|
PureBytes Links
Trading Reference Links
|
Hi Folks,
If I may put in an oar as well... The correspondence here has been all
about computing Pivot Points (PP). Maybe we should think a bit about the
assumptions behind them and their validity - if any.
So, what's the issue here? If my Indicators or the Funnymentals (or my
Tummy) tell me after the close (I don't use the realtime version of MS)
to take some action on the next day, I observe the first hour or so of
trading for confirmation. I guess most of us do this in one form or
another. The problem here is of course what one judges to be a
sufficient price move to be considered as confirmation. PP's certainly
are one convenient (because extremely simple to calculate) instrument
for this, eg. buy on the next day if the Open is above Resistance Level
One (R1), or R2, or R3 (easy enough to invent an R3 & S3).
Okay, on a practical level I wonder why Open is excluded from the
computation. Because it wasn't available to the author at its
conception? Imo. there is valuable information being thrown away here -
or doesn't everyone consider Candlesticks superior to HiLoClose-Bars?
On a more fundamental level the theory behind, say R1 & S1 could be
stated like "Price moves of more than a certain amount from yesterday's
'True Price' are significant; smaller moves are just random clutter on
the chart". 'True Price' meaning the average of H, L & C. Why this
number should be regarded as more meaningful than yesterday's Close or
today's Open ecapes me but lets continue...
Re. the size of this of this 'significant Move', why it's just a
distance of more than half *yesterday's range* from the 'true price'
which can easily be seen by:
p = (H+L+C)/3 ; R1=2*P-L
Signif. Dist. Resist. = R1-p = (2*p-L)-p = p - L ; the same goes for S1:
Signif. Dist. Supp. = H - p ; so Total Significant Move (combining
upside & downside) is a stunning ;-)
H-L
R2 & S2 are just fixed at a larger but equally arbitrary distance from
yesterday's 'true price'. The only reason prices may turn on these
'Pivots' is because 'everybody does it' (a surefire way for getting
skinned).
There are better methods of separating significant moves from noise; why
some of them are even built into MetaStock. What about ATR or using
r-squared with a sufficiently loose confidence interval (as a practical
tip: Use appropriately narrowed Bollinger Bands), and don't forget Bill
Williams' Profundity with its Squat- & Fake-Bars which takes Volume into
account as well.
Kind Regards to Ye All & Happy Trading,
Jan Willem Roberts
|