{"id":10041,"date":"2012-02-19T22:18:57","date_gmt":"2012-02-20T03:18:57","guid":{"rendered":"http:\/\/drduru.com\/onetwentytwo\/?p=10041"},"modified":"2012-07-09T00:47:33","modified_gmt":"2012-07-09T04:47:33","slug":"trading-after-overbought-period-ends","status":"publish","type":"post","link":"https:\/\/drduru.com\/onetwentytwo\/2012\/02\/19\/trading-after-overbought-period-ends\/","title":{"rendered":"Trading the S&P 500 After An Overbought Period Ends – Using the Percentage of Stocks Trading Above Their 40DMAs (\u201cT2108\u2033)"},"content":{"rendered":"

Over the past year or so, I have worked to further refine the application of T2108<\/a>, the percentage of stocks trading above their 40-day moving averages (DMAs), to short-term trading. Last November, I updated my analysis of overbought periods<\/a>, periods where T2108 equals 70% or more. That analysis provides a complete framework for understanding the odds of duration and returns during overbought periods. In this post, I am filling in a large gap by analyzing what happens BETWEEN overbought periods. Understanding these dynamics helps a trader determine the appropriate amount of risk to take on during overbought periods.<\/p>\n

Here are the fundamental conclusions I generated from analyzing the S&P 500’s behavior between overbought periods since 1986. For convenience, I will call this the “interim period.”<\/p>\n

    \n
  1. Extended rallies are typically characterized by a series of overbought periods.<\/li>\n
  2. There is no relationship between the duration of the overbought period and the maximum decline before the next overbought period. However, the largest declines have all occurred after overbought periods that lasted 20 days or less. After longer overbought periods, the maximum decline during the interim period has been 10%.<\/li>\n
  3. There is also little relationship between the length of the interim period and the maximum decline during that time. However, declines of 10% or greater have ALL occurred when the interim period lasted between 50 and 155 days.<\/li>\n
  4. After excluding interim periods that included oversold periods, there is a stronger relationship between the duration of the interim period and the maximum decline. The vast majority of interim periods have lasted less than 50 days.<\/li>\n<\/ol>\n

    Given these observations, traders should treat rules for trading between overbought periods as general guidelines. The risks are worth the rewards IF traders size positions modestly, prepare to make multiple trades, take at least some profits after moderate declines, AND stop out by the time T2108 has begun a new overbought period.<\/p>\n

    Specifically, traders should follow these guidelines<\/strong>:<\/p>\n

      \n
    1. Short early but in moderate size once an overbought period begins.<\/li>\n
    2. Never build a “full” trading position just in case the overbought period lasts longer than 20 days, and\/or reduce the size of the bearish position after 20 days. Hedging against the bearish position counts as a reduction in risk.<\/li>\n
    3. Start taking profits no later than a decline of 5% or so on the S&P 500 during the interim period. Stop out by the time the next overbought period begins.<\/li>\n<\/ol>\n

      Here are the specific charts supporting my conclusions along with some more detailed observations. (Click on the smaller charts to see the larger view).<\/p>\n

      First, here is an overlay of the S&P 500 with the duration of all overbought periods since 1986. Since the chart is difficult to read, I also zoom in on the period since 2003. (Graphing difficulties in Excel prevented me from properly labeling price levels for the S&P 500 given it is on a log scale). Visually note the lack of any clear relationships between the S&P 500 and the duration of the overbought periods EXCEPT that extended rallies tend to include a multitude of overbought periods. So, if a trader finds himself constantly fading the market, it is probably time to take a step back and accept that a major rally may be underway and to plan accordingly!<\/p>\n


      \n
      \"S&P<\/a>
      S&P 500 Versus Duration of Overbought Periods Since Sept, 1986<\/figcaption><\/figure>
      \n
      \n
      \"S&P<\/a>
      S&P 500 Versus Duration of Overbought Periods Since Jan, 2003<\/figcaption><\/figure>
      \n<\/center><\/p>\n

      The chart below is a scatter plot showing the maximum decline in the S&P 500 in the interim period indexed by the duration of the preceding overbought period. In other words, this is an attempt to answer the following question: “what is the maximum correction a trader might expect after an overbought period ends?” Returns are measured from the close of the day the overbought period ends.<\/p>\n


      \n
      \"Returns<\/a>
      Returns between overbought periods can vary tremendously<\/figcaption><\/figure>
      \n<\/center><\/p>\n

      First, notice that there is no correlation between the duration of the overbought period and the maximum decline in the interim period. Also notice that in SOME cases, the maximum “decline” is actually a gain. I confirmed these cases manually and found that when the duration between overbought periods lasts only a few days, the S&P 500 sometimes drifts, sometimes lower, sometimes higher, while consolidating before the next move. During uptrends or sideways action, the interim period can easily generate marginally positive returns.<\/p>\n

      Although the above chart does not generate any overall correlation, I can characterize three distinct areas of the chart (I have changed these definitions from the last time I referenced this chart):<\/p>\n