Message Font: Serif | Sans-Serif
No. of Recommendations: 4
What is the excel formula for RRS?

This is the formula I use, It has slightly different values than The etf screener but close enough.

D is price
A is Date of price (0-252)

where can I find info on that variable?

Definition of Screen Structures:

DHxx This refers to Kentstutzman's pre-filtering of "dropping the high xx" performers over the last week. Macros here implement this
by measuring total return over the last 5 trading days and then excluding the highest performers. "xx" is used as a percentage, where
"DH10" means to drop the highest 10% of performers over the last 5 trading days. For 100 T1 stocks, this is the same as
dropping the top 10.

This filter is applied before sorting by relative strength, as described below.

RRSxx "RRS" refers to regression relative strength; "xx" refers to the number of trading days used in the regression. This approach regresses
the daily log-prices against trading days to determine the slope of the regression line for the log-prices.
This approach essentially changes the definition of the word "strength" in relative strength. Traditionally, relative strength is determined
by calculating total return over a historical period, then sorting by that measure for a set of stocks. RRS finds the regression line slope
for each stock and then sorts by this measure. Backtests indicate that this approach may be better at identifying stocks to hold
for the next month (compared to using total return to measure relative strength).

XGxx "XG" refers to exponential growth, a term introduced by LorenCobb. This indicates that the traditional relative strength approach,
using total return over "xx" days, is used to sort stocks. "RSxx" or "TRxx" could have been used here, but "XG" is used to make it notably
different from the alternative approach "RRS" and to be consistent with notation developed for LorenCobb's screens.

Sigma (volatility adjustment) This indicates that relative strength (total return or regression line slope) is adjusted for volatility before being used to sort for the
top stocks. This concept was introduced by LorenCobb, where the "-2 Sigma" approach is labeled Risk Averse.
The DataEntry_Results tab allows the user to specifiy the level of risk aversion desired by changing the "Sigma" parameter.

Lag This refers to measuring relative strength on a day before the "current" day (through which historical data is available). Robbie Geary's
(rgearyiii's) backtests show that using lagged relative strength can improve results for the shorter term lookback periods. An example
is: Lag5 / RRS42, which ranks stocks by the 42-day regression relative strength of each stock as of the close 5 days ago.

GTD This refers to a screen approach labeled "Grab The Dogs". It is also referred to as a "Lag/GTD" approach because it implicitly uses
a lagged relative strength measure as a first step. An example best explains the approach: Lag5 / RRS42 / GTD10 means the following steps
are used: Stocks are ranked by 42-day regression relative strength as of 5 days ago; the top 10 stocks are then sorted ascending by the
total return over the last 5 days. The top n (e.g., 5) stocks are then selected.

References Concepts implemented here are taken from The Motley Fool's discussion boards, specifically the Mechanical Investing Board
in the Investor's Roundtable section. You can find these posts by starting at the following URL and entering the desired
post number in the box to the right of "Number:" in the upper right of the message area.

Relevant posts include:

# Author Description
83661 BarryDTO Defines the "Regression Relative Strength" approach and shows an historical analysis of returns
for 1984 - 1998. Also provides pointers to earlier posts on analyses using the daily data provided by

87544 BarryDTO Presents comprehensive 1984 - 1998 backtests of various screens that combine the "RRS", "DH" and volatility
adjustment approaches.

89970 BarryDTO Examines screens with larger risk aversion factors (from -1 to -7). Demonstrates that these "annual"
risk aversion factors can be viewed as "monthly" risk aversion factors ranging from -0.3 to -7.0.
Also shows that for shorter-term RRS screens, screens with higher risk aversion factors (-5 on an annual basis)
show historical results with improved reward to risk profiles.

102113 rgearyiii Presents the results of using the GTD approach for a 20-day lookback. Provides a table of links to Robbie's
prior posts, which explain his notation and show other significant results. Robbie's backtests provide
a very thorough roadmap of historical performance of various screens, showing results averaged for all
start days and also showing results for small changes in individual parameters (e.g., lag days, DHxx, etc.).

85421 Kentstutzman Presents the concept and detailed analysis of dropping the recent 1-week high performers. It is presented
specifically in conjunction with a 13-week relative strength screen.

70943 LorenCobb These are the initial posts by LorenCobb describing his framework for blending historical return and volatility
71176 data to select stocks according to a desired level of risk aversion. You can read more updated material at his web

78518 Sux2BeU Provides a description of and pointer to the daily stock return history for 1984 - 1998 that serves as the basis
for the 15-year backtests I have done on the various screen combinations that can be used by this workbook.
Thank you Sux2BeU!

Print the post  


When Life Gives You Lemons
We all have had hardships and made poor decisions. The important thing is how we respond and grow. Read the story of a Fool who started from nothing, and looks to gain everything.
Contact Us
Contact Customer Service and other Fool departments here.
Work for Fools?
Winner of the Washingtonian great places to work, and Glassdoor #1 Company to Work For 2015! Have access to all of TMF's online and email products for FREE, and be paid for your contributions to TMF! Click the link and start your Fool career.