Over or underperformance by TA signals

M

MrAgi1

I just created a thread some minutes ago. https://www.elitetrader.com/et/threads/are-there-verifiable-ta-strategies-with-high-win-rate.370722/ However it appears I did not properly explain my question, so I locked the the thread(reported it), since I cannot delete or edit my initial question.



Okay now I would use an example to make my question clearer.

Let’s say I have 2 separate accounts. No stop losses and take profits are used on placed trades. However every trade automatically closes after 10 days on both accounts.

Let’s say in both accounts A and B I only buy or sell stock X and hold it for 10 days.

In account A, trades are entered randomly(either buy or sell) and held for 10 days. In account B trades are entered either based on chart patterns or indicator signals but the trades are closed after 10 days also.


Over 1000 stock X trades(either buy or sell) have been placed both in account A and B. Is it possible that
Account B would have a significantly different result(much worse or much better) than account A depending on the indicator use? Or would all indicators yield the same results as a random entry?


That’s the summary or the idea I get from this Tastytrade video:

However, I think I have come across a research paper in the past(unfortunately I don’t remember where) that shows a certain indicator signal consistently underperformed a 50/50 coin flip. That is the opposite of what Tastytrade is saying that video. Hanging man strategy is 66% but could have been 50 on another underlying or timeframe(no consistency).


Key word is consistency( on different underlying or timeframes). So is it possible to find an indicator/chart pattern that would always consistently underperform or over perform random entry(50/50 coin flip), no matter the underlying or time frame? Example: Is it possible a certain indicator X would always result in a 30% win-rate(over different underlying or timeframes) as compared to a coin flip that is always around 50/50 over large samples.
 
Last edited:
Handle123 said:
Trend is your friend, ever hear of this? People spend so little time on developing exiting routines and best ways are never exposed in forums. Hundreds of ways to enter but few really good ways to exit.
More...
%%
Exactly+ most likely the daily shooting star does not do a anywhere that well in a downtrend.
More like a shooting sink that sinks. May have done better in OCT, but havent even tested that last sentence
 
MrAgi1 said:
Or would all indicators yield the same results as a random entry?
More...

Absolutely, your account would be seriously depleted, if the transaction costs don't get you the markets will, they are designed that at some point, an event will occur to induce a loss of profits and if not exited, which no one ever does, then capital.

I have access to a system that kind of has a 10:1 win rate, was development consulting for some people (crypto arb and hedge fund trading) who had access to it, they let me keep using Forex, last week generated 85% return on account in to the October volatility.

The markets are arbitraged to one outcome, you become poor, they become wealthy, the only way to avoid this is be under the 'limit' of markets acceptable returns which today is pathetic, or you can time when the 'event' occurs and exit before.

Unfortunately even billionaires and the highest level elites can't do this, they can time the upswing but the inversion has too many moving parts, basically Account A and account B at some point will give the correct answer, and then diverge in cycles ebbing closer and further away, the 'event' will make them diverge at the maximum spread where you will capitulate, it is a forgone conclusion.
 
I finally designed a system that wins between 1-4% every year on the underlying BUT with correct hedging, overall losing is approx 2%. Took me 30 years to be totally happy with it.

If you can't develop risk management, you stumble through life.
 
Using the S&P500 universe, buy at the close when RSI2<7. Sell at the close when RSI2>73 or after 10 days. Going back to 1996, you would have a 7,462% return on your money with a CAGR of just under 20%.
 
Let's say you are targeting reward : risk ratio of > 3:1

Even if you use chart pattern,
trade with a trend (the trend is your friend),
hanging man / shooting start / dark cloud .... thing,
price action,
Miracle indicator X,
moving average cross,
blablablabla,

your success rate is going to be like 1%.

Now talk about 'trend is your friend'.
Unfortunately, the market is mostly unkind;
the market could move in an erratic, chaotic, jerky... manner.

Occasionally the market trends/moves very nicely and
is easy to earn money.
even a blind can make money from trading such market.

work very very hard so as to be a successful trader.
And you need some form of filter / confirmation / macro view
to improve the win rate.


upload_2022-11-2_12-54-17.png

The above % win looks abnormal.
I'd shift the decimal place to 2 digits to the left.
so the win % will be like 0.5%
If you see such a video, quickly run away from it.

PS : I didn't watch that video.
 
alistera said:
I have access to a system that kind of has a 10:1 win rate, was development consulting for some people (crypto arb and hedge fund trading) who had access to it, they let me keep using Forex, last week generated 85% return on account in to the October volatility.
More...
Is that somewhere available for the public because I would be interested in such a system with 10:1 win rate on Forex. Let me know this. Thanks.
 
You must log in or register to reply here.
Back
Top