39 lines
3.1 KiB
Plaintext
39 lines
3.1 KiB
Plaintext
816 Part VI: Measuring and Trading Volatility
|
||
600 days of implied volatility history for the purpose of determining percentiles, but
|
||
a case could be made for other lengths of time. The purpose is to use enough implied
|
||
volatility history to give one a good perspective. Then, a reading of the 10th per
|
||
centile or the 90th percentile will truly be significant and would therefore be a good
|
||
starting point in determining whether the options are cheap or expensive.
|
||
In addition to the actual percentile, the trader should also be aware of the width
|
||
of the implied volatility distribution. This was discussed in an earlier chapter, but
|
||
essentially the concept is this: If the first percentile is an implied volatility of 40% and
|
||
the 100th percentile is an implied volatility of 45%, then that entire range is so nar
|
||
row as to be meaningless in terms of whether one could classify the options as cheap
|
||
or expensive.
|
||
The advantage of buying options in a low percentile of implied volatility is to
|
||
give oneself two ways to make money: one, via movement in the underlying (if a
|
||
straddle were owned, for example), and two, by an increase in implied volatility. That
|
||
is, if the options were to return to the 50th percentile of implied volatility, the volatil
|
||
ity trader who has bought "cheap" options should expect to make money from that
|
||
movement as well. That can only happen if the 50th percentile and the 10th per
|
||
centile are sufficiently far apart to allow for an increase in the price of the option to
|
||
be meaningful. Perhaps a good rule of thumb is this: If the option rises from the cur
|
||
rent (low) percentile reading to the 50th percentile in a month, will the increase in
|
||
implied volatility be equal to or greater than the time decay over that period?
|
||
Alternatively stated, with all other things being equal, will the option be trading at
|
||
the same or a greater price in a month, if implied volatility rises to the 50th percentile
|
||
at the end of that time? If so, then the width of the range of implied volatilities is
|
||
great enough to produce the desired results.
|
||
The attractiveness to this method for determining if implied volatility is out of
|
||
line is that the trader is "forced" to buy options that are cheap ( or to sell options that
|
||
are expensive), on a relative basis. Even though historical volatility has not been taken
|
||
into consideration, it will be later on when the probability calculators are brought to
|
||
bear. There is no guarantee, of course, that implied volatility will move toward the
|
||
50th percentile while the position is in place, but if it does, that will certainly be an
|
||
aid to the position.
|
||
In effect this method is measuring what the option trading public is "thinking"
|
||
about volatility and comparing it with what they've thought in the past. Since the pub
|
||
lic is wrong (about prices as well as volatility) at major turning points, it is valid to want
|
||
to be long volatility when "everyone else" has pushed it down to depressed levels. The
|
||
converse may not necessarily be true: that we would want to be short volatility when
|
||
everyone else has pushed it up to extremely high levels. The caveat in that case is that |