San Diego bobl 15:18 GMT October 28, 2008
answer to questions:
Good questions.... let me try to convey some thoughts that are a bit different from the perspective I believe you intend.
1) Trend Following and volatility........first, from my study of trend followers, starting with Jesse Livermore, Donchian and going through to more modern times of Richard Dennis, Ed Seykota, Peter Henry, Dunn and so on, I've deduced what I feel they all share in risk management. Volatility is a "by-product" of trends.........by their very nature they are extensions of shape, form, or pattern. Volatility naturally goes up on extensions. Sizing down when things are extending your way is a mistake imho........you want to build up size when things are going for you. All these famous trend followers work towards hitting the home run; they need to because mathematically they have more starts and stops than pure runs. I believe I am correct in saying that trend following models focus first on financial risk to the position versus the mathematics of volatility. The way this works is that if you risk 3% of total equity on a theme, be it stocks, bonds, commodities, currencies, or trees in a forest, that is barometer of size going in. Now, if a trend develops, and works for you, then as your position profits, your "degree" of risk demenishes. Therefore, to maintain risk levels and increase potential profits, you buy/sell more to that original level of risk. For example, Richard Dennis used to define his original entry level, deduced mathematically, as a value of N. The cost of "N" then became the base function, to which IF a winning position, it became "N y (value of increase)"...... when y=N, then another lot was put on. On the other hand, if risk was valued at -2N, then if that number is elected, then the trend following system exits.
2) As to your second question: Top down approach is always a must imho. In simple terms, what happens in one week is certainly more meaningful and trade valued more than the movement of a market vehicle than an hour. Personally, I use several time frames, and obviously like to see them all in line, i.e. monthly/weekly/daily/4 hr/ 2 hr/ 1 hr............
As to the character of the market, I've always advised traders to understand the character and meaning of that character to their trading. Some general market characters are: trending, impulsive, breakaway, momentum, stairstep, sideways, flat, channel and so on. Now, here's the caveat for me. I measure the "retracement" of any move to ascertain the value and potential of any move thereafter. For instance, if a trading vehicle moves say 100 pts, then retraces 20 % before getting on its horse and going the original direction again.................and in another example the same trade goes 100 pts, then retraces 60%........ well, which move do you think is stronger and has more upside potential? Of course the trade with the least retracement. This current market is absolutely littered with validation of this approach. You increase size on the stronger move, and decrease size and risk on the weaker, more flexible moves...as in more noisy trade. IMHO if you study the reality of the above, then you'll increase your trade confidence and results exponentially.
Good Luck and Profitable Trading
San Diego bobl