THE LOGIC BEHIND PARABOLIC SAR
In several recent articles in this section of the newsletter, we have explored the importance of RSI for swing traders. RSI is one of J. Welles Wilder Jr.'s most famous indicators. Starting with today's issue I going to begin a series of detailed discussions of Parabolic SAR, another of Wilder's well-known indicators. When applied correctly, Parabolic SAR can greatly enhance your swing trading decision-making.

Parabolic SAR stands for Parabolic Stop and Reverse. The indicator's name derives from the shape it takes on the chart. When SAR first gives a buy or sell signal, its slope is relatively flat. However, the longer one is in a successful long trade, the more steeply the SAR dots rise. Eventually they take the shape of a parabola, rising at virtually a 90% angle until at some point price and SAR point meet. At that point in time it is usually wise to exit the trade. (Invert the description above for a short trade.) 

Wilder was primarily a commodities trader, not an equities trader. On a relative basis, he therefore had far fewer "tradeables" from which to choose. His SAR system dealt with this limitation by always being in the market. The trader is either long or short, but is never out. When the SAR point and price meet, the long or short position is reversed and the trader takes the other side.

With this historical perspective in mind, equity traders might therefore want to adapt SAR and be more selective about which signals they follow. As a general principle, it is wise to listen to SAR signals only when they are given in the direction of an existing trend. The trend may be established by using a key moving average relevant to your time frame, a trendline or Wilder's ADX indicator. SAR and MACD also combine well. In future articles in this series I will explore using SAR with other indicators.

The mathematical calculation behind SAR is very detailed. Three keep concepts are necessary for the calculation: what Wilder calls a "significant point," a daily "extreme point," which occurs when the stock makes a new high or low, and an "acceleration factor." To keep things relatively simple I am going to describe a trade from the long side; the instructions are reversed for a short side trade. The example I will use is ChevronTexaco (CVX, $56.74), which is currently in a very strong Minor uptrend.

In late November, CVX peaked at a high of $55.21. Several days later, the stock declined enough to trigger an SAR sell signal. At that point in time traders following SAR would have reversed positions from long to short. This is seen by the fact that the dots, which held below CVX's share price in late November, reversed and were then above the price action.

In very early January, CVX hit a low of $50.40. This low constitutes what Wilder calls the "significant point" -- the extreme hit while still in the trade from the short side. Note that at the "significant point" one was still in the trade from the short side.

The buy signal was triggered three days later. Again, observe how the dots switch direction from above price to below. Note too how the first couple of dots are nearly on an identical level with one another. The indicator is designed so that one is not whipsawed out of the trade near its beginning as the trend reverses.

The acceleration factor is what makes this indicator unique and is what gives SAR its parabolic shape. The acceleration factor increases each day the underlying stock makes a new high. It starts at 2%, and if enough new highs are generated it can ultimately expand to a maximum of 20%. The acceleration factor is multiplied by the difference between the high made on a specific day and the SAR point on the previous day. That amount is added back to the previous day's SAR, which causes the indicator to rise each day. If that computation sounds complex, then you're not too far off. Let's be happy the computer does the calculating for us!

SAR can never go backwards. The stop is always trailing the price action. As the trend matures the dots come closer and closer to price until the position, as stated before, is reversed.

In the chart of Chevron, SAR signaled an entry less than $1 from the $50.40 Significant Point low. It then remained "safely" below the rising trendline in the early and middle stages of the Minor uptrend. As of the most recent close, SAR was at $53.92. Note how the spaces between the dots were beginning to grow rapidly as the acceleration factor increased by 2% each time a new high was made (or in Wilder's terms, an "extreme point" was reached). As of Friday's close, CVX was still well above the SAR price and the swing trader who took the SAR signal just under $51 would have been riding a highly profitable position. 

OPTIMIZING PARABOLIC SAR
Throughout all of my educational articles, I am always careful to emphasize the importance of not making trading decisions based on any one indicator. Rather, the swing trader should apply multiple indicators to the chart. Each indicator sends its own message and gives the swing trader slightly different messages. Synthesizing all these messages together allows for effective decision making.

The advice to not use one stand-alone indicator is particularly true with Parabolic SAR, which provides a buy or sell signal that can last for many periods. Entering a position based solely on an SAR buy signal doesn't safeguard against the possibility that the stock could already be in an extended trend and close to reversing direction. One obvious answer to that potential problem would be to enter a trade on the very first change in the direction of the dots. There are times, however, when that can also be a dangerous strategy. I have seen many examples of what I call the "parabolic whoops." (The hourly chart of the S&P 500 above is a prime example.) This takes place when there is a trend in one direction, then Parabolic SAR gives a signal to enter from the other direction. After just one or two dots, the original trend resumes. If you follow SAR alone, you are whipsawed.

Part of the answer of how to optimize SAR lies in paying close attention to its visual signals. At the beginning of a trade, the SAR dots will go sideways for a period before they begin to accelerate. That acceleration will normally begin after the fourth or fifth dot. Swing trader's should consider a trend to be more "established" once the forth or fifth dot is reached and may want to use that dot as a signal.

Following such a strategy has a definite cost; traders will usually be forced to pay higher prices. The benefit, though, is that a more reliable entry point will be gained, one less prone to being whipsawed. By contrast, a more risky time to enter a trade based on a Parabolic SAR signal is when the dots have an increasing amount of space between them. As we saw in last week's article, when the dots are expanding the stock has typically already hit a series of new highs. The "acceleration factor" used in the SAR calculation is causing the dots to move parabolically.

A second way to optimize SAR is to pair this indicator with stochastics. When stochastics is close to the 80 level it signals that the stock is overbought; 20 is considered to be oversold. In a strong trend, a stock can remain overbought or oversold for an extended period of time. Therefore, combining SAR with stochastics can alert the swing trader to an entry that could be late in the game, where profits should be quickly nailed down if the trend begins to weaken. 

Above is an updated chart of Chevron Texaco (CVX, $58.16), which I used as an example in last week's article. Note that as of five periods ago, the dots below CVX were beginning to accelerate rapidly and stochastics was over 80. The combination of these signals warned the swing trader that the trend was becoming mature. CVX did move higher this week, but note also that the Parabolic stop and reverse point is at $55.91 -- about $2 below current levels. On Monday, February 7th, both stochastics and SAR were saying that it was getting into the late innings of the game. By Monday, February 14th, the contest will be entering extra innings, and the swing trader should be ready to nail down profits on signs of weakness.

ADX, MACD AND PARABOLIC SAR
One vital question all swing traders should ask is this: when has the trend truly reversed from down to up or vice-versa, justifying entry into a position? The swing trader's job is to find signs of trend reversal as early as possible and then ride that trend in the opposite direction for as long as possible. Enter a long trade too early and what seems like an apparent reversal can turn into a money-losing continuation of a former decline. Enter a short trade too late and success can be short-lived. The stock may go in your direction for a day or two, but then suddenly reverse course, taking your money along with it. 

The swing trader's goal is to identify that critical period when it is neither too early nor too late. Indicators are vital in spotting these optimum trading times. However, indicators also work in different time frames. Certain indicators like CCI and stochastics are designed for prompt action, giving almost immediate signals of trend reversal. However, on the down side, they are vulnerable to being whipsawed. 

Other indicators such as ADX or MACD give later signals, but are generally more reliable. Clearly, there is a trade-off here. If you follow the early indicator like CCI or stochastics, then it can suddenly reverse direction. If you apply a slower indicator such as ADX or MACD, then you have lost some of the profit potential of the trade. There is no ideal solution, and perhaps the proper choice of indicators comes down to the individual swing trader's risk tolerance. 

Last week we saw that Parabolic SAR was generally an early indicator, and that at the beginning of a trend it could be prone to what I labeled the "parabolic whoops." That happens when the stock changes direction for a few days, gives a Parabolic buy or sell signal, only to whipsaw the trader a couple of days later. I also pointed out that it was dangerous to rely on a Parabolic SAR signal when it was late in the game -- a point which is recognizable on the chart when the Parabolic SAR dots are spaced increasingly further apart. When that has happened, the acceleration factor in the SAR formula has kicked in, so the dots are beginning to ascend in a parabolic pattern. For that reason, I suggested combining SAR with stochastics so that late entry into trades was thereby avoided.

This week I want to discuss how combining SAR with two trend-following indicators, Wilder's ADX and Appel's MACD, provides greater reliability in entry. Wilder himself suggests combining SAR and ADX. Although the primary function of ADX is to represent the strength of the trend, it can also yield buy and sell signals when +DI crosses -DI and vice-versa. MACD is likewise a trend-following indicator and gives its trading signal when the "main" line and the signal line cross each other. I will use the 10-day moving average to represent the short-term trend and the 150-day moving average for the long term trend.

My rationale for entering into a long trade will be based on the following: Ideally the stock is above a rising 150-moving average. The stock must be above its 10-day moving average or penetrate above the 10-day moving average on the day the trade is initiated. A parabolic SAR buy signal must be confirmed by both ADX and MACD buy signals. To protect profits, the trade will be exited when the Parabolic SAR dots reverse direction. On a long trade, that will occur when the dots move from underneath the share price to above it. The reverse is true for a short trade.

USING PARABOLIC SAR TO SET STOP LOSSES
In this series of articles on Parabolic SAR, we have so far drawn several major conclusions. First, I pointed out that it can be dangerous to enter a position when the Parabolic SAR dots become spaced increasingly further apart. When that has happened, the acceleration factor in the SAR formula has kicked in, and the dots are ascending in a parabolic pattern. For that reason, I suggested combining SAR with stochastics to help avoid late entry into trades.

Second, I noted that combining SAR with two trend-following indicators -- Wilder's ADX and Appel's MACD -- provides greater reliability when entering a trade. ADX yields buy and sell signals when +DI crosses -DI and vice-versa. MACD gives its trading signals when the "main" line and the signal line cross each other. Combining these indicators with SAR provides a later entry to a trade, but gives greater assurance the true trend is being followed.

In addition to being a trend-following indicator, SAR was also designed to provide a point at which the swing trader would be stopped out of a trade. Several methods of setting stop losses are based largely on money management techniques and do not focus on the technical behavior of the stock. For example, one way of setting stop losses is to predetermine an arbitrary percentage of your purchase price you are willing to lose. One well-known figure -- suggested by William O'Neil, publisher of Investor's Business Daily -- states that you should never lose more than 8% of your position on any given trade. For example, in the case of a $10 stock, that would equate to a stop loss of $9.20.

Although predetermined amounts are useful for preventing unacceptably large losses in your account balance, they have nothing to do with the stock's behavior itself. The market doesn't know your purchase price for a stock and nor does it care. As such, stop losses set as arbitrary amounts leave traders vulnerable to being kicked out of a position for no reason other than normal market volatility.

Parabolic SAR provides a simple, but effective technical tool to aid in setting stop losses. A long as the dots are above the price of the stock in a short trade, you stay short. As long as they are below in a long trade, you remain long. Note that SAR is designed so that at the beginning of a trade there is sufficient time for the trend to mature. As the trend matures and the position becomes more profitable, SAR acts aggressively to protect these profits.

I use SAR as one factor in a setting a stop loss and adjust it according to other technical factors. Currently, we have a short position in Avocent in our portfolio. As of today, Parabolic SAR says to exit the trade only if the price rises above $36.20. 

My personal stop loss allows for a slightly larger move. It is at $37.11, just above the declining 50-day moving average that AVCT has twice failed to penetrate. In this case, the SAR stop and my stop are nearly $1 apart, but I find in many cases they are only separated by pennies. Next week I plan to conclude this series on SAR by discussing the limitations and drawbacks of the indicator.

LIMITATIONS OF PARABOLIC SAR
In our last several educational articles I have explored Parabolic SAR in depth. Daily SAR is a valuable tool in the swing trader's toolbox because it helps the trader follow the trend, immediately spots reversals, and gives a mathematical basis for setting the stop loss.

Is SAR a flawless indicator? No way! As helpful a tool as it is, SAR does has several limitations. Focusing on these limitations does not denigrate its value. Instead, it merely warns the swing trader to apply this technical tool with discretion. Mechanical judgments are to be avoided. Technical analysis is both an art and a science. In previous Inside The Black Box articles I have shown examples of where SAR worked effectively. In today's installment, however, I want to point out the fact that SAR does not hit a home run every time.

What limitations of SAR should the swing trader be aware of? There are three major ones I think bear further discussion:

-- During a period of sideways consolidation, SAR will give frequent buy and sell signals. Traders who mechanically follow these signals will at best breakeven and can possibly lose money by following SAR's guidance.

--SAR works best during a strong trend. Even then, however, SAR can give a signal to take profits, only to have the stock reverse soon after. Generally SAR leaves one or two dots going in the countertrend direction and then reverts to its original arc. I call these whipsaws the "parabolic whoops."

--SAR is a mechanical mathematical calculation. It is an effective first step in setting a stop loss, but should not be a final one. The swing trader should also integrate Intermediate and Minor trendlines, moving averages, "round number" support and resistance, and horizontal support/resistance areas.

Biotech stock Applera-Applied Biosystems (ABI, $20.80) provides an interesting example of the limitations of Parabolic SAR. In the last three months ABI has traded in a very narrow range, with a high of $21.27 and a low of $19.92. Note, how the 50-day moving average is largely trending sideways, reflecting the consolidation. 

During this time, however, SAR has given eight separate buy and sell signals! If followed, most of them would have resulted in a round trip to nowhere. The swing trader would have exited at breakeven or with a small loss each time. Further, valuable capital would have been tied up which could have been employed pursuing profitable opportunities.
SAR, like all indicators, is a mechanical tool based on a mathematical calculation. In the large majority of cases, it allows the trader to analyze a stock with greater insight than if it were not used. However, swing traders ultimately need to remember to be a master of their tools, not a servant to them. Like all other indicators, the directives of SAR should be applied with discretion.

Income Security of the Month Our "Income Security of the Month" for August 2008 invests in a fast-growing overseas market that doesn't get much exposure in the mainstream financial press. And although it typically makes enormous annual dividend payments -- it has paid an average dividend of 25.5% per year over the past five years -- this fund is perhaps most appealing for its total return potential. Specifically, the fund has delivered total returns of +178.9% since 2003, and it ranks in the top 10% of its category over the past decade.

Top 10 Stocks for 2008! Since we began publishing this report back in 2003, the picks we've featured have consistently beaten the broader market -- delivering average gains of +21.3% per year and outperforming the S&P by a nearly 2-to-1 margin. Act now to reserve your copy of our newest report -- Top Ten Stocks for 2008.