Improving your trading strategy using probability

Shortforbread · Nov 5, 2002

Eddy,

You can find some of this info in the "Batting .800" manual by Larry Williams on page 229. The page has a chart with these figures. He claims to use this for trading commodities like Pork Bellies and Copper. He also has info on ratios on page 224.

Chow

Mr Subliminal · Nov 5, 2002

Originally posted by dottom
Research on trend-following methods clearly show a higher % chance of losing trade following larger winning trade; and higher % chance of winning trade after series of losing trades.

Apparently not clear enough to convince Malkiel or Niederhoffer.

swoop[TR] · Nov 5, 2002

Anyone know the Feigenbaum constant ? Order in Chaos? Hello?
Saying that the DJIA has been going up for the past century doesn't mean anything, it is a little bit like the chaos theory. Now, the idea is: Can you find a way to consistently trade your strategy and let your profits run if it pans out and cut it when it doesn't. A system with more than 80% efficiency (Right to wrong ratio) is probably curve fitted IMO.

rtstrading · Nov 5, 2002

http://members.aon.at/tips/index.html

place your bets

dottom · Nov 5, 2002

Originally posted by Mr Subliminal
Apparently not clear enough to convince Malkiel or Niederhoffer.

Whether you utilize different position sizing techniques based on series of winners/losers of your system depends on the system itself. The edge that this type of adjustment to position sizing may give over standard money-management techniques may not be enough of an EV gain to warrant the extra layer of complexity (and like I said there is psychological effects as well that can affect your trading).

But to think that one trade is totally indepdendent from the result of a prior trade(s) is naive.

dottom · Nov 5, 2002

Originally posted by swoop[TR]
Anyone know the Feigenbaum constant ? Order in Chaos? Hello?
Saying that the DJIA has been going up for the past century doesn't mean anything, it is a little bit like the chaos theory.

Hmm, chaos theory completely refutes your previous comment:

Originally posted by swoop[TR]
Unfortunately, the market is RANDOM

Mr Subliminal · Nov 5, 2002

dottom, we are not in disagreement. As you yourself have cautioned, however, this is not readily exploitable by the typical ET reader.

dottom · Nov 5, 2002

Originally posted by Mr Subliminal
dottom, we are not in disagreement. As you yourself have cautioned, however, this is not readily exploitable by the typical ET reader.

I agree that Chaos theory is not readily exploitable by the typical ET reader. I only used swoop's own acknowledgement of market dynamics to refute his view that "markets are random."

However, regarding the discussion of varying position size in relation to winning/losing trades is indeed very applicable.

P2 has posted extensively on his trading techniques.

Many pairs traders and commodities seasonal hedgers increase their position size in similar fashion.

Before everyone throws out LTCM, Niederhoffer, or the many stat arb hedge funds that have shut down in the past as examples that these strategies do not work, I will say that in all these cases what we can agree on is that these entities did not properly account for risk and volatility (i.e. they were not truly hedged). But their failures does not in any way prove that varying position size in response to winning/losing trades is not a viable method of increasing returns. What you have to do is keep your risk in perspective and view such methods from return-on-risk, not overall return.

swoop[TR] · Nov 5, 2002

Chaos theory does not disprove anything.
The market is a random occurence, the underlying familiarity that exists in the market does not mean that it isn't random. It is random by all means just like a fractal is random yet yields a recognizable pattern. If you really believe that you can improve your strategy by statistically "filtering out" the bad trades then by all means do so but it has been proven time and time again that it doesn't work.

dottom · Nov 5, 2002

Originally posted by swoop[TR]
Chaos theory does not disprove anything.
The market is a random occurence, the underlying familiarity that exists in the market does not mean that it isn't random. It is random by all means just like a fractal is random yet yields a recognizable pattern. If you really believe that you can improve your strategy by statistically "filtering out" the bad trades then by all means do so but it has been proven time and time again that it doesn't work.

After you've read some basic books on chaos theory then get back to me.

Also, I'd like to see your proof that it has been proven time and time again that it doesn't work. I have statistical proof that trend-following systems show increase % chance of a losing trade after a large winning trade. This is from my own research when I was taking Russell Sands course many years ago.

What about P2's methods or stat arbs who vary position size based on previous previous trades result in relation to market movement??

If the markets are so random, every heard of fat tails? Explain that phenomenon please if markets are random as you say and follow Browinian motion.

Please explain why there is a non-Gaussian distribution of prices, if indeed prices are random.

"Everyone knows that prices in financial markets move up and down, but no one knows why or now. Economists claim these price moves are a random walk. They are the unpredictable product of efficient markets. Prices in these markets reflect the activity of rational, logical, and always equally well-informed investors. I don't know about you, but I'm not always a rational human being, and I think this is a pretty far-fetchecd view of the world."

(10 brownie points if you tell me who that quote is from?)