I bet that title got your attention! Although this is probably very good advice, it isn't the topic of this post. What I should have said is "Don't sell when you position is at a high". In this post I will demonstrate how you can get a little extra out of your trading system, just by following this advice.
Quant tools provided by Portfolio123.
This demonstration starts by simulating a simple Small Cap value system. First, set up a simulation with $10K trading capital, $10 commission per trade, and variable slippage. This system will rebalance weekly.
Set the position sizing to 10%. This means the port will typically hold 10 stocks at any given time.
Select the PRussell2000 as the stock universe. The PRussell2000 is Portfolio123's version of the Russell2000. The PRussell2000 will closely match the Russell2000 but there may be some minor differences.
Now select the Basic: Value ranking system. This is a standard value-based ranking system offered by the experts at Portfolio123.
The next step is to add the buy rules. The starting capital is fairly low so the minimum liquidity can be set to a low value without adversely affecting trades. In this case, the minimum liquidity is AvgDailyTot(65)>50000. In English this means that the stock must have traded $50000 average daily dollar-volume over the last three months. This tells us nothing about the liquidity for the next trading day but will help to avoid very low liquidity stocks.
The second rule Rank>99 filters out stocks that have a Basic: Value ranking of 99 (out of 100) or less. In other words, only buy the best of the best ranked stocks.
The sell rule is Rank<99. In other words, sell when the stock's ranking drops below 99 (out of 100). It is no longer the best of the best.
Set the time period to the maximum that your membership will allow and click on RUN SIMULATION.
Your results will be quite good for minimal effort, with an annualized return of 16.43%
Average profit of 4.03%
Now lets get to the point of this post. What happens if you don't sell your position when it is trading at a high? Lets find out. GO back and re-run the simulation. This time modify the sell rule to: Rank < 99 & PctFromHi < -1.
In English this rule says that the system will only sell the stock if the ranking drops below 99 AND the stock price is at least one percent from its high (since you owned the stock).
Now re-run the simulation.
The annualized return has jumped from 16.43% to 19.78%.
The average profit has jumped from 4.03% to 6.78%.
Now lets modify the sell rule again, this time to: Rank<99 & PctFromHi<-10.
The annualized return is now 22.02%, up from the original 16.43%
The average profit per trade is now 9.28%, up from the original 4.03%.
This is just a demonstration and is not offered as proof. However, this has been a favorite trick of mine for several years and still appears to work.
This demonstration was done with the benefit of hindsight, and may not result in future profits.