|
PureBytes Links
Trading Reference Links
|
Gitanshu,
my apologies for the delay - I had intended to respond sooner, but didn't
find the time for a post that would do justice to your questions.
A lognormal distribution based on Statistical HV is what I use. This sounds
very scientific, but much is in the eye of the beholder here, as you will
see presently. I usually have the curve centered in the current price of the
underlying; however, when I have a strong conviction regarding price
direction, I will allow myself some leeway (up to 0.5 sigma) in choosing a
center in the direction of the presumed trend.
Next thing to put into the calculation will be a volatility estimate for a
certain date in the future. The date is easy, because I hardly hold any
trade longer than 3 weeks, so it's going to be the expiration about a month
away. As for the volatility estimate, there are no hard and fast rules
again, but starting from the current statistical volatility I will look at
different scenarios to see how my intended position(s) would behave.
All this is comparatively easy to do with OptionVue which is my main tool
for this kind of research. In fact, the danger is that it gives you a
surfeit of quantitative information which makes your head swim. Remember
Buridan's ass - that metaphorical donkey introduced by a medieval
philosopher. The poor animal, though endowed with perfect reasoning and free
will, and though standing exactly in the middle between two exactly equal
bundles of hay, dies of hunger because it cannot find any logical reason to
choose one bundle in preference to the other. Well, I sometimes feel that
way with OptionVue.
That program also has a feature called "TradeFinder" which, using the above
inputs, will find and quantify with probabilities of profit all possible
positions pertaining to a set of given strategies. As an example, when I
expect volatility to rise, I will look for backspreads, long straddles and
strangles etc., and TradeFinder will ferret out those with the highest
probabilities. Again, these results have to be taken with a grain of salt;
some are totally unrealistic, with options incorrectly priced, or too deeply
in-the-money, etc.
A probability of profit > 50% is a necessary (not sufficient) condition for
me to enter a position. Once the trade is on I don't look at probabilities
much because usually it will quickly become apparent by visual inspection
whether the trade is working out or not. Like when you hire a new employee
you might depend on psychological tests and the CV and other information;
but once the new guy works in the office, it will probably be easy to see
what kind of a person he is.
You are asking for examples for my way of trading. I don't usually talk
about my trades publicly because I don't see any reason to, and also my
business partners don't want it, but to illustrate here is one trade I did
last November. On Nov. 1, PMCS was looking to me like it was going lower.
Actually, I had posted to that effect on the Metastock list.
PMCS seemed interesting for several reasons. The stock was in an obvious
downtrend, far below the major averages, and had just tested support twice
within only a few days' time. A three day retracement ending with a reversal
bar seemed to indicate that at least one more test of support was to be
expected. Statistical volatility was about 125%, an extremely high reading
for a stock which, just two months ago, had shown a reading of 55% or so. At
the same time, implied volatilities were comparatively low - not in absolute
numbers, but in comparison. If PMCS were to break support, a further rise in
implied volatility could be expected.
Furthermore, the stock had closed near the 1-sigma band which in my
experience is often indicative of an impending move the size of more than
one standard deviation within a few days. All seemed set for a move down, so
I didn't want to start out delta-neutral, but I also wanted to have some
insurance in case I was wrong.
Given this situation, I would have loved to do a Put Backspread at a credit;
this, in my opinion, would have been a perfect fit for my expectations.
Unfortunately, the volatility skew present in the puts made this strategy
unfeasible. After some deliberation, and trying out various scenarios with
OptionVue again, including simple long puts, I finally put on an asymmetric
long strangle wherein, if memory serves me right, I bought a number of Dec.
155 puts, and half that number of Dec. 165 calls. I don't remember the
prices, and the records are with my CPA now, but I do remember that the
trade worked out beautifully, and I closed the position a week later with a
smile on the face of the tiger.
Of course, I have made full use here of my constitutional right to choose a
winning trade for my example. Believe me I have losers, too, but the nice
thing is that with positions always protected, occasional losers cannot turn
into nightmares.
Have a nice weekend,
Michael Suesserott
-----Ursprungliche Nachricht-----
Von: onwingsofeagles@xxxxxxxxxxxxx
[mailto:onwingsofeagles@xxxxxxxxxxxxx]
Gesendet: Wednesday, January 24, 2001 01:01
An: realtraders@xxxxxxxxxxx
Betreff: [RT] Re: conservative?
Mike,
A few q's about the foll probability thing:
a/ Bell curve/ ln distro based on or decided by what? Historical IV?
Statistical HV? How does the trade probability adjust for a change in
the shape of the curve (eg from tall to flat, or vice versa). How
does the trade itself adjust once the probability changes?
b/ How does the trade look? I mean, where on the curve does the trade
start out? Median? Tails?
c/ Any examples? Historical are fins, in case you want to preserve
the privacy of the current trades.
Thanks much!
Gitanshu
<snip>
> (2) a probability of profit greater than 50%,
>
> ad 2: greater than 50% probability may not sound like much, but
remember
> this is the mathematical expectation based on the Bell-curve (or,
rather, a
> lognormal distribution). It doesn't take into account my
(hopefully) expert
> opinion on the market which, hopefully again, will raise the
probability of
> profit to an acceptable level.
To unsubscribe from this group, send an email to:
realtraders-unsubscribe@xxxxxxxxxxx
------------------------ Yahoo! Groups Sponsor ---------------------~-~>
eGroups is now Yahoo! Groups
Click here for more details
http://click.egroups.com/1/11231/0/_/152424/_/980630718/
---------------------------------------------------------------------_->
To unsubscribe from this group, send an email to:
realtraders-unsubscribe@xxxxxxxxxxx
|