Moving Average Crossovers May Not Be The Best Entry Signals

There are many ways of using moving averages to trade but by far the most common method is to trade when a short-term moving average crosses over a longer term moving average. For example, if the 10-day MA crosses above the 30-day MA we typically assume that we have a new buy signal.

Let's stop for a minute and think about what exactly is occurring at the point of a crossover. When the 10-day MA and the 30-day MA are at the same price, the trend is not nearly as clear as it should be. What we are really observing at the crossover point is that the average of the last 30 prices is exactly the same as the average of the last 10 prices. If we are looking for trends to trade, this equal relationship of the two moving averages is not a reliable or logical indication of a trend. In an upward trending market the average prices over the last 10 days should be much higher than the average of the last 30 days. By implementing new trades at crossover points we are limiting our trading to points that may not clearly reflect what we should be doing. For best results in a trend-following system we want to be trading when the trend is clear and reliable; not when the trend is confused and questionable.

Instead of trading at crossovers we should be implementing our trades when the moving averages are parallel or when the short-term moving average is moving farther away from the longer-term moving average. Perhaps the short term MA should remain a minimum of some units of Average True Range above the longer term MA for several days. I believe that this procedure would give us more reliable and more frequent entry signals in the direction of the prevailing trend, which is exactly what we want. To identify the most reliable trends we want to see the slopes of various moving averages all moving steadily in the same direction and not crossing back and forth.

Take a look at a chart of any market with a strong trend. You will see that the moving averages are not crossing back and forth repeatedly. They will be moving in the same general direction in a more or less parallel fashion. Now look at a chart of a non-trending market. As this market moves sideways the moving averages will be crossing back and forth very frequently. Look at the implications of this simple examination of the charts. If we are trading the crossovers we will be trading most frequently in non-trending markets and trading most infrequently in strongly trending markets. Is that what we want? No, it's obviously not what we want. We want just the opposite. We want frequent entry opportunities in trending markets and we want to avoid as many trades as possible in non-trending markets.

The error in the logic of trading moving average crossovers also extends to some interpretations of MACD (Moving Average Convergence and Divergence) and DMI (Directional Movement Indicator). If we are looking at MACD we want to see both lines (each line reflects a moving average relationship) moving in the same direction. We don't want to see them crossing. When looking at DMI we want to see the Plus DI lines and the Minus DI lines moving in opposite directions and definitely not crossing. Remember, when the Plus DI and the Minus DI lines intersect it is telling us that the market is in balance and has no direction; the amount of upward and downward directional movement are exactly equal. What makes our favorite indicator, the ADX, so effective is that it rises only when the Plus DI and the Minus DI are moving in opposite directions and the distance between the two indicators is widening.

With a little thought and effort I'm sure we can design some reliable entry signals that are based on moving averages but avoid the typical crossover signals. For example we could measure the slope of several moving averages and when all the averages slope upward we would have a buy signal.

We could also measure the distance between several moving averages and implement our trades when the averages are all headed in the same direction but start getting farther apart. This procedure would give us a series of entry signals within the same original trend. This should provide an excellent entry and re-entry strategy.

Contradictions in using ADX

In our last article we described how the ADX works best when a move develops out of a basing pattern. Someone sent us a very courteous email questioning this strategy and reminded us of our "25 X 25"Bond system where we want the ADX to be above 20 before looking for an entry. He commented that by the time the ADX gets to 20 any move out of a base pattern may be over. He is absolutely right and I can see how there may appear to be some confusion on how the ADX should be applied. But there really isn't any contradiction if you understand that what we are trying to do with the ADX is very different in the two examples.

In the "25 X 25"strategy the ADX is used as an important "setup"condition that tells us when the trend is strong enough that we can confidently buy on weakness. However, when we describe the basing pattern strategy we are using the ADX as the actual entry trigger to buy on strength. There is a big difference in the buy on strength and buy on weakness strategies. In the basing pattern strategy a low level of the ADX is preferred because the rise in the ADX is the trigger. If the ADX is at 12 and starts rising we very well could miss the majority of the move if we waited for it to reach 20. With the ADX already at 20 or higher it might only be safe to buy on dips and of course that is exactly what the "25 X 25"bond strategy does.

To sum things up: there is no contradiction. To catch a move out of a base you should enter as soon as the ADX starts rising. Just compare today's ADX with yesterday's ADX and the faster it is rising the better. At this point the the lower the level of the ADX the better because we are buying on strength and the ADX is our entry trigger.

System Results Update

By the way, our bond strategies have been making lots of money this year. Hope you are all trading them with real money. Back in February I had lots of critics calling and asking why we were offering long only bond systems when bonds were at 115 and that had to be the top. They said that they couldn't possibly go much higher than that and our long only results could not hold up in real trading. Now that the bonds are over 130 I am glad that we have the Serendipity system that does trade the short side because I suspect that we really are near the top. (Doesn't take a genius to make a dumb statement like that. I apologize.)

The Big Dipper system has been long since July and has huge open profits even after having to be rolled forward from the September contract into December. The "25 X 25"system is not doing bad either. When is the last time you had a free system make that much money for you? And last but not least, lets not forget the "Little Dipper". According to my recollection the last seven trades were all winners and it could be more because I can't remember the last loser. Not too bad for out of sample trading results. REMEMBER: PAST RESULTS ARE NOT NECESSARILY INDICATIVE OF FUTURE SUCCESS. I'm not just saying that because I have to. I really believe it and so should you. The bull market in bonds has made us look smarter than we really are. Bull markets do that.

Combining RSI and ADX

Now that I am spending seven hours a day doing trading for the new hedge fund I haven't had much time for research or writing new Bulletins. However a comment in one of the trading newsgroups that I monitor got me thinking about the potential benefits of combining our knowledge of RSI and ADX into a simple system. Both the ADX and RSI are valuable trading tools and a combination of the two would seem to offer some interesting possibilities. I like to use the RSI primarily as an indicator for buying on dips in an uptrend. The ADX is my primary indicator of trend strength.

Here are a few ideas on how the two indicators might compliment each other in a system that "knows" when to enter on strength and when to buys on dips. (I'm only going to use the long side for examples but the logic should apply to short trades as well.)

When the ADX is rising it usually indicates that a strong trend is underway. In many cases waiting for any sizeable dip would be costly because the market could run away and the dip entry would be too late to maximize our profits. In this case we must enter on strength. To make this idea into a simple trading rule we might state that if the ADX is rising (and we have some indication it is rising because an uptrend is underway) we will buy whenever the RSI is below some very high threshold like 85. This rule would give us a very prompt entry in most cases and the result would be almost identical to simply trading whenever the ADX is rising which seems to be a good idea. The RSI has little, if any, benefit in this situation except it might occasionally keep us from buying into an extremely overbought market where the RSI was above 85. In this case a slight delay on the entry might be prudent.

The RSI, however, can play a much more important role when the ADX is flat or declining. In this case the rule would be that when the ADX is not rising we should postpone our entry until the RSI is below some more typical threshold like 45 or 50. Since the ADX is not giving us a signal that the trend is unusually strong we would need some additional indicator to show that the market has some minimal amount of upward direction. Otherwise we would not be buying a dip within the framework of an uptrend. Something simple like an upward sloping 20-bar moving average might work in this application.

Now that we have combined the ADX and RSI for our entries we might also want to combine them for our exits. When a market is rising but the trend is not particularly strong any spike in the RSI represents a good opportunity to take a profit. For example when trading in stocks the 9-bar RSI rising above 75 or 80 often signals that a correction is imminent. If the market trend is not unusually strong we would probably be happy with taking our profit on strength rather than waiting to get stopped out on weakness. However if the ADX is rising we might want to risk a correction in hopes of riding the trend even further. In this case when the ADX was rising we would ignore the RSI signal to take our profit. However, once our patience has allowed us to accumulate a very substantial open profit we might be best served by acting on the next RSI signal and nailing down the big winner. Also, when the ADX is rising it would not make much sense to be buying at a high RSI level and also selling at a high RSI level. We would be in and out of our trades almost immediately. Therefore we need to ignore the RSI extremes until our profit has had a chance to accumulate.

In summary, the important concept to remember is that our knowledge of the ADX can make the RSI a much more useful trading tool. When the ADX is rising the RSI tends to get overbought and it can often remain overbought for a surprising length of time. On the other hand when the ADX is flat or declining any spike to the upside in the RSI is an opportunity to nail down a profit. Conversely, any spike to the downside can be a potentially profitable entry point.

Here is the logic of a simple little system based on this discussion. (Just the rules in text form, you will have to do your own coding.) The parameters selected have not been tested or optimized. For example the 20-day moving average is just a number I picked out of the air. This is enough information to get you started and you can vary the rules to make the system trade over whatever time frame you prefer.

Long Entries:

1. The 20-bar moving average must be rising.
2. If the ADX is rising (ADX today is 0.20 or more higher than yesterday) then buy if the 14 bar RSI is less than 85.
3. If the ADX is not rising (ADX today is not 0.20 higher than yesterday) then buy if the 14 bar RSI is less than 50. Here is where you can influence the frequency of trading. For more trades use a higher threshold like 60. For fewer trades use a lower threshold like 40.

Long Exits

1. If the ADX is not rising (ADX today is not 0.20 higher than yesterday) then sell (long exit) if the 9-bar RSI is greater than 75.
2. If the ADX is rising (ADX today is 0.20 or more higher than yesterday) and the open profit is greater than (pick some amount - maybe 4 ATRs or some unit of price) then sell if the 9-bar RSI is greater than 75.
3. You need some additional exit rule for the losing trades. Use your favorite loss-limiting exit or you might want to exit when the price goes below the 20-dat moving average or when the 20-day moving average turns down. (See entry rule 1.)

ADX Has its Limitations

(I'm referring to Welles Wilders Average Directional Index in case you are a "newbie".) After many years of extolling the virtues of the ADX in articles and lectures all over the world I have become closely associated with this indicator. That's fine with me and I don't mind being considered the resident expert on ADX. It is an excellent measure of trendiness and a good indicator to be linked with.

However, I think it is a mistake to try and over work or become too dependent on any one indicator. If you were going to build a house you would need more than one tool and you wouldn't try to do it with just a hammer. The same is true of building systems. The ADX can be a very valuable tool if used correctly but it has some major shortcomings that everyone should be aware of: We all know that the ADX is slow. This is because of all the smoothing in the formula. The basic ingredients are smoothed and then the results are smoothed again. For example I think it takes more than 30 bars of data to calculate a 14 bar ADX. This smoothing makes the ADX slow but there is an even greater problem than just the speed of the indicator. The logic of measuring directional movement makes the ADX very reliable at certain times and very unreliable at other times.

A rising ADX is a reliable indication of a trend when there has been an extended sideways period before the trend gets started. Before all the high tech computer mumbo jumbo we used to simply refer to this sideways period as a "basing pattern". The ADX is most effective when it begins to rise from a low level (low = 15 or less). This low level on the ADX indicates that there has been a basing pattern for a while. This interpretation is contradictory to those users of the ADX who want to see the ADX cross above a specified threshold (usually 20 or 25) to indicate that a trend is underway. This technique would make the ADX even slower and means you would be confirming a trend and entering your trade long after the basing pattern was broken. But even if you were late due to your method of interpreting the ADX, following the ADX after a base pattern is still quite reliable. The potential problem I want to bring to your attention in this article is the action of the ADX after major peaks and valleys.

The logic of the ADX is best visualized as measuring directional movement over a moving window of data on a bar chart. If we have sideways data in the window followed by recent trending data (lets think of rising prices but it could be the reverse), the rising prices would show directional movement relative to the sideways data at the beginning of our window. The ADX would promptly rise and call our attention to the fact that there is now a direction in prices that should continue for a while.

However, if the prices rise for an extended period and then begin to fall sharply (a typical scenario) we now have a window of data that shows rising prices followed immediately by falling prices. The ADX formula measures the rising prices in the window and compares them with the declining prices in the window. Because the two trends are about equal they cancel each other and the ADX does not detect any net directional movement. The ADX now begins to decline indicating that it is finding no net directional movement in the period measured by the window.

As the window moves forward, eventually the older rising price data falls outside the back of the window so that the window now contains only the more recent downward price movement. The ADX suddenly begins to rise rapidly because the data window at this point contains only one trend. The problem with this new signal is that the downward trend in prices has been underway for quite some time and only now has the ADX finally begun to rise. This is obviously not a good point to be entering a trade to the short side. We are probably nearer the end of the trend than the beginning.

Remember that the ADX works best after a basing period and is unreliable after a "V" bottom or top.
That"s all for now. I'll continue this discussion of the ADX in our next bulletin which should be out very soon.

ADX Has its Limitations

(I'm referring to Welles Wilders Average Directional Index in case you are a "newbie".) After many years of extolling the virtues of the ADX in articles and lectures all over the world I have become closely associated with this indicator. That's fine with me and I don't mind being considered the resident expert on ADX. It is an excellent measure of trendiness and a good indicator to be linked with.

However, I think it is a mistake to try and over work or become too dependent on any one indicator. If you were going to build a house you would need more than one tool and you wouldn't try to do it with just a hammer. The same is true of building systems. The ADX can be a very valuable tool if used correctly but it has some major shortcomings that everyone should be aware of: We all know that the ADX is slow. This is because of all the smoothing in the formula. The basic ingredients are smoothed and then the results are smoothed again. For example I think it takes more than 30 bars of data to calculate a 14 bar ADX. This smoothing makes the ADX slow but there is an even greater problem than just the speed of the indicator. The logic of measuring directional movement makes the ADX very reliable at certain times and very unreliable at other times.

A rising ADX is a reliable indication of a trend when there has been an extended sideways period before the trend gets started. Before all the high tech computer mumbo jumbo we used to simply refer to this sideways period as a "basing pattern". The ADX is most effective when it begins to rise from a low level (low = 15 or less). This low level on the ADX indicates that there has been a basing pattern for a while. This interpretation is contradictory to those users of the ADX who want to see the ADX cross above a specified threshold (usually 20 or 25) to indicate that a trend is underway. This technique would make the ADX even slower and means you would be confirming a trend and entering your trade long after the basing pattern was broken. But even if you were late due to your method of interpreting the ADX, following the ADX after a base pattern is still quite reliable. The potential problem I want to bring to your attention in this article is the action of the ADX after major peaks and valleys.

The logic of the ADX is best visualized as measuring directional movement over a moving window of data on a bar chart. If we have sideways data in the window followed by recent trending data (lets think of rising prices but it could be the reverse), the rising prices would show directional movement relative to the sideways data at the beginning of our window. The ADX would promptly rise and call our attention to the fact that there is now a direction in prices that should continue for a while.

However, if the prices rise for an extended period and then begin to fall sharply (a typical scenario) we now have a window of data that shows rising prices followed immediately by falling prices. The ADX formula measures the rising prices in the window and compares them with the declining prices in the window. Because the two trends are about equal they cancel each other and the ADX does not detect any net directional movement. The ADX now begins to decline indicating that it is finding no net directional movement in the period measured by the window.

As the window moves forward, eventually the older rising price data falls outside the back of the window so that the window now contains only the more recent downward price movement. The ADX suddenly begins to rise rapidly because the data window at this point contains only one trend. The problem with this new signal is that the downward trend in prices has been underway for quite some time and only now has the ADX finally begun to rise. This is obviously not a good point to be entering a trade to the short side. We are probably nearer the end of the trend than the beginning.

Remember that the ADX works best after a basing period and is unreliable after a "V" bottom or top.
That"s all for now. I'll continue this discussion of the ADX in our next bulletin which should be out very soon.

The Parabolic Trigger for V Tops and Bottoms

Our previous article about using the ADX for V shaped tops and bottoms was surprisingly well received. We had a great deal of very favorable feedback from our members who experimented with it. This very valuable pattern seems to do an excellent job of spotting major turning points in almost any market from Palladium to Natural Gas to Soybeans or even Lumber. This pattern seems to work extremely well in almost all futures, stocks and even the hard to trade stock indexes.

Much like a kid with a new toy, we've been having fun scanning through our charts and finding all the important signals that have been generated. For example, when looking at the stock index charts we had some very timely and important signals that the strong bear market in stocks was finally reversing. Let's review very briefly the conditions that create the pattern we are looking for.

REVIEW OF SETUPS and TRIGGERS: For those of you who are new to our work, we strongly recommend a two step process for entries. The first step is to identify some "setup" conditions that tell us that an entry is near. The second step is the "trigger" that tells us we must enter the trade NOW.

Just to refresh your memory from the previous Bulletin, let me review the "setup" conditions that we are looking for. Remember that we are trying to anticipate important "V" shaped reversal patterns. We want to be able to trade as near as possible to major tops and bottoms. As most of you are aware, a major directional price move will cause the ADX to rise to a high level. Depending on the direction of the price movement, either the Plus DI or the Minus DI will also move to an unusually high level. As the market peaks the DI will begin to decline while the ADX is flat or still rising. Near the top (or bottom) the ADX will become the highest line and will be above both the Plus DI and the Minus DI lines. This is our "setup" and alerts us that an important change in direction is likely in the very near future. The relationship of the three lines with the ADX being the highest tells us that there has already been a very extended price move that is running out of gas.

FINDING A TIMELY TRIGGER: While studying the charts using our new ADX pattern we found that our setup conditions often occurred early and that our DMI triggers were sometimes a little bit late. We don't mind having the setup conditions occur early. After all, lead indicators are rare and very hard to find. However to make this entry pattern even more exciting we thought we would see if we could make the triggers occur sooner.

Now that we have our lead indicator in place we want to find a timely entry trigger that gets us started in the direction of the new trend that should just be getting started. In Bulletin 45 we suggested that the crossing of the Plus DI and Minus DI lines could be the entry trigger. Although this method is acceptable and produces excellent results we observed that there might be room for further improvement. In many cases, by the time the Plus and Minus DI have crossed some profits in the new direction have already been left behind.

After some trial and error we found that the Parabolic indicator did just what we wanted. We believe we can use the Parabolic indicator instead of the DMI crossovers to provide much more timely entry triggers.

We have never liked the Parabolic stop and reverse (SAR) method as an independent trading system which was the intent of J. Welles Wilder, its originator. However we do like to use the Parabolic indicator for exits. As a system we find that the Parabolic reversal points occur much too frequently and this reversal system would drive us crazy with far too many false changes in direction. However the features of the Parabolic indicator that make it useful as an exit strategy are exactly what makes it the timely trigger we need for our ADX reversal pattern.

The Parabolic indicator accelerates steadily as the prices trend until the reversal points are very, very close to the peak of the move. The stronger the trend the closer the Parabolic gets to the prices. That is exactly what we want. When the Parabolic indicator is close to the prices and we have fulfilled our ADX setup conditions we are all set for an outstanding trade. Even a very small countertrend move will now quickly cross the Parabolic and signal our timely entry.

AN IDEAL ENTRY PLUS AN ADD POINT: We view the marriage of the ADX setup with the Parabolic entry trigger as an ideal combination. The entries now occur in a much more timely fashion than when we relied on the DI crossovers for our trigger. In fact, once the Parabolic has been crossed we can use the DI crossover as a confirmation and add to our position. I don't normally believe in pyramiding positions but in this case we are trading a very reliable pattern that is designed to identify a major reversal in direction, so I think that adding to positions early is a very good strategy.

As you can probably tell, I really like this ADX and Parabolic entry technique and I think that we have a lot of good concepts working for us here. We have an early setup, a timely trigger and now we can use the delayed confirmation of the DI crossovers as a point to pyramid the position. You can't ask for much more than that for an entry strategy that does a great job of catching big moves early. Take a look at this pattern on your favorite market and give me your comments.

ADX for V Tops and V Bottoms

We have often described how the ADX (J. Welles Wilder's Average Directional Index) can be a useful tool for measuring the strength of trends. (Please review: Contradictions in using ADX and ADX Has it's Limitations if you are not familiar with our recommended use of ADX.) To briefly summarize our previous advice, we have found that when the ADX begins to rise it is telling us that a strong trend is developing.

We have often described how the ADX (J. Welles Wilder's Average Directional Index) can be a useful tool for measuring the strength of trends. (Please review: Contradictions in using ADX and ADX Has it's Limitations if you are not familiar with our recommended use of ADX.) To briefly summarize our previous advice, we have found that when the ADX begins to rise it is telling us that a strong trend is developing. A rising ADX has proven to be a particularly reliable indicator after a market has been going sideways for a while and then begins to trend. For best results, the ADX should begin its rise from a low level (less than 15 or 20) because the low level of the ADX indicates that a sideways basing pattern has been formed. Most of our applications of the ADX strategy have been predicated on finding these highly profitable patterns where a trend suddenly emerges after an extended sideways period.

Unfortunately not all trends begin with a sideways pattern. There are many V tops and V bottoms that our rising ADX strategy fails to capture. In a V pattern the ADX rises and then peaks out and declines. The ADX does not begin rising again in time to catch the change in direction in a timely fashion. By the time the ADX falls and then begins to rise again a major portion of the new trend will have already been completed. As we have pointed out in our previous Bulletins, any entry on a rising ADX that was not preceded by an extensive sideways period is not a very reliable pattern.

Recently in our research on using the ADX for trading stocks we have observed another ADX pattern that we believe shows great promise. This new ADX pattern signals very timely entries that allow us to profit from possible tops and bottoms that are V shaped.

Here is how these V shaped top and bottom patterns can easily be recognized:

1. Make sure that your plot of the ADX also includes the plot of the Plus DI and the Minus DI. The pattern begins when the ADX is above both the Plus DI and the Minus DI. Most often when the ADX is above both the Plus and Minus DI the ADX will be at a high level, perhaps greater than 30 or 35. The high level of the ADX indicates that the previous trend was a very strong one. Now we are going to try and catch the reversal of that strong trend.
2. With the ADX at a high level and declining, look for a crossing of the Plus DI and Minus DI. If the Minus DI crosses above the Plus DI it indicates that a strong up market has ended and weakness has set in. If the Plus DI crosses above the Minus DI it indicates that a strong downtrend has ended and a new uptrend can be expected.
3. These reversal patterns should be entered only as the market moves in the new direction. (We suggest that you use stops for entry triggers.) Once you have entered the trade you should expect a substantial move in the new direction.
4. Be sure to use a stop loss at the recent low or high of the previous trend. Be willing to make more than one attempt to catch the new trend. (Sometimes the Plus and Minus DI will cross back and forth more than once before the new trend emerges.)

We have found that this simple pattern identifies major changes in direction in almost any market. However you should be aware that the change of direction pattern we have described is not as reliable as the typical rising ADX pattern that starts with a basing action. However the trades that do work are exceptionally profitable and we know of no other method that is as timely at catching major changes in direction. Most traders take a great deal of personal satisfaction in quickly recognizing major changes in direction. This simple entry method can produce some truly outstanding trades and provides a welcome change from typical trend following strategies.

Our Phoenix Bond system uses this technique to identify major bottoms in the bond market. The same system also spots bottoms in individual stocks. To catch major tops we simply reverse the logic. Just put up some charts with the ADX and look for the pattern we have described in this Bulletin. We think you will be very surprised at its accuracy. Give us your comments and observations on the Forum.