All,
Budonk started a post asking about the best Sell Rules. I have wanted to summarize my buy and sell rules so here goes.
My basic logic is to maximize Annual Return. For me the power of compounding has to dominate my approach. Giving up a few % Annual Return for a reduced drawdown becomes a big loss (lack of gain) over a 10 year period. I put little actual concern on drawdown. However, it is difficult to achieve a consistent annual gain if the approach continues to have high drawdowns. So my lower drawdowns that I achieve are not a goal, but a byproduct of chasing consistent annual return.
When I say I try to maximize my annual return, I am not speaking about the Annual Return on the Summery page. What I am referring to is maximizing the daily return of each stock that I buy. If I am willing to risk cash to buy a stock then it needs to meet the highest standards that I can find. For this I use the factor Gain/Stk/Day or the average gain on each stock I buy for each day that I hold the stock. To calculate this factor go to the Statistics; Trading page and divide the Average Return for the Realized Trades by the Average Days Held. This gives you a value of the % gain per stock traded per day that they were held. Or what I call the Gain/Stock/Day. If I can optimize this one parameter then I have optimized my annual gain relative to my risk of buying stocks in the first place.
Although I don’t try to time the market, I set up my Buy Rules such that if the market is in a decline the Sim will find very few stocks to buy. To do this I use restrictive Buy Rules. I limit the stocks I buy to those that will have the highest probability of good gains in the shortest period of time.
I use the Buy and Sell Rules below in almost all of my Ports. I use additional Rules in most of my Ports that are tied to the Ranking System’s approach.
Rank > 99
Since the highest ranked stocks, on average, out perform lower ranked stocks, why settle for stocks that, on average, will have lower performance. I’ll use Rank > 98 occasionally if it improves the Gain/Stock/Day. If there are not any stocks that meet this criterion and the other buy rules then I will either stay in cash or use the cash in another Port that finds stocks that meet all of its rules. I will not reduce the rank value just to find a stock to buy.
Price > 1
Lower priced stocks, on average, out perform higher priced stocks. So why would I want to restrict my buys from some of the best returns with a Price > 5 rule. However, Penny stock’s prices are all over the place and are very difficult to pick winners consistently.
MktCap > xx & MktCap < xxx
Smaller Cap stocks, on average, out perform larger Cap stocks. There are very few billion dollar companies that can double in price in a few months. However, in any giver year there are dozens of small and Micro Cap stocks that do it. I use; MktCap > 50 & MktCap < 500, for most of my ports, and MktCap > 100 & MktCap < 1000 for a few Ports that still show high performance with the higher Mkt Cap maximum.
AvgDailyTot(20) > xxxxxx
Set the min value for the smallest range that you can buy with the dollar amount that you need to spend on each stock. I need to buy a minimum of $10,000 per stock so I set it at 10 times that; AvgDailyTot(20) > 100000. Although this level, only 10 times what I am buying, will effect the buy and sell price somewhat, I find that overall I do better than if I only bought stocks that had an AvgDailyTot(20) > 200000, or 300000.
Pr4W%ChgInd > 0
This is a very restrictive rule. If the industry that a stock is in has not increased in value over the last 4 weeks I don’t want to buy it. This rule is very helpful during market corrections because it only allows the Port to buy stocks in industries that are not correcting. This rule, combined with the ones above, will frequently, during market corrections, prevent the Port from remaining fully invested. However, it improves my % winners by avoiding stocks in losing industries. This is my way of timing the market automatically. In my most aggressive Ports I will use; Pr4W%ChgInd > 5.
The Sell Rules I use are complementary to the buy rules and my desire to maximize my Gain/Stk/Day.
Rank < 99
If I don’t want to buy a stock that has a lower probability of gain why should I keep a stock that has a lower probability of gain, especially a stock that has declined in rank.
With some of my Ports I will use a Sell Rank value that is 1 less than the Buy Rank value if it improves the Gain/Stk/Day.
PctFromHi < -xx
I use a trailing % stop loss with all my real trades. I lost way too much money many years ago in Micro Caps that tanked after bad news. If you are going to trade Micro Caps I feel that a stop loss on your real trades is mandatory. Therefore, I use a stop loss Sell Rule in all my Ports. With the proper stop loss value you can normally increase Gain/Stk/Day and reduce the drawdown. I will use 20% in most of my Ports, 15% on the ports that trade the largest cap stocks, and 30 % on the Ports that trade the smallest of the Micro Caps.
GainPct < BenchPct & NoDays > 30
I use a Laggard rule in most of my Ports. 60 to 70 % of the highest Ranked stocks will outperform the Benchmark. Since the Benchmark performance is the average of the stocks that it benchmarks, and I am trying to maximize my annual performance, why would I keep a stock that can’t even do as well as the Benchmark? Sell it and buy another one. I have a 60 to 70 % chance of the next stock I buy out performing the Benchmark.
Pr4W%Chg /20 < 0.x & NoDays > 20
This is an aggressive sell rule. It requires a stock to maintain a minimum consistent gain over a rolling 4 week period. This will cause the port to sell a stock if it doesn’t meet a minimum performance criterion that the rest of the stocks in the Sim meet. I base the number I use for the x in the rule above on the Sim’s Gain/Stock/Day. This will be a number between 0.1 and 0.5 (hopefully), the higher the better. If the number is 0.2 I will generally use about 0.1 for the x in the sell rule above. In other words if a stock can’t do at least as well as ½ the average gain then I sell it and try another one.
This approach results in a higher turnover than most investors think is acceptable. However, it only takes me about 2 hours a week on Saturday or Sunday night to Rebalance my 10 Ports, review the stocks, and set up my trades. I trade on average 4 stocks a week out of 25 to 30 stocks total. I execute the trades first thing Monday morning.
The above buy and Sell Rules are based on my logic for the best way to achieve the highest Gain/Stock/Day. This approach is what results in my high returns in the P123 Weekly Performance email.
Good Trading,
Denny 
