As an active trader, I manage risk largely through position sizing. In other words, my "casino money" comprises 10 % or less of my total portfolio. The remainder of my portfolio is very conservatively managed.

In the past, position investing served me well. During the early 90's, you could take a buy and hold position and be reasonably well assured it would perform well over the long run. This approach does not do so well today, as uncertainty in market environments throughout the world has created a choppy investing environment. Even the so-called "safe haven" vehicles like gold and bonds require constant vigil today.

Nevertheless, I am constantly searching for a relatively safe haven to park some passive money. Recently I have been experimenting with a mindless system that will automate a passive portfolio -- all I would have to do is just wait for a signal and buy an equity when I get the signal.

As a day trader that actively uses candlestick and chart patterns for my entries, I have become interested in the morningstar candlestick pattern. According to Thomas Bulkowski ( the Pattern Site ), the morningstar is a 78 percent probability bullish reversal pattern that ranks number 12 out of 103 patterns in performance. The downside is that it ranks 66 out of 103 candlestick patterns in frequency, so it is not a daily occurrence.

I chose to back test a morningstar swing trade strategy over a 9 year period using a high return on equity (ROE), low price earnings to growth (PEG) screen sample from Finviz. The screen produced 44 US stocks for my sample base. The rationale was to pick under priced, high performers which theoretically should show resilience over time. My controls were the S&P 500 index, and the top largest market cap equities in the S&P Index. I also set up exit parameters for an absolute stop, profit targets, and trailing stops which would kick in after a target was met.High return on equity, low PEG stocks were chosen on the theory that they will spring back with resiliency.

This Finviz screen kicks out 44 US equities.

My back testing was accomplished through the Zignals web site (Zignals). This site has recently added (beta) a strategy engine which will let you back test as far back as the early 2000's various strategies based upon a slew of technical and fundamental parameters. The strategy tester does have some quirks -- for example, the trailing stop is not real-time intra day. Even though I receive an email alert intra day when the trail stop is hit, the actual sell will not happen until end of day. Also, the data is static - the 44 companies stay the same even though over time you may have some net additions or subtractions.

So if I go live with this scenario (which I plan to), I would probably manually execute my stops and I would revise the equities as the dynamics change.

The results of my study were impressive, as shown below.

As you can see, the high ROE low PEG morning star group performed positively and consistently over the entire period that the S&P DECLINED -- plus 170% versus minus 16 %. The other morning star control group, the top 50 cap S&P equities, also performed positively, although not as well as the high performing ROE sample.

So I have developed and published a morning star trading strategy using the high ROE base of 44 stocks on Zignals that I am going to go live with using the following parameters:
 

I chose $35,000 as my available capital and a position size of $5,000. because I really don't need any more. With this strategy, I would likely have no more than 3 or 4 positions open at any one time. Of course, I could make the position size whatever I wanted.

I set a stop of 10% (yes, I know that sounds like a lot, but this pattern rarely fails, and this stop percentage gave me the best results). I also set a trailing stop of 5 % to kick in after a target level of 15 % was met. I set an outright sell target of 26%. I set a daily limit of 30 days to re-buy each equity to prevent whipsaw.

I back tested this strategy to 1/4/2005. The results, again, were impressive:

2005: 10.11%
2006: 32.87%
2007: 28.31%
2008: 0%
2009: 25.00%
Total Period: 137.66 %

 Here's an example of two trades on the same equity, CBI that occured  in this strategy:

A couple of notes on this strategy:

It is slow, meaning you could go weeks without a buy or sell. But this is OK for a passive, high return approach to swing trading.

The morning star pattern, while very reliable, does and will break down occasionally. It is important to note that no pattern will hold up during a general market sell off.

The beauty of this strategy is that if the market does correct, I will be mostly in cash and will have a number of opportunities that will come to me on the next upswing in the market.

I'll post my trades when I start this strategy.