|
PureBytes Links
Trading Reference Links
|
I meant to copy the O-list as well.
Regards,
Bill Vedder
X-Mozilla-Status2: 00000000
Message-ID: <3928229B.9BEB0AF4@xxxxxxx>
Date: Sun, 21 May 2000 12:53:31 -0500
From: Bill Vedder <bved01@xxxxxxx>
X-Mailer: Mozilla 4.72 [en] (Win98; I)
X-Accept-Language: en
MIME-Version: 1.0
To: Paul Altman <paulha@xxxxxxxxxxxxx>
Subject: Re: Tharp's Expectancy
References: <4.3.2.20000519143917.00c80a90@xxxxxxxxxxxxxxxxxx> <4.3.2.20000520201827.00b38c80@xxxxxxxxxxxxxxxxxx>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
I've read the posts on this thread and will reply to this one.
First, Mel is correct; I meant to refer to the mathematician Dr. Edward O Thorp
and not "Tharp". It's Dr. Thorp's book to which I was referring to. Thanks for
catching that Mel.
See below...
Paul Altman wrote:
> Bill:
>
> >This thinking is absolutely incorrect. A negative expectancy game is just
> >that; over time the returns are negative. No position sizing routine, no
> >betting algorithm, no stop-loss strategy can change this. You will lose
> >playing a negative expectancy game.
> >You can change the rate of loss, but the final outcome is completely
> >determined.
>
> Sorry, but I still don't see it. Let's do a thought experiment. We look
> at a hypothetical system that's had 100 trades historically. 50 of them
> returned 5%, 50 of them returned -10%. Obviously, a negative expectancy
> based on V. Tharp's definition (we've backed out and normalized the
> position sizing, by examining only the return).
I need to get very picky here because both our arguments rely on very precise
definitions. I want you to understand how I arrive at my conclusions.
In the above example, you are assuming that each of the 100 trades had an equal
dollar bet. No problem; that's ok. With that assumption, then we both agree; the
above system has a negative expectation. Some of the trades are winners and some
losers but the long term "expectation" is you will lose money.
> But what if you had had an algorithm that somehow "knew" to leverage your
> winning bets by 4/1. Now we have a very profitable game, but Tharp's
> "expectancy" is still negative.
This logic escapes me. How can you know ahead of time which trades will be
winners? If you know this, don't bother trading the system when it's projecting a
losing trade.
If you're saying that, based on your system testing you have developed filters
(or whatever) that work to identify winning trades then you have now completely
changed the premises! You have two completely different systems (one without the
algorithm - the negative expectancy game and one with the alogorithm - the
positive exp game). Tharp's expectancy has now changed as well... it positive for
the second game.
>
> The "knowing" which bets to leverage is not absurd. There are many folks
> who load up on what they consider to be higher probability bets.
>
Ok. I'll buy that. But... their expectation is that these trades are winners.
They are (maybe subconsciencely) adding a filter to their system, much like your
example above.
>
> After reading many thoughtful replies to my letter, I still believe I'm
> correct, though I was missing a critical point. Phil Lane helped clarify
> the situation for me, when he replied:
>
> >If you know which trades to load up on, then you should reprogram the system
> >to take those trades and skip the rest. And then you should be able to
> >demonstrate a "positive" expectation w/o regard to size..
>
Yes.
>
> This helped me realize that, while it is theoretically possible to have a
> very profitable negative expectancy game over long periods of time,
No. By definition.
> it could only occur by loading up on high prob trades.
You're changing the system.
> And if you were clever
> enough to do so consistently, then you'd be frivolously hurting your equity
> curve by continuing to make the smaller trades.
>
Yes. Why trade when you "know" you're going to lose?
>
> I conclude that V. Tharp's expectancy is indeed critical to track, because
> a negative expectancy game is highly likely to be a loser over time,
Not highly likely. It's a certainty. By definition.
> AND
> because if a game has a negative expectancy and yet is profitable over
> time,
Impossible.
> it can be immediately improved by simply cutting out the smaller
> trades.
If he's saying that (I haven't read his stuff) he's wrong.
> If this thinking is fuzzy, please say why but please do respond to
> the thought experiment, also.
>
> >For a better explanation than I can give, follow this link to the online
> book,
> >"The Mathematics of Gambling" by Edward Tharp. One of the later chapters in
> >the book discusses this very topic.
> >
> >http://www.bjmath.com/bjmath/thorp/tog.htm
>
> Thank you very much, I'll d/l it tomorrow and read it thoroughly.
>
> Thanks to all of you for taking the time to respond. I would welcome any
> add'l criticism, as I would like to get my thinking about this very clear.
>
No "criticism" intended here. You've obviously put some thought into your posts
and I've enjoyed responding. It's definitely one of the better threads lately.
>
> BTW, I have learned a lot from Van Tharp's book, and recommend it highly.
>
> Paul
Regards,
Bill Vedder
|