Tuesday, January 3, 2012

10,000 Charts

They say that it takes 10,000 repetitions to become proficient at anything, be it Tiger Woods (before his fall) on the golf course, or Venus and Serena Williams on the tennis court. Or my grand-daughter taking her first of many thousands of steps.

When Woods or the Williams sisters do their thing, they don't think about it. I mean, Tiger doesn't recite in his mind the steps of hitting a golf ball, nor do Venus and Serena silently rehearse the elements of a swing. Each acts from an unconscious mind that has been honed by repetition into an unseen expert in the sport.

David Brooks, in his excellent book The Social Animal: The Hidden Sources of Love, Character, and Achievement, presents some recent findings of scientists studying the unconscious mind.

The research finds that the unconscious mind works by finding patterns, and that it is far more powerful than our conscious minds, can handle complex tasks without any help from consciousness, works faster than the conscious mind, and produces fuzzy results, often from fuzzy input.

Brooks writes: "Conscious processes are better at solving problems when the factors are concretely defined. Unconscious processes are better when everything is ambiguous."

What does this mean for traders like us?

Ambiguity is soul of the markets. When a trader looks at stock chart with the goal of picking out the trend, he or she is confronted not by a single trend, but by a barrage of contradictory possibilities. Any market at any given time is in many bull phases and bear phases, simultaneously.

Traders deal with this by using technical tools, like moving average crosses or the relative strength index, to produce yes-or-no -- binary -- results, reducing ambiguity by moving away from the reality of the price movements.

But I've concluded that such technical tools generally produce bad trades. (See my essay "Technical Trading Doesn't Work".) And certainly their clearly defined, binary results do little to engage the powerful unconscious mind of the chart-reader. A bare-bones chart, by contrast, gives full scope to the unconscious mind's skill in pattern recognition -- patterns on such a chart are all it has to work with.

So I propose an alternative method that will 1) ensure that the trader has analyzed 10,000 charts in less than a year and half, and 2) engage the unconscious mind with the goal of developing the ability to judge a chart in a glance as a single gestalt of the current reality of the market.

I do this by running a 16-stock bracket, much like the NCAA Men's Division I basketball championship, a single-elimination tournament.

Here are the steps.

1) Form an opinion of the broad market's direction. 

This can be done by listening to market commentators, or analyzing the S&P 500 chart, or for that matter flipping a coin.

2) Select a universe of stocks from which the bracket will be seeded.

I generally use a scan of stocks ranked by Zacks, choosing stocks expected to outperform if I'm bullish on the market as whole, or under-perform if I'm bearish. But I could as well use any method, or simply use a grab-bag of stocks selected using no method, except volume to assure liquidity.

I generally go for 60 stocks or more, and I want them to be in a form that I can copy and paste their symbols into a vertical list in a spreadsheet.

3) Select 16 stocks at random.

Get 16 random numbers in the range from the first item in your list (generally #1, or #2 if the first row is a label) to the last item.

The website randomizer.org is a great place to do this, and it's free. The results will look like this:

       5, 11, 49, 2, 10, 65, 56, 58, 25, 7, 59, 27, 39, 67, 50, 3

4) Associate each random number with a stock symbol.

Looking at your list, put the number 1 beside the 5th item, 2 beside the 11th item, 3 beside the 49th item, and so forth to 16 beside the 3rd item. (It helps to sort the random numbers into numerical order, to save jumping back and forth in the spreadsheet.)

5) Seed the bracket.

Put the 16 selected symbols in the first column of the bracket. The bracket I did today looked like this:






6) Disengage the conscious mind.

Remember, one goal of this method is to bring the unconscious mind into play, to look at and judge the charts without the inner narration that accompanies us most of our lives.

I do this by directing my attention elsewhere. I'll plug my iPhone into my ear and play an episode of NPR's "Planet Money" or "Freakonomics" and concentrate on that while judging the charts. My conscious mind is listening to the program, so my unconscious mind takes on the task of judging charts.

For some reason, White House briefings by Jay Carney on C-SPAN seem to be especially effective.

7) Compare charts.

What follows is just like the basketball tournament, except rather than games I'm using bare-bones stock charts, containing only the daily prices (I use candlestick charts myself, but any kind will do) and the daily volume.

My market opinion was bullish in the bracket above, so I was looking for the most bullish chart in each case.

Compare each pair of charts from the seed column, bringing the most bullish into the quarter-final column. Repeat, and move the results into the semi-final column, and so forth until there's a winner.

In the bracket above, I compared DFS and LULU, and moved DFS the next column. And then I compared SE and COH, and moved SE to the next column. And so forth down to CREE and ASML.

If the market opinion is bearish, then move forward the most bearish chart of each comparison.

Once I have a winner, I go back to the semi-final column, and compare the two losers, which provides 3rd and 4th place for my ranking. The losing finalist is 2nd place, and of course the winner is 1st place.

Why worry about anyone but the winner? Sometimes the winner can be disqualified, such as by an earnings release scheduled the same week, or by bad news that casts doubt on the company's quality, or by a poor selection of options for the winning symbol.

In that case, then I'll go down to the 2nd place finisher, or further down to the 3rd or 4th if necessary.

8) Get conscious.

Once I've picked a winner, I unplug my iPhone.

First, I immediately write down why I chose the bracket's winner instead of the runner-up. By the time a bracket reaches the finals, the charts tend to be very similar. It's always a tough choice.

By writing down an ex post facto justification, I gain skill in determining what to look for in a chart.

Then I start looking at the chart consciously, narrating why I should or should not trade this stock, drawing trend lines and channels on the chart, looking at company finances, slapping up Fibonacci retracement ladders, even bringing in binary technical tools just to understand where the stock really stands.

Based on that conscious analysis, I'll make the final decision whether or not to take the trade.

Altogether, each bracket gives the trader an opportunity to do 31 chart comparisons, picking the 'best" chart in each case. At that rate, it will take 323 days to reach 10,000 charts.

Stocks trade 252 days a year on the American exchanges, So it will take one year and 71 days, a couple of weeks over 15 months, to reach the goal.

But why a bracket? Why not just pick 31 stocks and run down the list, picking out the best chart of them all?

I find that my mind, especially when distracted, as I've proposed above, works better when choosing from a limited field: This one is better than that one.

And I think we see that every day in our daily lives, where we tend to compress our choices into this-or-that, or possibly a trilateral this-that-or-the-other. People rarely try to choose the best out of a dozen choices without somehow organizing the selection into intermediate choices.

Anyhow, that's the method. It can be a training method, with paper trades, or an operating method, with real trades. I use real trades, because I find that I respond differently when I have real money on the table.

There are always far too many stocks for a trader to analyze in any one day, especially without binary technical indicators, so I find that a random sample from a universe that is biased in the direction I'm interested in provides a workable process.

Finally, I always look back to see how my choices fared a week or so down the line. That provides the feedback to improve my ability at chart reading, much as the young Tiger Woods, the novice sisters Serena and Venus Williams, and the Oregon Ducks football team have relied on post-mortems of their contests to see what they did right or wrong.

No comments:

Post a Comment