In the first installment of this series, we looked at using RS to find the hottest stocks and sectors. In this second installment, we will look at using ADX to find potential candidates and sectors to short.

Shorting, The Rules Are Reversed

If you buy XYZ 100 and later sell it for 102, you make a profit, less commissions, of 2 points. In shorting, the rules are reversed, you sell the stock first (known as selling short), then, hopefully, buy it back (known as "covering") cheaper in the future. So if you sold XYZ short at 102 and it later dropped to 100 (vs. rising in our first example) you would make a profit of 2 points (102-100 = 2). For a refresher on shorting, see Shorting Stocks: The Art of Playing Both Sides of The Market.

RS vs. ADX

As a momentum player on the short side, you should look for stocks that are in established downtrends--the complete opposite of a momentum player on the long side. Last time, we used relative strength (RS) to find candidates on the long (buy) side. However, because relative strength (RS) is a measure of strength not weakness, stocks with low RS aren't necessarily in downtrends. Often they are simply being outperformed by other stronger stocks. So a stock that has been going sideways for quite some time may have a low RS reading. This is especially true in strong market conditions. Therefore, in order to find stocks in downtrends, it's usually better to use an indicator such as Average Directional Index (ADX) which measures trend.

ADX Refresher

Before we get into the actual scan, a quick rehash of ADX is in order. ADX is based on directional movement (DM) calculations. If more of today's range is above yesterday's range then that range (above yesterday's high) is measured as positive directional movement (+DM). Conversely, if more of today's range is below yesterday's range, then this range (below yesterday's low) becomes negative directional movement (-DM). This is illustrated in Figure 1 below.

Figure 1

The ADX is calculated by taking the smoothed difference (also known as DMI) of the +DM and the -DM. There are special situations on how the DM is calculated but understanding the above should be enough for the purposes of this article. For those interested in the details, refer to the article: Capturing Trends With The ADX.

ADX measures trend, not direction. The higher the ADX, the higher the trend. To determine direction, the positive directional index (+DMI) vs. the negative directional index (-DMI) is analyzed. If the +DMI is greater than the -DMI then the stock is considered to be in a uptrend. Conversely, if the -DMI is greater than the +DMI, then the stock is considered to be in a downtrend. Because we are focusing on potential shorting candidates for this article, let's look at an example of a stock in a downtrend as measured by ADX and -DMI > +DMI.

Figure 2

How High is "High"?

We stated that stocks with a high ADX are said to be in trends but how high is high? Most textbooks claim that a stock is said to be trending if the ADX is greater than 30. However, it's really a function of the market. If the overall market is choppy, then there will be fewer stocks with ADX readings over 30. Therefore, in order to find stocks that are trending (relative to the market), you would have to reduce your minimum ADX readings to, say, 25. On the other hand, if the overall market is in a strong trend, then you may have to increase the ADX settings in order to help reduce the number of stocks.

The Scan

In order to find candidates that were trending lower, I used an ADX of 35 or higher and required the -DMI to be greater than the +DMI (which equates to "down" in the scanner). As previously stated, there are no "exacts" when deciding how high of ADX should be used. Thirty five was used because lower minimums were producing too many stocks as the broader market was in a downtrend at the time the scan was run. Like last time, we used a minimum average volume of 500,000 (which equates to 5,000 in the StockScanner). This was used in order to give us a representative sample of stocks in downtrends (vs. measuring thinner volume issues, see Scanning For Profits Part 1 for more details on using volume in the scanner). Also, when shorting, you want to focus on somewhat higher volume stocks in order to help avoid being caught in a "short squeeze" which tends to happen more offer in thinner issues (refer to the article on shorting for details). For this scan, we used a minimum price of 45. This was used for several reasons, 1) it helps narrow the focus 2) like the volume requirements, provides more meaningful representation of stocks vs. lower-priced stocks 3) when shorting stocks, your profits are limited to the maximum amount the stock can drop, therefore, higher priced stocks tend to offer more opportunity, especially for the short-term trader (again, see Scanning For Profits Part 1 for a discussion on price).

Figure 3 below shows what the above scan would look like:

Figure 3

Results

Below is the results of the scan along with the closing price three days after the scan was run.

Like the scan in the first article, we see that momentum begets momentum but this time on the downside as 20 out of the 21 stocks produced closed lower three days later. Overall, the stocks lost over 7% of their value during this period. Be warned though, this doesn't guarantee that stocks produced by this scan will be lower in the futures(or stocks produced by any other scans for that matter will follow through).

Picking The Next Winner

Going back to 03/03/2000, we see that many of the stocks (seven to be exact) produced by this scan are financial related (i.e., banks, insurance). Like last time, it helps if you take the stocks produced by the scanner and look for some sort of confirming technical pattern. Putting the pieces together, we see that American Express (AXP | Quote | Chart | News | PowerRating), a financial conglomerate, is a pullback from lows (a) on 03/03/2000. Notice that the stock took out its prior lows on the following day, providing an entry for shorting and then proceeded to drop for several days (b).

Figure 4

Again, Experiment!

Above I have shown just one scan that could be used to find candidates on the short side. Like I said in the first installment of this series, experiment! Make the StockScanner your own. Try using higher or lower ADX. Increase or decrease your volume requirements. Run scans using looser parameters to get a broad view, then increase price/volume/ADX to whittle the list down. Again, it's your StockScanner, use it as you see fit.

Click Here For A Free 1-Week Trial To Dave Landry's Daily Swing Trading Alerts

Click Here To Learn How To Trade Dave Landry's Best Strategies