John Murphy is StockChart.com's chief technical analyst. He has written extensively about technical trading. He wrote ten laws for technical trading that I will quote here. The source is the following link: http://stockcharts.com/education/TradingStrategies/MurphysLaws.html
1. Map the Trends
Study long-term charts. Begin a chart analysis with monthly and weekly charts spanning several years. A larger scale map of the market provides more visibility and a better long-term perspective on a market. Once the long-term has been established, then consult daily and intra-day charts. A short-term market view alone can often be deceptive. Even if you only trade the very short term, you will do better if you're trading in the same direction as the intermediate and longer term trends.
2. Spot the Trend and Go With It
Determine the trend and follow it. Market trends come in many sizes -- long-term, intermediate-term and short-term. First, determine which one you're going to trade and use the appropriate chart. Make sure you trade in the direction of that trend. Buy dips if the trend is up. Sell rallies if the trend is down. If you're trading the intermediate trend, use daily and weekly charts. If you're day trading, use daily and intra-day charts. But in each case, let the longer range chart determine the trend, and then use the shorter term chart for timing.
3. Find the Low and High of It
Find support and resistance levels. The best place to buy a market is near support levels. That support is usually a previous reaction low. The best place to sell a market is near resistance levels. Resistance is usually a previous peak. After a resistance peak has been broken, it will usually provide support on subsequent pullbacks. In other words, the old "high" becomes the new low. In the same way, when a support level has been broken, it will usually produce selling on subsequent rallies -- the old "low" can become the new "high."
4. Know How Far to Backtrack
Measure percentage retracements. Market corrections up or down usually retrace a significant portion of the previous trend. You can measure the corrections in an existing trend in simple percentages. A fifty percent retracement of a prior trend is most common. A minimum retracement is usually one-third of the prior trend. The maximum retracement is usually two-thirds. Fibonacci retracements of 38% and 62% are also worth watching. During a pullback in an uptrend, therefore, initial buy points are in the 33-38% retracement area.
5. Draw the Line
Draw trend lines. Trend lines are one of the simplest and most effective charting tools. All you need is a straight edge and two points on the chart. Up trend lines are drawn along two successive lows. Down trend lines are drawn along two successive peaks. Prices will often pull back to trend lines before resuming their trend. The breaking of trend lines usually signals a change in trend. A valid trend line should be touched at least three times. The longer a trend line has been in effect, and the more times it has been tested, the more important it becomes.
6. Follow that Average
Follow moving averages. Moving averages provide objective buy and sell signals. They tell you if existing trend is still in motion and help confirm a trend change. Moving averages do not tell you in advance, however, that a trend change is imminent. A combination chart of two moving averages is the most popular way of finding trading signals. Some popular futures combinations are 4- and 9-day moving averages, 9- and 18-day, 5- and 20-day. Signals are given when the shorter average line crosses the longer. Price crossings above and below a 40-day moving average also provide good trading signals. Since moving average chart lines are trend-following indicators, they work best in a trending market.
7. Learn the Turns
Track oscillators. Oscillators help identify overbought and oversold markets. While moving averages offer confirmation of a market trend change, oscillators often help warn us in advance that a market has rallied or fallen too far and will soon turn. Two of the most popular are the Relative Strength Index (RSI) and Stochastics. They both work on a scale of 0 to 100. With the RSI, readings over 70 are overbought while readings below 30 are oversold. The overbought and oversold values for Stochastics are 80 and 20. Most traders use 14-days or weeks for stochastics and either 9 or 14 days or weeks for RSI. Oscillator divergences often warn of market turns. These tools work best in a trading market range. Weekly signals can be used as filters on daily signals. Daily signals can be used as filters for intra-day charts.
8. Know the Warning Signs
Trade MACD. The Moving Average Convergence Divergence (MACD) indicator (developed by Gerald Appel) combines a moving average crossover system with the overbought/oversold elements of an oscillator. A buy signal occurs when the faster line crosses above the slower and both lines are below zero. A sell signal takes place when the faster line crosses below the slower from above the zero line. Weekly signals take precedence over daily signals. An MACD histogram plots the difference between the two lines and gives even earlier warnings of trend changes. It's called a "histogram" because vertical bars are used to show the difference between the two lines on the chart.
9. Trend or Not a Trend
Use ADX. The Average Directional Movement Index (ADX) line helps determine whether a market is in a trending or a trading phase. It measures the degree of trend or direction in the market. A rising ADX line suggests the presence of a strong trend. A falling ADX line suggests the presence of a trading market and the absence of a trend. A rising ADX line favors moving averages; a falling ADX favors oscillators. By plotting the direction of the ADX line, the trader is able to determine which trading style and which set of indicators are most suitable for the current market environment.
10. Know the Confirming Signs
Include volume and open interest. Volume and open interest are important confirming indicators in futures markets. Volume precedes price. It's important to ensure that heavier volume is taking place in the direction of the prevailing trend. In an uptrend, heavier volume should be seen on up days. Rising open interest confirms that new money is supporting the prevailing trend. Declining open interest is often a warning that the trend is near completion. A solid price uptrend should be accompanied by rising volume and rising open interest.
"11."
Technical analysis is a skill that improves with experience and study. Always be a student and keep learning.
- John Murphy
5 BAR PROPRIETARY LOOK-AHEAD ENVELOPES:
--------------------------------------
In my last posted comment, I promised to explain how one promising trend-following trading approach works. It works like this: You run a five bar, 1.5 standard deviation look-ahead envelope on any tradeable. It doesn't matter what period of time a bar represents. That is dependent on your trading time horizon and desired win/loss size. Your wins and losses will be greater for bars representing longer periods of time.
At the end of each bar (for example, at the end of each day, if your bars are daily bars), you will obtain a top and bottom edge of the next bar's trading period (for example, for tomorrow's prices, if you are using daily bars). You wait until prices actually touch or go through either the top edge or the bottom edge of the look-ahead envelope. Then, you enter a limit order to go long (bottom edge) or short (top edge) at the next bar's look-ahead envelope edge on the same side. The first and second "touches" don't have to be on consecutive bars. If prices move to the other side of the envelope before the second touch, change your limit orders to trade the other way on the opposite projected edge of the envelope (or establish a "straddle" position between the "touches"). In other words, always wait for the second "touch of" or "pass through" the edge of a look-ahead envelope to take a position. Your orders will always be limit orders. You should never have ANY slippage on executed orders. If prices touch or pass through the same edge on subsequent bars, add double the units of your last "bet" to your new position. If you had one unit (for example, one contract), add two new ones. If you had three (one plus two), and four units. Add the units at the limit prices represented by the envelope edges. This is the Martingale aspect of the trading approach. Exit all positions at the FIRST "touch" or "pass through" of the opposite edge of the envelope (top edge, if you are long or bottom edge, if you are short). That's basically all there is to it.
You might ask, why did I use five bars and 1.5 standard deviations? The reasons are as follows: First, if you run a five bar envelope, it is almost impossible to have more than 3 or 4 "touches" of an envelope edge all go against you. Therefore, you don't have to have as much backup capital as you would otherwise need if you used 15 bars, for example. I have never had more than three touches against me on any of the charts which I have looked at, although it would technically be possible to have more (but not very likely). Second, the choice of 1.5 standard deviations pretty much insures you that you will get position exit "touches" soon after taking a position. This also keeps capital requirements down.
Using notations from my historic chartbooks to plot out an example of this approach on a computer, make a 100 day daily bar chart of the S&P 500 stock index ending 03-Feb-94. Price bar charts work just as well as Japanese candlesticks (I just happen to prefer using a candlestick chart). The plot should use 5 days and 1.6 standard deviations (close to the 1.5 standard deviations which I later adjusted/refined my trading model to reflect, as mentioned earlier). Total winnings for this 100 day period were 50.10 S&P 500 points. There were 13 winning position series taken. There were three losing position series taken. The total number of winning points during this period would have been equivalent to about $23,834 (after reasonable commissions) if you had been trading S&P 500 futures contracts, as I was. That would have been about $59,585 annualized. About $70,000 margin would have been required for worst-case positions taken. These numbers assume you started each series of trades with a single contract. You would have been out of the market a significant part of the time. This approach is not limited to short-term trading. Using the same approach with monthly S&P 500 bars, obviously this is trading at a much slower pace. But there are no losses plottable on that same chart, even though several short positions were taken in a generally uptrending market. Six trades were taken over about a four year period. Total winnings for completed trades were 248.49 S&P 500 points. The last trade of the series was not completed as of 93-Feb-94 (the same termination date used for plotting out our 100 day bar chart), but it exits a short position at 462.16.
This approach is also not limited to stocks or stock indices. Plotting the March 94 sugar futures contract (one of the most volatile choices available) for the same time horizon, it has a 6 to 1 win/loss ratio.
In summary, the advantages of this trend-following trading approach would appear to be: (1) You use limit orders with no slippage. (2) The method actually works very well in congestion price ranges. (3) It also seems to work on many kinds of tradeables with the same parameters. (4) It can be used for both long and short-term trading. (5) You always know your entry price well ahead of execution. (6) Losses can be kept controlled at relatively low levels.
So look into this interesting trading technique yourself (or have it checked out by other expert technicians) and tell me what YOU think about it.
This approach has certainly not been exhaustively checked out against a large number of tradeables, however it has been subsequently researched over an extended period of time. It appears to have genuine promise as a mechanical trading system. Two of the chart studies I used as examples were of stock indices; one was of March Sugar, which I used just for grins, folks, since my success trading THAT contract is legendary among the other locals who knew me back then! (How to lose a quarter of a million dollars in less than three months)...
In my next comment, I intend to labor the point with further consideration of what I shall codify as "Envelope Trading." Hang in there.
Posted by: Bob Hansell | March 2, 2006 12:18 PM | Permalink to Comment