Probability and Trading

[panzerman]

Hello sridhga,

Just take the trade and do your best.

No, that's what this whole thread is about. Just taking your best shot is at best a 50/50 proposition at best. Smallfil is correct about positive expectancy, but that entails detailed back testing, possibly for years, before trading.

Of course there are whole threads and web articles that go into detail about the whole back testing paradigm being bullshit. After years of trying, I'm still only breakeven, and have not found a positive expectancy methodology that fits my demeanor.

I've tried various trend following methods, but always manage to give back the gains of a good trend when price action enters a whipsaw phase. I now use an extreme value methodology, but still have yet to "make it."

 

[SimpleMeLike]

Smallfil is correct about positive expectancy, but that entails detailed back testing, possibly for years, before trading.

Hello panzerman,

And who has years?

 

[smallfil]

%%
Trading+ investing is not nearly as simple as a coin flip, or even much higher probability coin toss.
IT'S much more like a forest or logging business .
Good day for day trades, even though most logging jobs are more than any one day/LOL

You do have to follow the trend and make sure your trades are aligned with the major trend. Major trend has been down. Just follow the trend and make monies like the big boys (mutual funds, banks, brokers, hedge funds). I have been sitting on my trades for a while now. As long as the stockmarket is tanking, I am a happy camper.

 

[murray t turtle]

You do have to follow the trend and make sure your trades are aligned with the major trend. Major trend has been down. Just follow the trend and make monies like the big boys (mutual funds, banks, brokers, hedge funds). I have been sitting on my trades for a while now. As long as the stockmarket is tanking, I am a happy camper.

%%
\SAME here\
except i took most of my profits lately . Good day to daytrade also \ most commercial loggers stop trading i mean working when it rains/too easy to get stuck. Down trend$ are fun\ but can easy claw back profits much faster than uptrends.
Enjoyed my old QQQ chart of half hour trends in MAY 9, 2002. Gapped up 3.33%/one day that week/ then up 10% for that day.......

 

[sridhga]

If your trade has a set of rules to provide specific measurable data points, then you can calculate some backtested odds by looking at a population of past occurances of that setup and trigger. From that you can estimate the probability that the next x number of trades will provide a similar result in toto. ymmv. The waiting for your setup is the hardest part, lol.

In my own case, there is not much wait. I generally get 20 trades a day. I am too exhausted to trade all of them. The trading day lasts 14.5 hours. I trade 3 min charts. That is 290 bars a day.

The pattern I described is called 3 push pattern (also called 'wedge') as discussed by Al Brooks in his books. There is not always a climax. The pattern reverses sometimes with leg count. The probability issue in that case is slightly complicated since the previous trend can go for one more leg before a reversal. This can take out the SL. It will force one to re-enter after that 4th push. Further, if there is a channel forming in the previous direction, this pattern builds within a lower time frame and you then have to trade intuitively.

In my original post, the discussion about probability is not specific to this trade. So, I did not delve into the specifics earlier.

 

[sridhga]

Okay, now let us extend this further. What if I say that a trend has climaxed in one direction on a given time frame and I take a trade after the signal bar on the opposite side formed. assume, my stoploss is the previous climax and my target is twice that distance. Now, is the probability skewed in my favor? What could be the odds?

Let me take a simpler example.

Suppose, there is a trend in a specific direction, say in a higher time frame like daily and I take a with trend-trade once every time the scrip touches 20 ema on hourly. I use reasonable SL and 3 times the target. What are the odds?

 

[SunTrader]

Tooooo vague.

 

[sridhga]

The thing is that markets have several simultaneous trends. One cannot trade successfully without being aware of the full picture.

1. In your example a trader would have to know where this top occurred in relation to the rest of the chart. Trading based on reliable somewhat repetitive behavior will put you well over the 50% line. On a single trade I know I am more likely to win than lose, but I can't put a number on it.
2. In your example we know a top has been made but we don't know if it's the top. Therefore I would say the chance of a wining trade is much less than 50%.
3. Also, don't put your stop on the high. There were sellers there and they could still be there.
4. Last point your target is arbitrary. You need a target that has a logic to it.
​

Thank you for your detailed points.

Point no. 1: You are right. I described that trade in the above post of mine. Sometimes it goes for one more leg in the direction of the earlier trend.
Point no. 2: Agreed. Same as point no. 1
Point no. 3: You are right. This happens and it results in double top/ double bottom at that extreme.
Point no. 4: How would you suggest target setting?

Once again, Thank you for in-depth observations.

 

[sridhga]

Tooooo vague.

???

 

[comagnum]

A good edge applied to a large enough number of trades will have your equity curve hitting new highs often. It takes a great deal of experience & fierce drive/persistence to do this somewhat consistently.

Winning pro gamblers & professional traders have learned how to become the casino.

Pro blackjack players & traders think in terms of the long run. They see it as a marathon, not a sprint. Their math edge will reward them over a large enough number of bets if they can endure the losing streaks which are certain due to variance which is the random outcome of winning bets. The focus is on playing as close to perfect as possible, as long as they do this they will grow their bankroll despite the losing periods.

They do not make money every day, week, or month & do not try to make small gains on a daily basis and stop when ahead, this would only increase how long it takes to get to the long run - they see this as trying to trick the math edge which will not work.