12/02/2006

Lessons from best and worst performing stocks

The Tickersense churns out lots of statistics on daily basis. Many times it is focused on finding seasonality or correlations or historic tendencies. Today they have a interesting statistics about the components of Russell 3000 over the past 5 days, the past 3 months, and year to date.




Now If you look closely at it, you can clearly see that the best performing stocks over worst performing stock ratio is 2 or more. You can run this kind of scans on last several years (which will represent both the bull and bear markets) and you will find the same thing in top 10 best and worst performing stocks. Even if you run it during the worst bear year the 10 best and worst performing stocks will still favor long side. That is the way the market works.

The implication is clear, put your heart and soul in to developing a long strategy. You have overwhelming advantage on long side.Second implication is monitor stocks up 100% or more in a year, if you miss the first 100% there is still opportunity to capture some part of rest of the move. Currently there are around 300 plus stocks up 100% or more and 70 stocks out of that up 200% plus and around 20 stocks up 300% or more for the year. Even if you just concentrate on buying 5% plus breakout on high volume( 2 times 50 day average volume) on the 100% list and cut your losses judiciously, you should outperform the market by wide margin.

The most important implication is be wary of bears and perma bears. They simply do not understand the markets. They just want to trade their beliefs and not interested in making money. As I have said earlier a mediocre long strategy has a high probability of beating the best designed short strategy. So If you have to follow a guru, follow a bullish biased guru. Probability of you beating the street is extremely favorable.

Update :
Try this scan in TC2007 and lets compare notes after a quarter to see if buying stocks from this list with a risk management strategy(risking less than .5 of your account per trade) beats the market or not. Try it also against your favorite newsletter and see if it beats it or not.

100 * ((C + .01) - ( MINC270 + .01)) / (MINC270 + .01) >= 100 AND ( 100 * (C - C1) / C1) >= 5 AND V >= 2 * AVGV50.1

2 comments:

Michael Remmerde said...

Interesting...but is this a valid indicator of anything?

Nominal positive percentage gains don't linearly correlate to nominal negative percentage gains.

For example, if you bought NBIX and were down 85%, what percentange gainer do you have to have on the next stock to make that up?

The answer, of course, is not +85%. It's +567%.

That said, I do strongly agree that in the long term, being long will beat being short.

Pradeep Bonde said...

My point is different. If you were a short seller and picked the best 10 shorts for a given year you still have to look at what was the best opportunity to make money for that period. If the 10 best longs outperform the 10 best shorts by 2:1 or more, should you focus your attention on shorts. For any given period of time, the objective for a trader should be to try and capitalise the best money making opportunity, in a market in which short side profit is structurally capped while long is theoretically unlimited, what should one chose.