4/03/2007

Why choice is a big problem

No matter the time frame you trade, one of the biggest or the most critical problem when trading the stock markets is one of choice. You have around 10000 stocks to chose from. There are now over 375 ETF to choose from. So the ETF investor still has to grapple with choice problem.

Unlike the commodities market where the number of tradable futures are limited, in the stock market the wide choice creates a problem. That is also one of the reason some of the things which work in futures and commodities market do not work in stock market.

Most methods will give you too many stocks to chose from. At the end of the day a method like 100% plus or Episodic Pivot still gives you too much choice. You still need to decide to take just 1 or 2 trade out of the list with best potential.

While entry, exit, risk management, equity selection is critical, the first and foremost task for developing a strategy is to reduce the choice. Choice creates complexities. Choice creates sub-optimization. The choice in number of equities is just one problem. The other problem is choice of style, choice of time frame, choice of entry techniques, choice of exit techniques and so on.

Every choice which you make has flip side to it in such case. The other choices may be better. More than that the most important thing is one must chose the best method which will give you best return for a defined period. In the sense for a given time frame of say a year, for a given capital base and risk and number of trades, if the maximum possible opportunity in the market is to make say X%, anyone making less than X is sub optimizing his or her effort.

So choosing the variable to reduce your choice becomes very critical. One of the simplest choice reduction technique used by most traders is by using either volume or price as a cut off. Now that choice reduction method is inherently flawed, because it is not based on potential opportunity. Stock price growth is not a function of starting price. Similarly your need for liquidity and the market opportunities are two different concept altogether. Also one should realistically answer the question for his own trading capital base, how much liquidity do you really need.

The other choice reduction techniques are buying at or near 52 week high or low only. Again it is at best a choice reduction technique and not necessarily a valid indicator of opportunity in the market. Same logic applies to those promoting buying only all time high. While buying all time high reduces your choice, in fact if you test the validity of the idea of buying all time high only, then you will notice, it under performs the market. Same way using relative strength to narrow your universe again reduces your choice, but at the same time it has its own limitation and it again does not necessarily solve your problem of too much choice.

Valuation and financial factors can also be used to reduce choice. Measures like earnings, earnings momentum, P/E, P/S, book values are all used to reduce the choice universe. A combination method like the one used in CANSLIM is primarily aimed at choice reduction.

Traders who just trade one instrument like market makers have eliminated the choice problem. That is why for shorter time frames many traders just reduce their choice to few stocks and periodically change their choice.

The other choice problem, I have frequently seen is traders using a Chinese buffet approach and trading a variety of time frames and techniques. It gives them an illusion of lot of diversification, but may not improve results. Reducing number of traded methods, scans and time frames is one of the simplest way to improve your profit in my experience.

Choice reduction requires careful strategy. Choice reduction is one of the main goals at least in my scheme of things. My ultimate dream is to fine tune the selection procedure to such an extent that one takes only 10 trades in a year, but those 10 trades are in the best performing stocks for the year and one trades only one method. The one which is the most profitable.

8 comments:

ybn said...

Pradeep,

You hit the hammer right on the nail. I have seen very few articles about this problem. I use just about all the techniques you mentioned in the article. Price, volume, checking IBD Daily Graphs for fundamentals, and carefully studying the chart pattern. With all this work it still seems that the stock I didn't choose always performs better. After carefully think about this problem I came to the conclusion that this is the same problem you have when you drive on a 4 lane freeway during rush hour. The other lane always seems to move faster then yours. The reason is obvious. You only have 1 in 4 chance of being in the fastest lane at any given time. The same thing works with stock selection. If you choose one out 20 good ones, your chance of being number one, even after careful elimination, is still only 1 in 20.

Pradeep Bonde said...

1 in 20 is still good. I have thought about it for sometimes and over period of time improved the odds a lot. It is bigger problem if you trade longer time frame because in reality one is taking say first few signals, gets fully invested, now unless you get stopped out you are going to let other signals go , which in many cases might have higher probability. When you want again to pick a trade because one of the long term trade has closed, you are essentially taking what is available. Frequent re balancing is one of the solution, but it increases transaction cost.

MissTrade said...

Choices and yet, still establishing how much risk vs the possible reward is as important in my eyes. You can be wrong over and over if the risk is less than the reward potential. Establish it, use game theory to manage it. You will win some races and lose some, but the winners should outweight the losers all day. Give yourself the proper reward to risk scenarios and choices will be easier. Undertrade should be your goal.

Pradeep Bonde said...

The issue is slightly different. It is about for a given risk and time frame how to maximize results with a already profitable system. Finding profitable trades is not a problem. Finding several profitable trades is also not a problem. The problem I am talking about is of prioritizing within several possible profitable trades.It is about even better performance or the best possible performance for given risk and time frame.

Bryan Matthews said...

pradeep,

10 trades a year would be nirvana in the trading world. That's a great goal to shoot for.

Pradeep Bonde said...

Joseph
Those are the formula for Episodic Pivot conditions.

patterndaytrader said...

Hi Pradeep,

In my opinion, pattern recognition plays a critical role in consistently weeding out the best performers from a pool of profitable ideas. If you look at documentary by the greats like William O'Neil, Dan Zanger and Mark Minervini for instance; you will attribute the bulk of their out performance to the ability to recognize patterns and assess the quality of patterns and price volume action especially within the context of the price action of the overall indexes themselves!

patterndaytrader said...

Here is an example of what I speak of with respect to 1) pattern recognition and 2) assessing patterns with in the context of the price action in the overall markets from Mark Minervini's blog: http://markminerviniblog.blogspot.com/2010/03/how-i-played-ascending-handle-in.html
In this post he speaks of an "ascending handle"...i.e. a pattern he looks for, very interesting stuff...