LevelTrading

Michael Covel - Trend Following

Chapter 1

Let's put change and Trend following in perspective. Markets behave the same as they did 300 years ago. In other words, markets are the same today because they always change. This is a philosophical underpinning of Trend Following.
Follow the trend - don't try to guess how far a trend will go. You can't. "Price makes news, not the other way around. A market is going to go where a market is going to go" =>(this guy is stating that you should follow the market, while i believe i must anticipate it...)

Let's say you saw a stock go from 5 to 100. When it was at 5, you didn't know it was going to go to 100. And trend followers didn't know it was going to go to 100 either. But they were buying all along, knowing that it could go to 100 even though it might not. You can't time the trade. No one can pick a top or bottom

A wise trend follower once told me a story with a new trader who wanted to learn the secrets. The experienced trader took the newbie out to the beach. They stood there watching the waves break against the shoreline. The neophyte asked, 'what's your point?' The trader said, 'go down to the shoreline where the waves break. Now begin to time them. Run out with the waves as they recede and run in as the waves come in. Can you see how you could get into rhythm with the waves? You follow the waves out and you follow them in. You just follow their lead.'

Chapter 2 - Great Trend Followers

Bill Dunn - Only 2 systems. The first he made in 1974, 2nd in 1989. The major strategic elements - how and when to trade, how much to buy and sell - have never changed in almost 30 years. We expect change. None of the things that have happened in the development of new markets over the past 30 years strike us as making the marketplace different in any essential way.

John Henry - No one consistently can predict the future. Prices, not investors, predict the future. We rely on the fact that other investors are convinced that they can predict the future, and i believe that's where our profits come from

He studied 18th and 18th century price data to prove to himself that there was only one successful way to approach trading
Long term trend identitifcation
Methodology is designed to keep discretionary decision making to a minimum
Risk management - strict formulaic risk management system that includes market exposure weightings, stop loss provision, and capital commitment guidelines that attempt to preserve capital during trendless or volatile periods.
Global diversification - participates in 70 markets in many countries
long term - 'There is an overwhelming desire to act in the face of adverse market moves. Usually it is termed 'avoiding volatility' with the assumption that volatility is bad. However, I found avoiding volatility really inhibits the ability to stay with the long-term trend. The desire to have close stops to preserve open trade equity has tremendous costs over decades.
stocks - the current thinking is that stocks have outperformed everything else for 200 years. But there is no one in the year 2000 that you can convince to jettison the belief that 200 years of performance will not cause stocks to grow to the sky. What will be new to them is an inevitable bear market.
On his system:

time frame is long term, with the majority of profitable tardes lasting longer than 6 weeks, some lasting several months
the system is neutral in markets until a signal to take a position is generated
it is not uncommon for markets to stay neutral for months at a time, waiting for prices to reach a level that warrants a long or short position
predefined levels of initial trade risk. if a new trade turns unprofitable, risk parameters will force a liquidation when a preset level is reached. A trade can last for as little as one day in this situation.

The changing world is not going to hurt if you have principles designed to adapt. So the markets Have changed. But that's to be expected and it's good.

Chapter 10 - Trading Systems

after a trend has begun. the goal is to ride the trend
technical indicators are a part of the system not the system. And only about 10% (MACD, %R ect...)
you will probably have more losses then gains because you don't know which trend will be the big winner. You accumulate many small losses trying to find it.

Trading system you design for one asset must be able to work in different asset classes.

If you design it for T-bonds, when you apply to Euro, corn, gold or anything else it should also work reasonably well.
Design parameters should also work well. If CCI at 20 works very well, but 19 and 21 don't, it is not a robust model.
You should be able to describe the strategy in relatively simple terms. (ie: CCI anticipates lows and highs in the market based on the frequency of previous lows and highs)

Appendix - Trend following research

Tested stocks 1983 to 2004, includes delisted issues, adjusted for dividends, volume / $15 min price filter, $250K minimum traded / day in 1983(inflation adjusted for future periods)
Entry at all time high
Exit at previous high - 10*ATR(14 of previous high) [for volatile stocks could be 55%, non-volatile, could be 20%]

15% return, had some years where loss was much greater than stop loss because of a gap down or something else. 1987 was almost 5% of trades. However, there were regular periods were avg $ gain / stop loss in $ was 60% of trades.

50% of all stocks significantly underperformed the index
25% of all stocks that have ever been in the index are responsible for ALL the gains
Most of the big winners spent a disproportionate amount of time making multi-year highs (from $20 -> $200 happens in increments)

Pages