Dancing with the Trend

Building a Rules-Based Trend Following Model - 13

Greg Morris

Greg Morris


I may seem to be wandering around when it comes to this series on model building, but I think that keeps the interest a little higher; maybe not.  Rules and guidelines are a critical element to a good trend following model; in fact, any type of model.  Once you have the weight of the evidence measures telling you what the market is currently doing, the rules and guidelines provide the necessary process on how to invest based upon those measures.  If there was a simple answer as to why they are necessary, it is to invoke an objective approach, one that does as much as possible to remove the frail human element in the model.  Rules are mandatory, while guidelines are not.  That being said, if a guideline is to be ignored, one needs to ensure there is ample supporting evidence to allow it.  Basically, the strategy I use is one of a conservative buyer and an aggressive seller.  After many decades in aviation and the always increasing use of checklists, the rules and guidelines are no different for maintaining a non-discretionary strategy than a checklist is for a pilot.  In aviation, checklists grew in length over time because as accidents or incidents happened a checklist item was created to help prevent it from happening in the future.  There is an old axiom about checklists that said behind every item on a checklist, there is a story.  Same philosophy goes for rules and guidelines in an investment strategy.  A checklist (rules) ensures portfolio managers follow all procedures precisely and unfailingly.  This overcomes the problem with experienced managers thinking they can accomplish the task and do not need any assistance.  That attitude is costly.  Below is a sampling of my rules and guidelines.


Buy Rules

No Buy Days are (1) FOMC announcement day, (2) First/Last day of calendar quarter, (3) Days in which the market has reduced hours.  FOMC announcement days are typically high volatility days and the end / beginning of a quarter involves a lot of window dressing.  Because I use a lot of breadth, on reduced hour days, you will still receive a full complement of breath data.

No buying unless 50 (this can also be a percentage) tradable ETFs (not counting non-correlated) have:

        Weight of the Evidence Color:  Red: Trend > 60, Orange: Trend > 55, Yellow/Green: Trend > 50

I call this the “soup on the shelf” rule.  If you have been to a large grocery store lately and strolled down the aisle that has soup, you probably noticed there are thousands of cans of soup with hundreds of blends, styles, etc. to choose from.  Now imagine your spouse has sent you to the store to buy soup.  When you turn down the soup aisle, you notice they are essentially empty except for two cans of rhubarb turnip barely in cream sauce.  You probably aren’t going to buy any soup that day.  The market is similar, especially during the initial stages of an uptrend, there just isn’t much to choose from.  In addition, the initial stages have stricter buying requirements so the number of issues to pick from could be very small, if any.

No buying on days when stops on current holdings are hit and assets sold.  This is usually the first hint that the ensuing uptrend is faltering.  It just doesn’t make sense as a trend follower to be buying on the same day as you are selling something that has hit its stop.  The argument that one holding might not be correlated is weak in this example as with proper trading up, weak holdings should have been traded.

No buying on days when the Nasdaq or S&P 500 is down greater than 1.0% (the indices used need to be tied you what you are using in the trend measures). Simply, this means that if the market as determined by the S&P 500 and / or Nasdaq Composite is down over 1% for the day, something is wrong with the uptrend and it is better to not buy that day.  An argument from bargain hunters or value investors would be that one would get a better price on that day if the uptrend resumed.  I can’t argue with that but I’m not a value investor or a bargain hunter.  It seems many investors want to buy stocks at bargain prices and I can understand that.  However, we are not buying soap at a discount store; we are buying a tradable investment vehicle whose price is determined by buyers and sellers.  Moreover, you only want to buy what is going up.

Sell Rules

If stops are hit with End of Day data and still in place at 30 minutes (this time period is based solely on your comfort level) after the open the next day, a sell is initiated; if not in place at the 30-minute point, the issue falls under intraday monitoring.

Intraday monitoring of Price and Trend between the hours of 45 minutes after the open until 60 minutes before the close, will invoke Sell order sent to brokers for execution.  Once an issue hits its stop, then a 30-minute period is allowed before it is sold.  With the constant barrage of internet and financial media trying to be first with breaking news, often the story is presented incorrectly, and it can influence a large stock, an industry, or even a sector and cause a big selloff.  Usually, if the story was reported in error or incorrectly, and then reported correctly, the issue quickly recovers.  Most of this happens in a very short period of time.  The 30-minute rule will help avoid most of these short-term selloffs with quick recoveries.

In broad-based sell off and stops are hit, holdings hitting stops can begin liquidating before the 30-minute limit.

Guidelines

Note: Guidelines are used as reminders and offer the opportunity to be ignored but only after considerable deliberation and examining all other possibilities.  The absolute most important Guidelines is the first one below.

In the event a situation arises in which there is not a rule or guideline; a conservative solution will be decided upon and implemented based upon immediate needs.  A new guideline or rule will be developed only after the event/conflict has totally passed.  This is a critically important guideline to ensure the “heat of the moment” is not used to create or change a rule.  The absolute worst time to create or change a rule is when you are emotionally concerned about something that just seems is not working correctly.

European ETFs need to be monitored closely after 1pm to ensure adequate execution time.  This is because when the Europe markets close, liquidity in those issues becomes a problem.

All buy candidates should be determined by:

        - Rising mandatory ranking components using a chart of the Ranking Measures, or the 5 day trend change.

        - An awareness of the issue’s price support and resistance levels.

Always be aware of the Prudent Man concept.  This is sort of a catch all to make one think about an action that has not been adequately covered with rules or guidelines.  If deciding to do something as far as asset commitment or ETF selection, one needs to be prepared to stand in front of the boss and explain it.

I also maintain a set of Trade Up rules.  These are the rules on when to sell a holding and replace it with another holding; one that is performing better.  When the weight of the evidence is Green (>80), then trading up is fairly active.   Once in Yellow (between 51 and 80), then trading up is limited considerably.

There are a host of additional rules and guidelines that can be created.  I would caution you on trying to develop a rule for every inconsistency or disappointment that surfaces while trading with a model.  There is probably a good equilibrium about the depth and number of rules that is best.  I strongly suggest adding rules rationally and unemotionally.

Tushar Chande has written another great article and this time offers you his Chande Trend Meter that is reasonably close to my Weight of the Evidence.  His only uses price whereas mine uses price and breath.  The similarity shows that measuring trend is the goal and there are many ways to accomplish it. 

Dance with the Trend,

Greg Morris

Greg Morris
About the author: has a 50-year investing career as a technical analyst, a developer of indicators and trading systems. He is also an accomplished author of books on trend analysis, breadth, and candlesticks. Morris worked with N-Squared Computing from 1982 to 1993. During his time there he produced over 15 technical analysis and charting software titles. Learn More