Trend change days (top and bottom picking) part 2

The second technique is one mentioned in Dynamic Trading by Robert Miner. It uses the time duration of previous swings to calculate the duration of the current swing. I prefer to weight the swing in the same direction two swings back more heavily then the swing in the same direction immediately previous to the current one.

For the three indexes shown so far here is the swing duration in days for the first swing and the most recent "high" day to count from:

$INDU 33 trading days 2/01/08

$NDX 9 trading days 2/01/08

$NYA 33 trading days 2/27/08

Now extending by various Fibonacci ratios from the high dates you get dates like these:

Index Ratio Result

$INDU 1.000 3/19/08

$NDX 4.236 3/26/08

$NYA 0.618 3/26/08

That's the result for using the "first" swing from the highs of last fall.

For using the "second" swing I get the following dates counting swing durations both from the nominal tops of these index and also from the second lower tops in each case:

Index Primary Top: Ratio Date Secondary Top: Ratio Date

$INDU .5000 3/25/08 1.272 3/26/08

$NDX .618 3/25/08 1.272 3/27/08

$NYA .236 3/21/08 0.500 3/19/08
$NYA 0.618 3/25/08

Using these methods, a consensus of around 3/25 - 3/26/08 seems to be emerging, again with a somewhat less popular option of 3/19/08 - 3/21/08