The Perl Toolchain Summit needs more sponsors. If your company depends on Perl, please support this very important event.
 / 
stockmonkey-2.9409
River stage zero No dependents
5 ++
/ Math::Business::ConnorRSI

NAME

Math::Business::ConnorRSI - Technical Analysis: Connor's 3 tuple average RSI with PriceRank

SYNOPSIS

  use Math::Business::ConnorRSI;

  my $rsi = Math::Business::ConnorRSI->new(3,2,100);

  # Equivalent set functions
  $rsi->set_cdays(3);
  $rsi->set_sdays(2);
  $rsi->set_pdays(100);

  # or to just use the recommended settings (3,2,100)
  my $rsi = Math::Business::RSI->recommended;

  my @closing_values = qw(
      3 4 4 5 6 5 6 5 5 5 5
      6 6 6 6 7 7 7 8 8 8 8
  );

  # choose one:
  $rsi->insert( @closing_values );
  $rsi->insert( $_ ) for @closing_values;

  if( defined(my $q = $rsi->query) ) {
      print "CRSI: $q.\n";

  } else {
      print "CRSI: n/a.\n";
  }

RESEARCHER

The RSI was designed by Connors Research, LLC 2012

Connor desired to combine information about price momentum (via traditional RSI) relative price magnitude information (called PriceRank) and information about price rate streaks (an RSI of the streak information).

The way he has constructed it, the informations are all on the same scale (0 through 100), and he then combines them in the natural way: a mean.

   $close_rsi = RSI(close,3); # nothing unusual about this
   # this is the first parameter, or set_cdays()

Streaks are ore confusing. They are constructed as follows.

    if( $close_yesterday < $close_today ) {
        $streak = $streak >= 0 ? -1 : $streak-1;

    } elsif( $close_yesterday > $close_today ) {
        $streak = $streak <= 0 ? 1 : $streak+1;

    } else {
        $streak = 0;
    }

Such that, if our closing values are 20, 20.5, 20.75, 19.75, 19.5... our state would be: undef, 1, 2, -1, -2. This streak stream is then fed to another RSI object:

   $streak_rsi = RSI(streak,2);
   # this size is set via set_sdays()

Lastly, his PriceRank value is determined as a percent winning rate of change. Say we have an array like so:

    @items = (
      (20.5-20)/20,
      (20.75-20.5)/20.5,
      (19.75-20.75)/20.75,
      (19.5-19.75)/19.75,
    );

Suppose we get a closing value today of 20. Then the price rank would be computed as follows:

    $today_rank = (20-19.75)/19.75;
    $items_beat = 0;
    for(@items) {
        $items_beat ++ if $today_rank > $_;
    }

    $PriceRank = 100 * ( $items_beat / @items );

Finally, we compute CRSI as the average of these three items:

    $CRSI = ( $PriceRank + $close_rsi + $streak_rsi ) / 3;

The interpretation is rather similar to the standard RSI. If the CRSI is above some threshold, it is considered overbought. If it is below some threshold, then it is oversold. He states that this method adapts very quickly and has a high level of accuracy. His original paper has many charts to suport this conclusion.

THANKS

Robby Oliver <robbykaty@gmail.com>

AUTHOR

Paul Miller <jettero@cpan.org>

I am using this software in my own projects... If you find bugs, please please please let me know. There is a mailing list with very light traffic that you might want to join: http://groups.google.com/group/stockmonkey/.

COPYRIGHT

Copyright © 2013 Paul Miller

LICENSE

This is released under the Artistic License. See perlartistic.

SEE ALSO

perl(1), Math::Business::RSI, Math::Business::StockMonkey, Math::Business::StockMonkey::FAQ, Math::Business::StockMonkey::CookBook

"An Introduction to ConnorsRSI" by Connors Research, LLC, 2012