|
PureBytes Links
Trading Reference Links
|
I don't know about 1929, but 1987 exceeded any statistical measure. Well in
excess of 5 which describes almost anything possible. The question however is
incorrect in it's nature. When you have events like 1929 or 1987 or major moves
and non continuous pricing... the model you are using such as binomial or a B/S
benchmarks to wrong number. If 1987 was inches of a 5+ move than the reference
point of what would be +/- is invalid. That is why delta hedging in options
dies after 1987. Events cannot be 5 10 or 20 std. deviation moves .. they are
in fact 2 0r 3 std deviation moves and the perception of expected deviation
alters. For example if the 108 point move on 10/16 of 1987 wasn't
"extraordinary than the 500 point move on 10/19 wasn't really as "unexpected" of
an event.
There used to be a MIT PhD on the board of the CBOE who quoted 10/87 .. if you
benchmarked to implied volatility as a 22 std. deviation event. he then went on
to quote NATO's estimate of a war in Berlin with East German troops and tanks
breaching the Berlin Wall as 5 std. deviation event. So if you used traditional
volatility as benchmark 10/87 was much less likely than nuclear war using the
computer simulation. It is one of the reasons why "traditional" benchmarks like
historic volatility are only still used by former successful option traders.
Lionel Issen wrote:
> How many standard deviations did prices move in the crash of 1929 and what
> was the statistical probability of this happening assuming a normal
> distribution?
> Lionel Issen
> lissen@xxxxxxxxx
> ----- Original Message -----
> From: Michael Suesserott <MikeSuesserott@xxxxxxxxxxx>
> To: <metastock@xxxxxxxxxxxxx>
> Sent: Sunday, August 13, 2000 6:14 AM
> Subject: AW: What options to sell?
>
> > Gitanshu,
> >
> > thanks for your interesting post, and also for the friendly and
> constructive
> > approach you are taking. Like yourself, I entered into this thread in a
> > spirit of trying to be helpful. There is nothing I stand to gain through
> my
> > posting here, and I assure you I would immediately fall silent if any
> > flaming arose from it.
> >
> > As you surmised, I am well aware of the points you are raising. Though the
> > counterexample you were using does not describe my position quite
> > correctly - I would not think it advisable in Guy's case to buy puts with
> > only one week to go - still I agree that there are many different spread
> > positions that could be used to trade direction to good advantage.
> >
> > In fact, one additional strategy that I like to use sometimes to trade
> > direction is the sale of backspreads (provided volatilities are right, and
> > there is a certain volatility skew). So there are indeed many avenues to
> > explore in any given situation, as you were rightly stating.
> >
> > Now let me explain why I don't think these applicable in the case of Guy's
> > system.
> >
> > Anyone who has ever devised a mechanical trading system will have had this
> > experience: you change a certain parameter a little - just a little! - and
> > test results start diverging by a wide margin. Unfortunately, most of the
> > time this happens, results deteriorate.
> >
> > Now here we have this successful trading system that it took Guy and his
> > family 30 years to develop. Don't you agree that Guy would be well advised
> > to be extremely careful, even reluctant, to change his system?
> >
> > There can be no doubt that replacing futures with option spreads,
> especially
> > those where one gets short premium, thus limiting possible profits,
> > constitutes a fundamental change of main characteristics of a trading
> > system. To name only one effect that is immediately obvious - the big per
> > trade profits (up to 150 points, as Guy stated) just wouldn't have
> occurred.
> > It is true that losses (up to 62 points), too, might have been less, but
> > since the system had 19 winning trades and only 3 losing ones, the use of
> > credit spreads would very likely have led to a severe deterioration of the
> > system.
> >
> > Besides, many types of spreads require the use of stops, or at least
> > constant supervision and readjustment; readjustment "rules" are not really
> > clear-cut because there is a choice of option strikes and volatilities and
> > deltas, even different follow-up strategies, with the ensuing action to be
> > individually determined by the trader in each and every case. If you can
> > call this a "system" at all, it will certainly not be the same system Guy
> > had been trading before.
> >
> > That is why I didn't take spread trading into consideration in my posts.
> The
> > only option strategy that would preserve the characteristics of Guy's
> system
> > at least to a reasonable extent, would be the simple purchase of puts or
> > calls, as the case may be.
> >
> > This strategy may indeed prove useful in catastrophic situations such as
> the
> > October 87 crash where prices moved more than 12 standard deviations, an
> > event that statistically should have occurred less than once in the
> history
> > of the universe.
> >
> > Kind regards,
> >
> > Michael Suesserott
> >
|