Sand2Pirls Market Signal
S2P Market Signal Historical Data
Note that time runs from bottom to top in these spreadsheets with the most recent periods at the top. If you are using Internet Explorer as your web browser and have Excel 2003, you can click the green X "Export to Excel" toolbar button below to save our historical market calls in Excel format to your system.
To identify the two signals generated from our one mathematical model, the 4.268 S2P Market Signal model refers to "Up" and "Dn" as "Directions" and "Long", "Short" and "Out" as "Trends." This means that for users of the two phase signal, "Dn" is Short and "Up" is Long. (Obviously, for users of the three phase signal, "Short" is Short, "Long" is Long and "Out" is Cash.) Running % in Market for the 3 phase signal is a simple percentage of days in the market with "Out" days removed.
You are encouraged to compare these signal dates with others and back-test them to develop or refine your trading method. See our Comparison page for a comparison of QQQQ Rate of Return using S2P 4.268 and seven (7) other market signals.
S2P Market Signal 4.268 Trend Periods (Excel 2003 Spreadsheets)
S2P Market Signal 4.268 Weekly Data (Excel 2003 Spreadsheet)
Note: The 'Direction' column is the 2-phase signal and the 'Trend' column is the 3-phase signal.
Why have we made changes to the signal calculations?
Models based on logical switches such as S2P Market Signal are fundamentally different from models based on "curve fitting" of moving averages in which parameters of the model are periodically "optimized" to provide the best fit.
The S2P Market Signal model is one that can grow as new conditions arise, but does not require periodic "optimization" by making changes to parameters such as the length of time for calculating moving averages.
Because S2P Market Signal was designed with logical space between conditional switches, when new conditions arise (either moving forward or backward in time), those new conditions can be added to the model as "logical switches" making the model more accurate as input data is increased. For instance, time-price relationships within each week were not analyzed until S2P 3. We found that adding linear regression analysis of Monday through Friday index close prices enabled us to more accurately detect the first stages of a trend.
The changes made with each S2P Market Signal version are listed below.
1/8/06 S2P 4.268 made a minor modification to linear regression analysis rules for non 5 day market weeks. The only resulting difference in historical data was the elimination of a 2-phase 'Up' signal call on 6/7/99 that 'whipsawed' back to 'Dn' the following week.
4/4/05 S2P 4.265 both two and three phase signals in a single model were made possible by increasing the responsiveness to trend directional changes by the use of a variable length look-back period. Also, a bullish bias in flat periods was removed.
7/18/04 S2P 3.2 improved trend shift detection accuracy by adding weekly price direction criteria to trend exit threshold.
7/5/04 S2P 3.1 improved accuracy of early trend detection by adding linear regression analysis of Monday through Friday index closing prices.
6/12/04 S2P 2.2 improved performance by adding a top detection algorithm.
5/16/04 S2P 2.0 improved performance by increased analysis of price enabling earlier trend detection.
5/7/04 S2P 1.1 added a bottom detection algorithm, making possible shorter and more profitable shorting periods.
5/3/04 We went live with S2P 1.0 when we saw that it out performed a fairly well known two-phase signal.