Trader's mashup gadget

Zack's Earnings and Margins screen

Zack's Investment Research has an area of their website devoted to stock screens. One of the screens that they are currently featuring is the Earnings and Margins screen. The screen looks for stocks with growing earnings, earnings above zero, a good Zacks rank, and a good broker rating.

Here's the parameters of the screen:
- The most recent completed year's earnings higher than the previous year's earnings.
- Net Margin above zero.
- Zacks Rank <= 2
- Average Broker Rating = 1.0

And here's the results (rebalanced every 4 weeks):
2008 (through 3/21/08): -13%
2007: 7.7%
2006: 22.4%
2005: 27%
2004: 11.1%
2003: 103%
2002: 29.1%
2001: 58.5%

It's interesting to note that this screen has always had a positive year until this year. This year is proving to be especially tough for investors and traders. Good luck out there!

Calendar market timing model (Halloween indicator)

Here is a well-known strategy for timing the stock market: Sell in May and buy in November. (AKA "Sell in May and go away" & the Halloween indicator)

In their paper entitled "The Halloween Indicator, 'Sell in May and Go Away': Another Puzzle", Ben Jacobsen and Sven Bouman document this pattern in 36 of 37 global markets studied. This is a very robust timing method that works well in most markets, although researchers are puzzled as to the reason.

Conservative investors could develop a strategy that would simply reduce and increase exposure to stocks via asset allocation in lieu of being 100% or 0% invested.

Novice traders might ask, "Well what happened this year?". The answer is that no strategy works all the time. Good strategies have been proven to work most of the time.

See also: Quant Investor: Calendar Based Market Timing Model

Collective2 trading systems

Collective2 is an interesting website that offers subscription trading systems. These systems are often technical in nature. Caveat emptor - anyone can submit a system, so there are a lot of subpar systems on there. Since you don't know the rules of the system, they are black box trading systems. Many systems I have noticed are probably data mined, because you can look at the chart and watch it fail when it operates in real time.

That said, the website lets you find systems several different ways. Once you go to the website, take a look at the options under "Find a system". There are quite a few good systems out there. Even if you don't want to subscribe to one, you can read the descriptions to get ideas for your own trading strategy.

Relative strength stock screen

Here's a stock screen from the Motley Fool Mechanical Investing boards which uses relative strength as it's sole criteria. This is a market crushing strategy for people with little time to trade, since it is rebalanced every 3 months. 30% annual return is not too shabby...

Actual screen code:
Define {RS13WKT12}
Uses [Timeliness Rank] [Total Return 13-Week]
Deblank [Timeliness Rank] [Total Return 13-Week]
Keep :[Timeliness Rank]<=2 Sort Descending [Total Return 13-Week] ; Top :10 End   Results from backtest.org from holding the top 10 stocks for 3 months at a time:

CAGR	30	12
GSD	38	16
Sharpe Ratio	0.83	0.54

Screen		S&P
1986		22	24
1987		17	6
1988		17	10
1989		37	35
1990		-2	-5
1991		107	31
1992		11	7
1993		31	10
1994		13	2
1995		48	39
1996		56	23
1997		-5	28
1998		37	35
1999		209	18
2000		21	-11
2001		-33	-9
2002		-1	-22
2003		114	28
2004		57	10
2005		12	8
2006		33	14
2007		11	6