PRICE ACTION vs BACK TESTING

I have written a small program to allow user to back test and looking at technical indicators. Basically when user presses "new trade" random stock symbol is picked at random dates. At that point you can buy, short "new trade" or fast forward one day. This allow you to see if you can pick winners by looking at real stock data using technical indicators.
I have written this program for my own use but you are welcome to use it.

Charts are taken from bigcharts.com as images and sock data it taken from yahoo.com


http://sites.google.com/site/testmytrades/
 
Quote from Swan Noir:

Here is my conceptual problem. My experience is that in business to be marginally profitable may not mean you are on the verge of success. It is just as likely you have had a run that is essentially random and that you have no read on where you are. If anything you may be at more risk than if you are losing because it is easy to convince yourself you are doing well for a novice and believe that as you build on your experience the deal starts to deliver more money and more consistency. Random success can, and frequently does, prevent novices from looking for a truly profitable business model.
More...

You are already ahead of the pack in this respect -- that is, you are anticipating the reasons why long-term success in trading is so elusive. Can you already see the problems with backtesting itself as so-called "proof" of an edge?

You seem to want to pit a longer-term trading time frame based on research vs a shorter, more immediate methodology based on direct observation. Whether these two are diametrically opposed at heart is questionable, but I think the instinct you have is correct. Backtesting, unfortunately, can only be done with archived data, looked at in isolation, whereas those trading primarily off price action get nearly immediate reinforcement that includes all other conditions (technical, fundamental, news flow) in real-time.

I guess what I'm saying is this: if you can trade directly off price action itself, you will be far less prone to falling into that trap of random luck masquerading as long-term edge that you fear. The nature of PA itself is never stable for long, as different market conditions even day to day will change the way prices move. But if you can master this mutability and come to terms with uncertainty, you will always necessarily be on your toes, always a bit unsure of yourself but also wary of never becoming complacent. That is how you keep yourself from being fooled by randomness, so to speak.

For me, the glaring difference between using backtesting vs price action is not the technique in itself, but the attitude of the trader in relationship to the market. Does a trader need certainty in order to prosper long term, or can he succeed by just always staying ahead of an ever-shifting playing field?
 
Quote from illiquid:

[B

For me, the glaring difference between using backtesting vs price action is not the technique in itself, but the attitude of the trader in relationship to the market. Does a trader need certainty in order to prosper long term, or can he succeed by just always staying ahead of an ever-shifting playing field? [/B]
More...

For me it is PA.

I have heard the rationals -- largely good ones -- for the scientific approach that is able to quantify probabilities in ways that lead to calm unemotional, sometimes even automated, entry and exit points.

I have even suffered through those that state we are in a new age and hour after hour of quiet uninterrupted observation of nuance as well as the more obvious went out the buggy whip.

I wonder if the fact that in 1928 the words "New Era" appeared in the names of newly started business entities (corporate and otherwise) more often than any other two descriptive two word set would give them pause.

I am convinced those who truly master PA, as a discretionary trader must, are destined to make money in any market. And very serious folding money when the market allows. I am sure there will be times when I will have to test and be guiding by the numbers on discrete points. But the overarching concepts that interest me intraday are very difficult to learn and do not easily lend themselves to testing. That does not make them less reliable or any less profitable.

They, as you point out, require frequent modification -- a sort of bending to the NOW -- and the ability to execute on what many would say is insufficient information ... grossly insufficient to some.

I have all my life seen that decisions are largely made without all the information but with what the decision maker perceives as enough information to allow him to calculate the risk/reward "close enough". To say I'm comfortable putting money on the line at those points would, frankly, be a stretch.

Yet I have observed over many years that I am decidedly less uncomfortable than any of my colleagues.

I can and have lived there while still sleeping soundly almost every night.

I thank you for your post and taking the time to pass on your view. It has value to me.
 
Quote from Swan Noir:

It may very well be to soon to tell anything from it but I consider it interesting that no one has stepped up to say that they trade price action profitably.

Anyone care to tell us that they trade PA profitably?
More...
I trade only PA (on a chart) to positive effect. My approach is almost entirely systematic, however, it cannot be traded across all markets or in all time frames. I have backtested it manually and followed it in real time to establish a familiarity and comfort level before committing to it.

You seem familiar. Have we met?
 
I'm an idiot savant. If you are idiot to spend as much time looking at charts and backtesting things as I have eventually you become a savant... I can put an idea up on a chart, scroll through and either I see a lot of value in it or I know right off it's random... I have stuff I use that never is backtested.. the value of it is apparent from the get go... I've tested an awful lot of stuff that looked great initially and was truly random... now I don't even have to test anything..

The biggest lesson I learned is "Bulls make money, Bears make money but Pigs get slaughtered".. If you are not happy with grabbing some steady gain on a few lots often enough and you want to optimize to make it better, you will get slaughtered...
 
Quote from illiquid:

You are already ahead of the pack in this respect -- that is, you are anticipating the reasons why long-term success in trading is so elusive. Can you already see the problems with backtesting itself as so-called "proof" of an edge?

You seem to want to pit a longer-term trading time frame based on research vs a shorter, more immediate methodology based on direct observation. Whether these two are diametrically opposed at heart is questionable, but I think the instinct you have is correct. Backtesting, unfortunately, can only be done with archived data, looked at in isolation, whereas those trading primarily off price action get nearly immediate reinforcement that includes all other conditions (technical, fundamental, news flow) in real-time.

I guess what I'm saying is this: if you can trade directly off price action itself, you will be far less prone to falling into that trap of random luck masquerading as long-term edge that you fear. The nature of PA itself is never stable for long, as different market conditions even day to day will change the way prices move. But if you can master this mutability and come to terms with uncertainty, you will always necessarily be on your toes, always a bit unsure of yourself but also wary of never becoming complacent. That is how you keep yourself from being fooled by randomness, so to speak.

For me, the glaring difference between using backtesting vs price action is not the technique in itself, but the attitude of the trader in relationship to the market. Does a trader need certainty in order to prosper long term, or can he succeed by just always staying ahead of an ever-shifting playing field?
More...

Sorry I just saw this thread




I trade PA profitably, and I agree with this gentleman completely


FWIW

What I did – along with spending a bunch of hours in front of a live screen – I would also capture mkt data in camtasia, (and similar products before camtasia), then play it out at night (after mkt) – studying it at great length and doing what if trades.


Aside

I had a hard time printing out charts and covering them up (the right side with a sheet of paper) – as I would inevitably see the whole chart and be prejudiced before playing it out.


eta - PA is all about the attitude - you approach the mkt/ trade with - imo


Regards

RN
 
There is not really any such thing as Price Action. So called PA uses nothing but price? Wrong. If it did all you would need is one only single market price on the screen changing as price goes up or down. So called price action uses the history of price on the screen up to the current latest second price. That price is on a chart and the preceding patterns up to the latest price are assessed for determining whether you buy or sell. So instinct or remembered observation is being used instead of indicators to decide buying or selling. Either such human observation or alternatively indicators may not be good enough for good trading, especially so for amateurs.

Now lets consider back testing. It is limited and fallible. It relies on usually OHLCs of daily sessions over a given period. OHLC is not useless but it is very insufficient as an input. Hence the frequent disappointment experienced by the hordes of back testers always sustaining themselves on obstinate ignorance and blind faith.

So where is the starting point? The starting point is observing where the money is. As obvious as it is, it is not obvious to many. The money is in the daily gyrations of a volatile and liquid market. A gyration is a swing up followed by a swing down (or vice versa). Generally I use, arbitrarily, 35 points, bar close to bar close, as the minimum for a swing (ie CL,NG,YM). The gyrations are where you begin your studies with a view to making yourself rich.
:)
 
You must log in or register to reply here.
Back
Top