Instead of using simple ROC or a average two period ROC we can use a weighted ROC. In weighted ROC, ROC's for different periods are added together and assigned weights. So for example you can add ROC of 1 year and 3 month ROC. One of the reason to do this is that recent period price gains or loss get additional importance.
ROC 3 month+ ROC1 year.
Now you can vary the importance you give to the periods. So you can give say double the weight to recent price action. That will make the above formula
2*ROC 3 month+ROC 1 year
Investor's Business Daily (IBD) uses a front weighted ROC to calculate its Relative Stregth ranking. It adds up the ROC for a quarter, 2 quarter, 3 quarter and a year and gives double the weight to recent quarter.
IBD advise focusing on top 20% stocks using above ranking or stocks with 80 plus momentum ranking which it call Relative Strength ranking. On those 20% stock it uses other fundamental factors to narrow down the candidates.
On that narrowed down list it uses chart patterns to time entry.
What does a weighted ROC like this does is it gives you stock with trending characteristics. They have high momentum on all four time frames. When you buy a continuation breakout on such stocks , both probability exists:
- it will follow through and you will have very profitable trade
- it will not follow through and you possibly entered a long running trend in stock at a time where it is about to revert
In order to avoid the second probability IBD also advises use of another criteria, it says buy breakouts after first or second base and avoid third stage bases.
Mark Boucher also uses front weighted ROC in his trading approach.
Mark Boucher in his book details his momentum based system for trading the world markets, the commodities, and the stocks. At the heart of his equity selection method is momentum. He use a weighted average ROC to rank stocks. He offers two methods for equity selection based on whether you are long term traders or short term traders.
For Long Term Traders
( (2 * C * 100 / C5) + (2 * C * 100 / C25) + (2 * C * 100 / C40) + (2 * C * 100 / C65) + (C * 100 / C130) + (C * 100 / C195) + (C * 100 / C260)) / 11
He advises taking top 20% stocks ranked by this kind of ROC scan.
For Short Term Traders
( (2 * C * 100 / C5) + (2 * C * 100 / C25) + ( C * 100 / C40) ) / 5
He advises focusing on top 20% stocks by this kind of ROC scan. After first using momentum to narrow the universe of stocks, he uses earnings and other fundamental criteria to narrow this universe. Within that narrower universe he uses a Runaway Pattern criteria "TBBLBG" , which stands for Thrust Breakout (TB), Breakaway Lap (BL), and Breakaway Gap (BG).
If you look at his momentum scan closely, he uses a front weighted ROC. A scan like above gives you stocks which already are in rally mode and then you enter them based on chart patterns.
The exact period used in such calculation and weightage used is a matter of trading time frames used. If you are say a day trader, you would use a minutes or hours time-frame and weigh them differently based on whether you want to fade the trend or ride the trend.
If you are primarily a pullback trader, you can use a back weighted ROC to find stocks which are top ranked by one year momentum but are currently having a pullback.
| Momentum trading using Back weighted ROC |
Depending on your objective and holding periods, you can create various kinds of weighted ROC and test their effectiveness. One of the observed tendency in stock price momentum is that short term momentum (less than 3 month) is bad, it often leads to mean reversion. So a back weighted ROC can be designed to find such stocks.
Ford Equity Research uses one such kind of weighted ROC. It is one of the well known equity research provider on the street and has been using a price momentum model since 1991. It uses the price momentum (which it calls PRM) figure to rank stocks and advocates buying top 10% ranked stocks. It is calculated as follow: PRM =PGY - PGQ - 3 * PGN PRM Price Momentum PGN (past month price gain) PGQ ( 3-month price gain) PGY (12-month price gain) So it gives negative weights to 3 month and 1 month price growth. A Weighted ROC calculated like this will give you stocks with high relative strength which have pulled back in last 3 months and especially in last one month. So this is an example of back weightedROC , where there is more emphasis on past one year price growth but less on short term price growth. This kind of approach is suitable for institutional investors who like to buy and accumulate shares during correction or pullback traders.
There are limitless possibilities to using momentum. You can create various combination of time periods and weightage. Some combination work better than others. The combination you use is a function of your objective and risk tolerance and frequency of trading. |
Posts
Post a Comment