Dear brethren: While I myself feel more inclined to the discretionary approach, as it is the path I have taken from the start, I would like to know the way the funds' machines are thinking. I guess the rules applied to mechanical trading have a predictive value given the amount of money involved in the funds' operations. As far as I know, a mechanical system needs quantifiable technical devices. Hence the popularity of moving averages, expansion and contraction cycles, time cycles, the four-week rule, etc. It would appear that moving averages are primarily used for trend assessment and to determine overbought or oversold zones through their spreads. Expansion and contraction cycles and the four-week rule would be used to identify increasing momentum and trend strength through the breakout from established resistance and support levels; and time cycles would be used for timing purposes. In short, which do you think are the criteria used to assess a trend strength and whether more money will be committed to a given market? How is timing calculated? Are all the mechanical system continuous in nature, i.e., they are always long or short with additional criteria for scaling in or out? If so, which criteria? Thanks. OPC
OPS, what exactly is your objective? Are you interested in learning about Mechanical Trading in general or how it is applied by large funds? Application of mechanical trading by individuals like us is very different than how large funds operate. Typically a large fund would trade 30-40 markets around the world, most utilize trend following strategies with money management models. A good example is John Henry. http://www.jwh.com/
I am interested in learning about mechanical trading as applied by large funds. It would appear that most funds are using the same criteria to make investment decisions, and they seem to be purely mechanical/systematic in nature. I know it's a tall order asking for these criteria in a public forum, but it sounds like that all they are doing is tracking volatility contraction/expansion along with some breakout setups and long term cycles. And if we are talking about commodity markets, I think that kind of information has a lot of predictive value, for large funds seem to be in a constant tug of war with their counterparts: the large hedgers. OPC
For me, discretionary, as applied to trading, means using the human judgement for trading decisions. Mechanical, on the other hand, means systematizing the human judgement about market behaviour and translating it into specific instructions for computer execution. OPC
Mechanical trading is based on the theory that not all aspects of the market are random. Through testing we can then find this non randomness which takes place over long term tendencies.
ET: Would you give me only three principles upon which non randomness is translated into mechanical instructions? I mean, if you had a large fund under your control, and you could give only three instructions for your computer, which ones would you give? OPC