On this blog we commonly refer to the overbought and oversold characteristics of stocks, indices, commodities - anything with a price. There has been some question about what methodology we use to determine the thresholds of overbought and oversold. Birinyi Associates uses a system similar toBollinger bands, but not identical. A stock is classified as being "overbought" when it is more than one standard deviation above its 50-day moving average, likewise a stock is oversold if it is more than one standard deviation below the 50-day average. Theoretical highs and lows that are sometimes mentioned are derived using the extremes over the last 18 months and a proprietary formula to determine price.
The chart below shows the S&P 500 based on the sum of its members. We analyzed all 500 stocks' trading ranges over the last year and calculated if they were overbought or oversold on a given day. Stocks that were oversold received a score of -1, while those that were overbought received a score of +1. By adding up the scores of every member for each day we arrive at a Net OB/OS score for the S&P 500; where a reading of -500 would mean that every member is oversold and conversely if they were all overbought the indicator would read +500.
As shown, in terms of broad market extremes, the indicator bottomed at -496 on October 9th, 2008. The recent peak occurred on May 4th at +372.
As more stocks become overbought or oversold the number of stocks likely to see negative or positive reversals increases. We can also see from this chart how the market can experience a correction in time and not necessarily in price. Using the S&P 500 as an example, the 50-day average is now trending up and if the index trades sideways it will return to its normal range. The same holds true for the stocks, the lack of big gains since May 4th has seen roughly half the index members return to their normal ranges.