[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

GEN - volatility, hist. vs imp.



PureBytes Links

Trading Reference Links

The new year brings a new insight.  Attached is a quick study of the ODDS
Probability cones by Don Fishback as implemented in MetaStock 6.5.  Below
the OEX chart is a Projection Band oscillator by Mel Widner.  Briefly
reviewing the concepts and thesis of volatility and price forecasting we
need to look at historical volatility derived from price behavior over a
specified period of time in the past and implied volatility derived from
options that expire in the future.  We also need to select a price range
expectation defined as %Probablity generally measured as 1 or 2 standard
deviations, 68% and 90%. 

In beginning this simple experiment the question arose as to which provided
a more correct view of the future pertaining to price behavior, i.e. "does
the past foretell the future or does the implied volatility of the options
foretell the future?"  The first set of cones was laid out using the blue
and red cones which both have the same 68.269% probability, statistically
meaning that 68.269% of future prices would be contained within the cones.
The blue cones have a 22 day lookback period for determining volatiltiy and
the red VIX cones have a 22 day look forward implied volatility derived
from OEX options pricing.  Clearly the VIX cones came closer to containing
prices within a +1,-1 Standard Deviation.  

Two questions arose from this first study.  One, is the look back period
for historical volatility in error, and or if the lookback period is
correct, then what probability would be needed to make the cones
coincident.  The ODDS MetaStock cone indicator permits a quick study of the
lookback period by clicking and dragging a stem from the apex of the cone
backward in time.  As the stem is lengthened or shortened the cone's
"mouth" widdens and narrows as historical prices vary with the time period
selected.  Ranging from 1 day to a hundred days would not bring the
historical cones at 68% probability into coincidence with the VIX cones at
the same probability.  So the Historical cone lookback was left at 22 days
and its % probability was varied to find the coincidence.  In all cases a
lookback of 22 days with a % Probability of 90% produced a cone that was
coincident with the VIX cone at 68% probability.
Is that orgasmic or what.

Going another step further with the mathmatical crystal ball, the Widner
Projection Band Oscillator(PBO) was inserted below the cone plot to attempt
some directional analysis.  To quote from page 480 of the MetaStock 6.5
manual, "The Projection Oscillator is basically a slope-adjusted
Stochastic.  Where the Stochastic Oscillator shows the relationship of the
current price to its minimum and maximum prices over a recent time period,
the Projection Oscillator shows the same thing, but the minimum and maximum
prices are adjusted up/down by the slope of the price's regression line.
This adjustment makes the Projection Oscillator more responsive to
short-term price moves than an equi-period Stochastic."  In keeping with
the same historical and implied periods of 22 days, the same period was
selected for the PBO. Overbought/oversold levels can be used as well as 50%
line crossings for signal generation.  As of the low on 12/26 prices are
climbing the VIX 22day/68% cone and the HIST 22day/90% cone.  The PBO has
crossed above the 50% line and is reaching for the arbitrary overbought 80%
line.  Even the PBO will saturate as the standard stochastic does when
price is trending.  It appears that there is price time probability
potential for higher prices ahead.

Cheers,
BobR
http://www.oextrader.com


Comparisons of the two and their use in timing of trades and strike price
selection
Attachment Converted: "c:\eudora\attach\cone1231.gif"