is there some site where this can also be learnt easily,at least to the extent required from Traidng angle?

Shri.
 

sh50

Active Member
#3
You(credit) have again given some very useful tips. I know fundamentals well but am new to technicals. My understanding of backtesting is that once you plot the indicator on a graph you have to see how it behaves with various ups and downs in giving reasonably accurate information in all markets and keep changing the parameters of the indicator till it is accurate enough? Why should one know programming for this.? I thought programming was required for fancy explorations. Maybe if you could give an example of how exactly you have backtested, one could understand what you are saying.

You are right about this trial and error causing monetary loss. Cant the losses be minimized by attempting intra-day under somebodys guidance. Intraday maybe the toughest but one can learn much faster. Some people say that the fundas of intraday are different. I feel that perhaps it may hit the oversold, overbought more frequently and the price objectives may be smaller compared to end of day trading. Purely from a learning perspective, it should be better?
 
C

CreditViolet

Guest
#4
shrees said:
What if one knows nothing about programming? Is there some site where this can also be learnt easily,at least to the extent required from Traidng angle?

Shri.

Shrees ur doubts have much to do with the myths the trading industry has perpetuated so long .Now take a look at this
If the Close > High of 1 Bar Ago Then Buy at Market;

Was that tough.I am not saying that it is easy either but hey man what in life is.To learn it the best way is to mess around with it.Google a bit and u find lots of info how to get started.It also depends on the program u r using.
If u learn just one all will look similar so i suggest get started and ofcourse i am here to help anyway

CV :cool:
 
C

CreditViolet

Guest
#5
sh50 said:
My understanding of backtesting is that once you plot the indicator on a graph you have to see how it behaves with various ups and downs in giving reasonably accurate information in all markets and keep changing the parameters of the indicator till it is accurate enough?
Actually sh u can program any kind of pattern like triangles,ross hooks whatever u wish and then test its reliability over historical data.

sh50 said:
Why should one know programming for this.? I thought programming was required for fancy explorations.
U can change parameters like days or no. of bars but to access the full
understanding of the concept u do need to learn a little abt pro.

sh50 said:
Maybe if you could give an example of how exactly you have backtested, one could understand what you are saying.

This is written in Tradestation 2000i Easy Language

This trading signal is designed to buy when the RSI has double dipped into
the oversold territory and sells when the RSI has doubly risen into overbought territory. Once a position is initiated an initial money management stop and profit objective stop is invoked.}
Inputs: rsiLength(14), overSold(40), overBought(60), moneyManStop(1000);
Vars: myRsiVal(0),longProtStop(0),shortProtStop(99999),obCount(0),osCount(0);
Vars: longProfitStop(99999),shortProfitStop(0),takeProfitStop(2000);
{Calculation Module}
myRsiVal = RSI(close,rsiLength);
if(myRsiVal > overBought and myRsiVal[1] < overBought) then
begin
obCount = obCount + 1;
osCount = 0; {Reset the OS counter since we are OB}
end;
if(myRsiVal < overSold and myRsiVal[1] > overSold) then
begin
osCount = osCount + 1;
obCount = 0; {Rest the OB counter since we are OS}
end;
if(marketPosition = 1) then
begin
longProtStop = entryprice (moneyManStop/PointValue/PriceScale);
longProfitStop = entryprice + (takeProfitStop/PointValue/PriceScale);
osCount = 0; {Since we are long reset the OS counter}
end;
if(marketPosition = 1) then
begin
shortProtStop = entryprice + (moneyManStop/PointValue/PriceScale);
shortProfitStop = entryprice (takeProfitStop/PointValue/PriceScale);
obCount = 0; {Since we are short reset the OB counter}
end;
{Order Placement Module}
if(obCount = 2) then Sell Short tomorrow at open;
if(osCount = 2) then Buy tomorrow at open;
if(marketPosition = 1) then
begin
Sell ("longLoss")next bar at longProtStop on a stop;
Sell ("longProfit")next bar at longProfitStop on a limit;
end;
if(marketPosition = 1) then
begin
BuyToCover ("shortLoss")next bar at shortProtStop on a stop;
BuyToCover ("shortProfit")next bar at shortProfitStop on a limit;
end;

Now if u look carefully u need to learn just a handful of concepts rest r repetitve.U will also notice that everything possible angle is covered including money management.And the beauty is that u can completely test whether this strategy worked or not.It may look a little complex but nebody can write this code in say 10 days of getting started.


sh50 said:
Cant the losses be minimized by attempting intra-day under somebodys guidance. Intraday maybe the toughest but one can learn much faster.Some people say that the fundas of intraday are different.
The advantage of intraday is u get to control risk.Unlike EOD u can get in and get out whenever u like.But that can be a useless advantages if u dont manage risk properly.The matter of fact is that any system in any time frame will work if it just has a good positive expectancy.Even flipping the coin if u learn well enough about positon sizing.

All great traders were loners and self taught.Guidance is alright if u dont get dependent on the person but as u know thats not the case.There is no substitute to self learning in the markets.As u would have realised that really good traders dont go around teaching their skills for peanuts coimpared to the profits they can make trading themsleves.It still is a advantage to get guidance for the first few times if it helps u find ur niche
 

sh50

Active Member
#6
Thanks for answering so specifically and explicitly. I will have to read and digest the programming part before I comment on it. You are right about not being dependent after asking for guidance the first few times and about traders not being able to spare time because of the penny wise pound foolish element. In that context, supposing you get somebody to guide you for a couple of months, would not intra-day trading be better because the number of situations you face in that are much more than in end-of-day where you take positions and got knows when the market may go up or down? Ultimately, learning is possible only through diverse practical situations and the more situations you get to face in the limited time span, the better isnt it. Intraday alone can ensure that-thats what my common sense says. You would know better. The other option could be to use as many scrips for day trading for back testing as possible?
 
Last edited:
#8
sh50 said:
supposing you get somebody to guide you for a couple of months, would not intra-day trading be better because the number of situations you face in that are much more than in end-of-day where you take positions and got knows when the market may go up or down? Ultimately, learning is possible only through diverse practical situations and the more situations you get to face in the limited time span, the better isnt it. Intraday alone can ensure that-thats what my common sense says. You would know better. The other option could be to use as many scrips for day trading for back testing as possible?
Hi,

Well I would also like to know what you all think about this because throughout this site there seem to be quite a lot of posts which say that intra day trading may not be profitable. In fact on one of his posts traderji had even suggested that you might in all probability lose a lot of money in intraday trading. There was some post which really said "please say no to intraday" So what is it exactly?

Regards,
snowy
 
#9
snowy said:
Hi,

Well I would also like to know what you all think about this because throughout this site there seem to be quite a lot of posts which say that intra day trading may not be profitable. In fact on one of his posts traderji had even suggested that you might in all probability lose a lot of money in intraday trading. There was some post which really said "please say no to intraday" So what is it exactly?

Regards,
snowy
What works for one person may not work for another.

The best way to know is try it for yourself. Like myself I tried it all and finally found what I was comfortable with.
 

sh50

Active Member
#10
Credit, the program that you have given is probably effective but at this stage one has to develop the basics first. Recently I read that Parthiv Patel is also having wicketkeeping problems only because they tried to catapult him straight to international cricket without making him go through the grind.

Strictly in that context, dont you think that if one could see various tops and bottoms with basic indicators like RSI, ROC, Stochastics, MACD etc with a couple of volume and volatility indicators, isnt good enough. One will face more situations in intra-day trading wont one?