39 lines
3.1 KiB
Plaintext
39 lines
3.1 KiB
Plaintext
828 Part VI: Measuring and Trading Volatility
|
||
make the move to profitability (or not make the move into loss territory, if you're sell
|
||
ing options). This is where historical volatility plays a big part, for it is the input into
|
||
the probability calculator. In fact, no probability calculator will give reasonable pre
|
||
dictions without a good estimate of volatility. Please refer to the previous chapter for
|
||
a more in-depth discussion of probability calculators and stock price distributions.
|
||
The use of probability analysis also mitigates some of the problems inherent in
|
||
the method of selection that compares implied and historical volatilities. If the prob
|
||
abilities are good for success, then we might not care so much whether the options
|
||
are currently in a low percentile of implied volatility or not (although we still would
|
||
not want to buy volatility when the options were in a high percentile of implied
|
||
volatility and we would not want to sell options that are in a low percentile).
|
||
In using the probability calculator, one first selects a strategy (straddle buying,
|
||
for example, if options are cheap) and then calculates the break-even points as
|
||
demonstrated in the previous section. Then the probability calculator is used to
|
||
determine what the chances are of the underlying instrument ever trading at one or
|
||
the other of those break-even prices at any time during the life of the option position.
|
||
It was shown in the previous chapter that a Monte Carlo simulation using the fat tail
|
||
distribution is best used for this process.
|
||
An attractive volatility buying situation should have probabilities in excess of
|
||
80% of the underlying ever exceeding the break-even point, while an attractive
|
||
volatility selling situation should have probabilities of less than 25% of ever trading
|
||
at prices that would cause losses. The volatility seller can, of course, heavily influence
|
||
those probabilities by choosing options that are well out-of-the-money. As noted
|
||
above, the volatility seller should, in fact, calculate the probabilities on several dif
|
||
ferent striking prices, striving to find a balance between high probability of success
|
||
and the ability to take in enough premium to make the risk worthwhile.
|
||
Example: The OEX Index is trading at 650. Suppose that one has determined that
|
||
volatilities are too high and wants to analyze the sale of some naked options.
|
||
Furthermore, suppose that the choices have been narrowed down to selling the
|
||
September options, which expire in about five weeks. The main choices under con
|
||
sideration are those in Table 39-2. The option prices in this example, being index
|
||
options, reflect a volatility skew (to be discussed later) to make the example realistic.
|
||
The two right-hand columns should be rejected because the probabilities of
|
||
the stock hitting one or the other of the striking prices prior to expiration are too
|
||
high well in excess of the 25% guideline mentioned earlier. That leaves the
|
||
September 500-800 strangle or the September 550-750 strangle to consider. The
|
||
probabilities are best for the farthest out-of-the-money options (September 500-
|
||
800 strangle), but the options are selling at such small prices that they will not pro- |