Asymetric Trend Trading

How do you know when a trend has started? Ended? This forum is for discussions about trend indicators and signals.
Post Reply
ForrestGump
Senior Member
Senior Member
Posts: 31
Joined: Mon Jun 09, 2003 10:50 pm
Location: Sydney Australia

Asymetric Trend Trading

Post by ForrestGump » Tue Feb 03, 2004 4:37 pm

G'day all,

A couple of disclaimers to start with.

1. I am still only learning (avidly) about trend trading.

2. The following is only a half formed idea but my past experience has taught me that by putting my thoughts down in a public readable format (ie a forum) helps me to get my thoughts straightened out in my own head.

Anyway...

This thought came to me only last niight after an incident yesterday. I occassionally do a little derivative playing (Australian Warrants) on a few ASX stocks, including RIO TINTO (ASX code RIO). About mid-afternoon yesterday, the price of RIO moved down a few cents and then, over just a few minutes, went into freefall. Now, the actual reasons for this may be many, including the wandering around of the $A and the anticipation of an Australian rate rise the next day (today).

However, when discussing this with a friend it was pointed out to me that Comsec have only recently introduced automated stoplosses. Comsec are one of the largest Australian on-line brokers with most of their clientele being private traders - with a huge range of skill levels (probably biased to the bottom of the skill level range). My friend suggested that the sudden plummet was due to the naive use of automated stoplosses by these traders. Interestingly, the plummet stopped at an even dollar amount ($35.00), which may support this theory.

Anyway, I started by wondering about if there is some way that I could take advantage of this phenomenon, which may be short-lived (those that got stopped out unexpectadly yesterday may be a little more carefull next time).

From that thought, I jumped to my interest in trend trading. (I'm getting close to the point now, I promise :D )

My understanding of OTTR and other trend following systems that I have seen are that they are symmetric ie they enter both up and down trends on the same signal set, they stop and exit on the same rules and use the same MM principals. My thought is - "Is there a place for an asymetric trend following system". This would be a system that understands that the dynamics are (perhaps) different when a trend is going down than when it is going up.

A number of points.

1. Even aside from the specific (Comsec stoploss) situation mentioned, the psychology of traders, I am sure, is quite different in a falling market than it is in a rising one. Stoplosses (automated or not) are not the inverse of "Startprofits", though I think that symetric trend trading systems tend to treat them as such (when looking at mass market psychology)

2. Thinking on it, this would probably have most relevance in the equity markets where the vast majority of positions in the market at any time would be "long". (Most equity traders are either long or out)

3. Even just looking at Position Sizing may be of some interest. Postion sizing in the OTTR is based on volatility. Could we look at "negative volatility" and "positive volatility" as two different things?

I'll leave it there but I am still thinking about it all

PS I love this forum (even if I dont understand half of it)

PPS Maybe the stuff above is useless or "old hat". However, on the off chance that I have stumbled upon something that makes somebody a fortune and they wish to thank me for providing this grain of sand that becomes an oyster then I would be happy for them to buy Veritrader for me in thanks. :D

Christian Smart
Roundtable Fellow
Roundtable Fellow
Posts: 50
Joined: Fri Apr 18, 2003 8:53 pm
Location: Huntsville, AL
Contact:

Asymmetry for Trend Following Longs and Shorts

Post by Christian Smart » Tue Feb 03, 2004 5:04 pm

Peter Aan has done some research in this area and it's published in the documentation for his "Mystery System." The optimized version of the Mystery System is asymmetric. [I have no relation to Peter Aan other than being a satisfied customer of his relatively cheap [$95] research report on the Mystery System.]

Forum Mgmnt
Roundtable Knight
Roundtable Knight
Posts: 1842
Joined: Tue Apr 15, 2003 11:02 am
Contact:

Post by Forum Mgmnt » Wed Feb 04, 2004 8:45 am

At least for trades using daily data, Stock Trading almost requires asymetric long and short rules.

The behavior of short trades and long trades is very different. There are plenty of reasons for this, the most obvious being that most people only buy stocks and don't short them.

- Forum Mgmnt

shakyamuni
Roundtable Knight
Roundtable Knight
Posts: 113
Joined: Wed Jun 04, 2003 9:44 am
Location: Somewhere, Hyperspace

Post by shakyamuni » Mon Feb 09, 2004 11:32 am

Forum Mgmnt,

I agree completely. One way to see this behavioral characteristic of stocks in action is to examine the relative results of going long 52-week highs and short 52-week lows.

Though I usually find most academic studies to tend toward the flimsy side, here is a study that I feel illustrates the point fairly well. (NOTE: This is not a statement in advocacy of the claims or methods put forward in this paper. Readers may want to do some of their own testing to obtain more robust results.)

http://www.afajof.org/Pdf/forthcoming/George-Hwang.pdf

Post Reply