Stock patterns are not something to make trades from

[rangerdanger12]

My opinion is that if you are manually day trading, meaning you are buying three or four stocks and trading them all day.

That executing a trade just based on a pattern makes no sense to me. Sure the stock formed a head and shoulders but what it does after that pattern forms is anybodies guess.

Now over an entire universe of 5000 stocks, it might be that statistically 51% of the time the head and shoulders pattern results in a particular type of move after its completed.

Which means if you can trade a large sample of stocks at the same time and design an algo that can recognize the head and shoulders pattern and execute the same time, then you can be right 51% of the time and then the pattern of a head and shoulders gives you an edge allowing you to be positive that day.

But if you trade three of four stocks your sample size is too small for you to just use a head and shoulders pattern to make a trade. Your stock choices for that day, that end up forming the head and shoulders pattern might be part of the 49% that didn't do what the pattern predicts and you are net loss for the day.

You're stock selection the next day may be fraught with the same problems selecting stocks that are not doing what their pattern predicts.

But then the next week you select all stocks that follow through on their pattern and break even with the week before.

You must risk the same amount of money every time in order for the statistics to make sense, if you for example keep putting 1000 shares one week trying to play patterns and they all turn out to be losers, and the next week you only put 100 shares and the patterns work out you will still be net negative.

I feel like a pattern is a good indicator to use, but that a stock has its own language everyday it speaks another language and using support and resistance and looking at charts as tracks of the big buyers and sellers you can see who wants to buy and sell at what price. The guys who move the market.

If you are trading three or four stocks in a day, I feel that the supply and demand fundamentals are the most important . At what prices are the most volume being done at.

What are your guys thoughts

 

[BeginnerJoe]

So what you are saying is that patterns are no use, therefore we should look at patterns (of other random factors) instead ?

 

[rangerdanger12]

No what I'm trying to say is that patterns only work with a large sample size of stocks being played at the same time. I was reading a book that had multiple case studies that shows that for example an ascending triangle in a short term period of 0 to 3 months had an upward breakout 38% of the time and a downward breakout 21% of the time, sample size 241 stocks, while in other cases doing nothing significant, that means a 5% move in either direction.

You're edge here for a 5% move up with an ascending triangle is 38-21 = 17%.

That means if you place 100 trades in one stock, or trade 100 stocks and place one trade that 17% of the time using an ascending triangle you will get a 5% upmove.

But in order for this strategy to work you need a large enough sample size, to utilize the edge.

If you see an ascending triangle one day in a stock you are trading, and put a 1000 share trade you are only 38% likely to have it go up, and 21% likely to go down, and the rest 41% do nothing significant.

Now thats the danger of sticking to one stock when trying to use chart patterns.

Suppose you win this week with that 1000 share trade and you make $500, next week it does nothing and the week after that only trading on an ascending triangle formation you lose $500.

But if you traded 100 stocks everyday instead of just one stock over a period of a week then you should have a net profit of 17%. Because 38% will win, 21% will loose and the 49% will be more or less flat.

My argument is that to just charts as an indicator for trades only works spread out over a large sample size and not for individual trades

 

[BeginnerJoe]

Does that mean if I go to a casino, I will win more if I play on all roulette tables simultaneously than if I just stick to 1 table ? Red or black has maybe 49% chance of winning. Why would you play with triangles when casino offers better odds ?

 

[BeginnerJoe]

Why would you play with triangles when casino offers better odds ?

I'll answer myself: because casino has higher odds of losing, even though it has higher odds of winning.

 

[wackypete2]

That executing a trade just based on a pattern makes no sense to me. Sure the stock formed a head and shoulders but what it does after that pattern forms is anybodies guess.

What price does on the next tick is anyone's guess all day long across thousands of stocks, commodities, currencies, financials, etc... You can't use that statement alone to justify patterns are no good.
Manage your trade, don't let it manage you.

Peter

 

[rangerdanger12]

I'll answer myself: because casino has higher odds of losing, even though it has higher odds of winning.

That makes no sense because net net you still have a 17% advantage in my scenario.

IF there was a game in a casino that had a 51% odds of winning you would have a 1% edge over the house and if you played all the tables with that 51% edge you will be positive over the long run.

You play one table and go all in then you can loose your shirt and never get another chance at any other table

 

[malaguti]

What price does on the next tick is anyone's guess all day long across thousands of stocks, commodities, currencies, financials, etc... You can't use that statement alone to justify patterns are no good.
Manage your trade, don't let it manage you.

Peter

I'm with Pete, whether we like it or not, all our trades are in some way a derivative of past behaviour and patterns and the likelihood of that patter reoccurring