Traders today have access to highly sophisticated computerized trading software at reasonable prices. With just a click of the mouse, most trading packages allow you to overlay numerous technical studies, adjust default settings and change timeframes. The novice trader normally is both awed and overwhelmed with these choices and can spend hours creating complex charts in exotic colors.
However, what many traders may not know is that most, if not all, of the available technical indicators were developed years ago on end-of-day data. Many were designed using only a hand-held calculator! The input was obtained from daily, weekly and monthly data. Traders would spend the evening if they had access to end-of-day data – or the weekends if they only had access to the weekly “chart books” – making these calculations and then applying the results to the next day’s or week’s trading plan.
Like many other things in our society today, the computer changed all of that. Indeed end-of-day data can be analyzed in countless different ways. First with satellites and then with the growth of the Internet, data now is accessed in real time, as well. This allows traders to apply these same technical indictors to data in timeframes that are as fast as the data can be released. And this is precisely what many traders do. However, this is exactly why many traders fail.
In this article, we’ll cover three of the most common technical tools: moving average convergence divergence (MACD), stochastics and relative strength (RSI). This includes review of the concepts underlying the indicators and then an illustration of a typical trading day, showing how these indicators would have fared in real time. This article does not address the idea that technical indictors may be programmed in real time for the purpose of computer-generated signals, instead assuming that these tools are in the hands of a discretionary trader.
Tick Tock, Tick Tock
One of the hardest concepts for day traders to understand is time. Most traders find their way to day trading through the stock market. Usually they’ve invested money in mutual funds or stocks. Perhaps they’ve learned the important ideas of P/E ratios, dividend growth over time and valuation. Maybe they’ve even dabbled in more exciting growth stocks and have learned about relative strength as it applies to stock picking. Then they discover day trading and have the mistaken belief that it can’t be too different from what they’ve already learned – only just a bit faster. That’s the first big mistake.
Here’s where the computer-trading package is purchased and the high speed Internet connection arranged. This is where the trader starts madly clicking his way through the dozens of technical analysis modules, and that’s where the problems really begin.
The second fallacy is that those same tried and true technical indicators developed on daily and weekly data will work on one-minute data. They don’t. Stop and think about what the trading day is really like. Even if you are alert and fresh, the complexity of the information and the speed with which it is delivered is overwhelming. Face it – even the sharpest person out there is only capable of synthesizing a limited amount of data in any given moment.
Once the data is obtained, analyzed and viewed graphically, the trader still needs to compare the information to price points and trading rules he may have developed, make an informed decision and then enter a trade. All the while, time is ticking by and money is on the line, offering even greater emotional factors with which to contend. But all this time, the data keeps flowing; the charts keep updating; and you are faced with yet more information to synthesize.
Without even touching on the various indicators, it should already be clear how difficult, if not impossible, this challenge is. Most intelligent humans are hard-pressed to remember a seven-digit phone number, a grocery list or the name of a new acquaintance. Before you shrug this off, remember – it’s a completely different matter to calculate relative strength, MACD and stochastics and then draw trend lines when the market is closed.
Relative Strength
For the purpose of this article, I decided to examine the results using the “default” settings for all three indicators. First off, relative strength as a technical tool was developed by J. Welles Wilder, Jr. (New Concepts in Technical Trading Systems, 1978). Although one of the most commonly followed indicators, it is misunderstood. In stock trading, the term relative strength implies a ratio or comparison of stock to stock or stocks to various industry groups. Similar use is made in classic commodity analysis, such as comparing various delivery months or markets to benchmarks like the CRB commodity index.
In day trading, though, relative strength is an altogether different construct. The original idea was to smooth out the fluctuations in momentum indicators that result from “drop off” of distant extreme price data and to provide for a more constant band for comparison of momentum over time. The formula is based on a ratio of the average of up closes to the average of down closes over a certain number of days (14 is the default). That product is then applied to a simple formula, which allows the result to oscillate between zero and 100. Typically, a number greater than 70 is considered “overbought” while a number less than 30 is “oversold.”
Simple enough? Traders use this information to take a buy signal on a cross back up over 30 and a sell signal on a cross back below 70. Others use the “50” line to generate or confirm a signal. Still others use divergence from price as the key to pulling the trigger. The RSI is an invaluable tool for analyzing end-of-day or end-of-week data, and when extreme levels are reached and divergences set up, it is one of the best. However, as an intraday tool it leaves much to be desired.
Next Up: MACD
In 1977 Gerald Appel devised the moving average convergence divergence, or MACD, the popular technical indicator that determines the direction of a trend in an individual security or an entire index. It remains at the core of his analysis. Again his work was based on daily and weekly data, not one-minute data. The idea behind this tool is to generate a value, which is the difference between two exponential moving averages, and compare that to a nine-period exponential average of that value to generate a signal. The default values of the two moving averages are 12 and 26. The result is an excellent moving average crossover system combined with an oscillator. In this way, a trader can see not only a crossover, but also how it occurs in the context of an oversold or overbought environment.
As an end-of-day or end-of-week tool, it is an excellent performer. However, as an intraday tool using one-minute data, it will generate multiple confusing signals. As with RSI, there are many ways to interpret the MACD including overbought and oversold crossovers and crossing of a zero line, but the best results come from spotting divergence from price. This is easy to do with end-of-day data and a ruler or chart tool in hand. On the fly it is a challenge, if not impossible.
Stochastics
Our last oscillator was developed by George Lane and was built on the idea that as a market goes up, prices should close at the upper end of the range, and when prices decline, they should be nearer the lower end of their range. Thus, stochastics tries to determine where price is relative to its price range over a given period of time. A trader once again can see how this will work well if one is able to spot divergence between stochastics and price.
Calculation of stochastics involves a ratio of the closing value and the 14-period low to the difference between the 14-period high and the 14-period low. The results, as in the RSI, use a zero to 100 scale. Lane suggested a second line which is a three-period moving average of the first result, which allows for a “signal” to be generated, much like the nine-period signal-generator for MACD. Volumes have been written about the nuances of the various crossover signals. Suffice it to say that like the other indicators already discussed, overbought and oversold crossovers generate signals. In general, a cross down through the 80 line indicates a sell, and a cross up through the 20 line indicates a buy. Traders will also look for a divergence from price to generate signals. Stochastics has become an extremely popular day trading indicator, yet conventional wisdom holds that 90 percent of all traders lose money. Is there a connection?
A Typical Day…
The following charts are taken from August 16, 2004. The day was chosen randomly and is an excellent example, as it turned out to be a fairly typical day. Remember what we are tying to show are the problems a trader might face by relying on what are supposed to be standard indicators. Default settings were used and the time frame is a one-minute chart. Time shown is Central time. First, we’ll show the indicators separately and then see if the indicators would have worked better when employed together as most traders use them. See how you would have interpreted the charts.
As illustrated in Figure 1, the market opened very strong, leaping higher from the opening bell. We’ll assume that the trader was unable to act until after 9:00 a.m. According to MACD interpretation, a long trade is entered at 1071.50 at 9:08 a.m. with an exit at 1075. So far, so good. However, look at the chart and see if you can figure out what you would have done next – in real time, not in hindsight. Would you go short? Short at 10:15? Break even. Short at 10:58? Break even. Long at 12:13? Break even. If a trader were dutifully trying to trade the MACD signals as they appeared live on his screen, each of these trades would have been taken and would have led to whipsaw and frustration.
However, what many traders may not know is that most, if not all, of the available technical indicators were developed years ago on end-of-day data. Many were designed using only a hand-held calculator! The input was obtained from daily, weekly and monthly data. Traders would spend the evening if they had access to end-of-day data – or the weekends if they only had access to the weekly “chart books” – making these calculations and then applying the results to the next day’s or week’s trading plan.
Like many other things in our society today, the computer changed all of that. Indeed end-of-day data can be analyzed in countless different ways. First with satellites and then with the growth of the Internet, data now is accessed in real time, as well. This allows traders to apply these same technical indictors to data in timeframes that are as fast as the data can be released. And this is precisely what many traders do. However, this is exactly why many traders fail.
In this article, we’ll cover three of the most common technical tools: moving average convergence divergence (MACD), stochastics and relative strength (RSI). This includes review of the concepts underlying the indicators and then an illustration of a typical trading day, showing how these indicators would have fared in real time. This article does not address the idea that technical indictors may be programmed in real time for the purpose of computer-generated signals, instead assuming that these tools are in the hands of a discretionary trader.
Tick Tock, Tick Tock
One of the hardest concepts for day traders to understand is time. Most traders find their way to day trading through the stock market. Usually they’ve invested money in mutual funds or stocks. Perhaps they’ve learned the important ideas of P/E ratios, dividend growth over time and valuation. Maybe they’ve even dabbled in more exciting growth stocks and have learned about relative strength as it applies to stock picking. Then they discover day trading and have the mistaken belief that it can’t be too different from what they’ve already learned – only just a bit faster. That’s the first big mistake.
Here’s where the computer-trading package is purchased and the high speed Internet connection arranged. This is where the trader starts madly clicking his way through the dozens of technical analysis modules, and that’s where the problems really begin.
The second fallacy is that those same tried and true technical indicators developed on daily and weekly data will work on one-minute data. They don’t. Stop and think about what the trading day is really like. Even if you are alert and fresh, the complexity of the information and the speed with which it is delivered is overwhelming. Face it – even the sharpest person out there is only capable of synthesizing a limited amount of data in any given moment.
Once the data is obtained, analyzed and viewed graphically, the trader still needs to compare the information to price points and trading rules he may have developed, make an informed decision and then enter a trade. All the while, time is ticking by and money is on the line, offering even greater emotional factors with which to contend. But all this time, the data keeps flowing; the charts keep updating; and you are faced with yet more information to synthesize.
Without even touching on the various indicators, it should already be clear how difficult, if not impossible, this challenge is. Most intelligent humans are hard-pressed to remember a seven-digit phone number, a grocery list or the name of a new acquaintance. Before you shrug this off, remember – it’s a completely different matter to calculate relative strength, MACD and stochastics and then draw trend lines when the market is closed.
Relative Strength
For the purpose of this article, I decided to examine the results using the “default” settings for all three indicators. First off, relative strength as a technical tool was developed by J. Welles Wilder, Jr. (New Concepts in Technical Trading Systems, 1978). Although one of the most commonly followed indicators, it is misunderstood. In stock trading, the term relative strength implies a ratio or comparison of stock to stock or stocks to various industry groups. Similar use is made in classic commodity analysis, such as comparing various delivery months or markets to benchmarks like the CRB commodity index.
In day trading, though, relative strength is an altogether different construct. The original idea was to smooth out the fluctuations in momentum indicators that result from “drop off” of distant extreme price data and to provide for a more constant band for comparison of momentum over time. The formula is based on a ratio of the average of up closes to the average of down closes over a certain number of days (14 is the default). That product is then applied to a simple formula, which allows the result to oscillate between zero and 100. Typically, a number greater than 70 is considered “overbought” while a number less than 30 is “oversold.”
Simple enough? Traders use this information to take a buy signal on a cross back up over 30 and a sell signal on a cross back below 70. Others use the “50” line to generate or confirm a signal. Still others use divergence from price as the key to pulling the trigger. The RSI is an invaluable tool for analyzing end-of-day or end-of-week data, and when extreme levels are reached and divergences set up, it is one of the best. However, as an intraday tool it leaves much to be desired.
Next Up: MACD
In 1977 Gerald Appel devised the moving average convergence divergence, or MACD, the popular technical indicator that determines the direction of a trend in an individual security or an entire index. It remains at the core of his analysis. Again his work was based on daily and weekly data, not one-minute data. The idea behind this tool is to generate a value, which is the difference between two exponential moving averages, and compare that to a nine-period exponential average of that value to generate a signal. The default values of the two moving averages are 12 and 26. The result is an excellent moving average crossover system combined with an oscillator. In this way, a trader can see not only a crossover, but also how it occurs in the context of an oversold or overbought environment.
As an end-of-day or end-of-week tool, it is an excellent performer. However, as an intraday tool using one-minute data, it will generate multiple confusing signals. As with RSI, there are many ways to interpret the MACD including overbought and oversold crossovers and crossing of a zero line, but the best results come from spotting divergence from price. This is easy to do with end-of-day data and a ruler or chart tool in hand. On the fly it is a challenge, if not impossible.
Stochastics
Our last oscillator was developed by George Lane and was built on the idea that as a market goes up, prices should close at the upper end of the range, and when prices decline, they should be nearer the lower end of their range. Thus, stochastics tries to determine where price is relative to its price range over a given period of time. A trader once again can see how this will work well if one is able to spot divergence between stochastics and price.
Calculation of stochastics involves a ratio of the closing value and the 14-period low to the difference between the 14-period high and the 14-period low. The results, as in the RSI, use a zero to 100 scale. Lane suggested a second line which is a three-period moving average of the first result, which allows for a “signal” to be generated, much like the nine-period signal-generator for MACD. Volumes have been written about the nuances of the various crossover signals. Suffice it to say that like the other indicators already discussed, overbought and oversold crossovers generate signals. In general, a cross down through the 80 line indicates a sell, and a cross up through the 20 line indicates a buy. Traders will also look for a divergence from price to generate signals. Stochastics has become an extremely popular day trading indicator, yet conventional wisdom holds that 90 percent of all traders lose money. Is there a connection?
A Typical Day…
The following charts are taken from August 16, 2004. The day was chosen randomly and is an excellent example, as it turned out to be a fairly typical day. Remember what we are tying to show are the problems a trader might face by relying on what are supposed to be standard indicators. Default settings were used and the time frame is a one-minute chart. Time shown is Central time. First, we’ll show the indicators separately and then see if the indicators would have worked better when employed together as most traders use them. See how you would have interpreted the charts.
As illustrated in Figure 1, the market opened very strong, leaping higher from the opening bell. We’ll assume that the trader was unable to act until after 9:00 a.m. According to MACD interpretation, a long trade is entered at 1071.50 at 9:08 a.m. with an exit at 1075. So far, so good. However, look at the chart and see if you can figure out what you would have done next – in real time, not in hindsight. Would you go short? Short at 10:15? Break even. Short at 10:58? Break even. Long at 12:13? Break even. If a trader were dutifully trying to trade the MACD signals as they appeared live on his screen, each of these trades would have been taken and would have led to whipsaw and frustration.
Relying on RSI in Figure 2 would not have allowed us to enter the early profitable MACD trade. Instead, we would have entered short at 9:20 a.m. as the signal crossed below 70 from overbought, and we would have stopped out. We enter short again just before 9:30 only to see RSI move lower and prices move steadily higher! Even if we had used price divergence as our tool, it would have led to a couple of points on the short at 10:20 as we broke below 50. But ask yourself if you would have stayed short at 10:45 as RSI rose back above 50, or if you would have held on until 11:00. Why? The rest of the day shows a series of unclear waves, unless you can figure out a trading rule that would cover the short at 12:20. I can’t.This indicator, stochastics, seen in Figure 3, is the most useless of them all in my opinion. Remember, we never know ahead of time what kind of trading day it’s going to be.
We want a reliable system where we can turn on the computer and begin generating effective signals in real time. Would you go short at 9:10 or 9:25 a.m.? Would you go long from 9:40 to 9:45 and then miss the bulk of the subsequent move? After that, I see more than 15 “typical” sell signals. Which one would you have taken? I see around seven buy signals. Which ones would you have taken? Which ones would have worked?
Combining the Indicators
Most traders who come to me for help are routinely watching more than three charts at a time. See Figure 4 to see what August 16 looks like when all three of these indicators are placed together. Around 10:20 a.m., there may be a sell signal, but what criteria would you have used to exit? By 10:20, the stochastic was already oversold and the MACD was turning up. If you had the conviction to stay short through the RSI drop, you would have made two points.
Combining the Indicators
Most traders who come to me for help are routinely watching more than three charts at a time. See Figure 4 to see what August 16 looks like when all three of these indicators are placed together. Around 10:20 a.m., there may be a sell signal, but what criteria would you have used to exit? By 10:20, the stochastic was already oversold and the MACD was turning up. If you had the conviction to stay short through the RSI drop, you would have made two points.
There might be a good sell signal at 1:15 p.m. following the spike in the RSI and confirmed by the MACD while the stochastic may have already been in a sell. It’s a lousy trade. At about 2:00, one would have been confused by the uptick in MACD. Would you have covered and gone long at 2:30? It’s hard for my eyes to move up and down through the price bars and the indicators simultaneously in an end-of-the-day chart! Imagine trying to synthesize this on a computer screen in real time!
Frustrated yet? The real kicker is that the Dow was up around 120 points that day, and there was a 12-point move in the S&P 500. Not one of these famous indicators would have captured any meaningful piece of that move. While not impossible to use, (and I am sure there is someone who has a “system”) the point is if you have not been able to make money trading off of one-minute charts, the problem is not that you don’t know how to read the charts right yet. The problem is that they don’t work.
This review does not intend to take away from these excellent technical indicators. The point is that it is naïve to think a trader can simply switch to an intraday timeframe and obtain the same benefits from indicators that were designed for something
altogether different.
What’s a Day Trader to Do?
So how do we trade? That’s a topic for another time, but stop and think about all the successful traders who never look at a computer screen. What about all the floor traders? Surely there is other information we can use. We use all of these indicators on end-of-day data as they were meant to be used, giving us a good feel for the market trend and strength. Using that data, a trading plan can be developed for each day, including price support and resistance. In addition, we use supporting indicators during the day – such as breadth, volatility and volume – to assist with trading decisions.
Don’t be fooled by the allure of sophisticated trading software. Traders who choose to use intraday technical analysis should make sure to understand how the indicators are calculated and what the output represents. It’s easy to forget that even with sideways price movement, these oscillators will continue to move between zero and 100! During periods of low volatility they will still generate buy and sell signals that can churn your account and fray your nerves. Some of the most successful S&P day traders I know follow price, Tick, Trin, VIX and volume and never look at a chart during the day. There is a place for technical analysis: at night and away from the action.
Frustrated yet? The real kicker is that the Dow was up around 120 points that day, and there was a 12-point move in the S&P 500. Not one of these famous indicators would have captured any meaningful piece of that move. While not impossible to use, (and I am sure there is someone who has a “system”) the point is if you have not been able to make money trading off of one-minute charts, the problem is not that you don’t know how to read the charts right yet. The problem is that they don’t work.
This review does not intend to take away from these excellent technical indicators. The point is that it is naïve to think a trader can simply switch to an intraday timeframe and obtain the same benefits from indicators that were designed for something
altogether different.
What’s a Day Trader to Do?
So how do we trade? That’s a topic for another time, but stop and think about all the successful traders who never look at a computer screen. What about all the floor traders? Surely there is other information we can use. We use all of these indicators on end-of-day data as they were meant to be used, giving us a good feel for the market trend and strength. Using that data, a trading plan can be developed for each day, including price support and resistance. In addition, we use supporting indicators during the day – such as breadth, volatility and volume – to assist with trading decisions.
Don’t be fooled by the allure of sophisticated trading software. Traders who choose to use intraday technical analysis should make sure to understand how the indicators are calculated and what the output represents. It’s easy to forget that even with sideways price movement, these oscillators will continue to move between zero and 100! During periods of low volatility they will still generate buy and sell signals that can churn your account and fray your nerves. Some of the most successful S&P day traders I know follow price, Tick, Trin, VIX and volume and never look at a chart during the day. There is a place for technical analysis: at night and away from the action.
No comments:
Post a Comment