• tracks option premiums on the S&P 100 (OEX) which trades on the Chicago Board Options Exchange
• available since 1986 for the OEX, also available for the Nasdaq (symbol: VXN) since Jan. 2001
• information is available daily from your data provider, or from www.cboe.com in the Market Statistics Summary section
• it is also possible to retrieve back data from this website
• if the VIX, or implied volatility, is 20%, it means that one year from now, the index has a 68% chance (1 standard deviation) of being 20% higher or lower than it is now
• we don't necessarily know in what direction the index will go, but volatility in either direction presents opportunities for making money
• implied volatility is almost always higher than actual historical volatility, which suggests that options on the OEX are almost always overpriced
• VIX is generally believed to be a contrarian indicator (I invert the y-axis scale in MetaStock, so that low readings show as tops, and high readings show as bottoms)
• when it is trending higher, investors are bidding up index option premiums (either calls or puts), presumably because they believe the market is about to make a significant move
• often, however, the market does just the opposite and settles into a narrow trading range
• conversely, when options premiums are trending lower, investors sell more index options in the belief that the market will remain quiet
• however, that is often the time when the market will move dramatically higher or lower
• the problem is that VIX only tells us when an explosive move is likely to occur, it doesn't tell us in which direction
• the intraday high for Sept. 17/2001 was 48, for October 1997 crash it was 55, for January 1991 (Gulf War) it was 38, and for October 1987 crash it was 150!
• the yearly average is around 25
• the biggest moves seem to occur when VIX moves under 25, and then closes below (on an inverted scale) its 200 day SMA
• market bottoms tend to occur when VIX hits values above 40-50

• a short term sentiment indicator developed by Marty Zweig in 1971
• data is available for CBOE listed options, and for the S&P 100 (OEX)
• daily numbers are available at the CBOE website
• institutional activity and arbitrage can sometimes cloud the picture
• many technicians use a 10 day SMA of the put/call ratio
• invert the y-axis scale, so that overbought levels (.35) show as tops, and oversold levels (.55) show as bottoms
• can use 21 day Bollinger Bands, and buy when the P/C ratio crosses below the lower Bollinger Band, and then turns up (on an inverted scale)
• the analogy is that like a rubber band, when sentiment is stretched to an extreme and then reverses, it will tend to reverse quickly and sharply
• since 1980, there have been 78 buy signals, with 67% of them turning out to be profitable
• unfortunately, the sell signals haven't been as good (correct 73% of the time, but not very profitable)
• some technicians feel that it is more bearish if low P/C numbers are found in a sideways/down market than in an up market
• psychology should be more bullish in a rising market, but if it remains as the market goes sideways or down, then it may take on more significance

• weekly poll of approx. 130 newsletter writers, and published weekly in Barron's
• primarily used as a contrarian indicator
• investment advisors usually aren't right about the direction of the market, especially at key turning points
• extreme levels are the most important ones to watch for (therefore, more of a long term sentiment indicator)
• John Bollinger uses net bulls (%Bulls - %Bears), instead of the individual %Bulls and %Bears readings
• it usually runs from +25 to -25 except during extreme times
• highest level was 41% in Jan. 1992, while the lowest level was -34% in Dec. 1988
• Martin Pring uses a 10 week SMA of %Bulls / (%Bulls + %Bears)
• Highest level was 76% in Jan. 1992, while the lowest level was 28% in Dec. 1988

• daily poll of approx. 170,000 members daily
• also, published weekly in Barron's
• used as a contrarian sentiment indicator (assumes the individual investor isn't right about the future direction of the market at key turning points)
• measures the percentage of individual investors who are bullish, bearish, or neutral on the stock market over the next six months
• the survey was started in July 1987, and is conducted weekly, and updated each Thursday
• for AAII, a bullish percent greater than 45% is considered bearish, whereas it is considered bullish if it falls below 20% (use an 8 week SMA to smooth the data)
• going back to July 1987, the maximum percent bullish was 66% (August 1987), and the minimum was 12% (Nov. 1990)
• the maximum %Bullish / (%Bullish + %Bearish) is 92% (August 1987), and the minimum was 16% (Oct. 1990)
• the maximum percent bearish was 67% (Oct. 1990), and the minimum was 6% (Aug. 1987)

• at some point, short sellers will become buyers of stock to cover their positions
• the trading levels of these players will then either support a declining market, or accelerate a rising market
• short interest ratio is calculated by dividing the total number of short sales outstanding on the NYSE (or an individual stock) by the average daily trading volume for the past month
• SIR for the NYSE is published monthly in Barron's
• a ratio above 1.0 is bullish and a ratio above 1.5 is a buy signal
• a value greater than 5.0 is bullish because it reflects excessive pessimism, and can represent potential demand if the market rises
• bullish readings tend to lead the market by several weeks
• this ratio worked beautifully, signalling nearly all the major bottoms until the 1970's
• then, stock options and stock index futures were introduced
• now, short selling may just represent hedge positions against these derivative products
• therefore, a high SIR no longer reflects pessimism (or potential demand) to the extent that it once did

NYSE Specialist Short Sales Ratio

• based on the principle that the smart money knows what is about to happen in the market
• therefore, heavy specialist short selling is bearish, and light specialist short selling is bullish
• calculated by dividing the weekly total of shares sold short by NYSE specialists by the total of shares sold short by all NYSE participants
• there is a two week lag in the publication of this data in Barron's
• when a one week reading falls below 33%, or the average of the past four weeks is below 35%, a buy signal is generated
• a sell signal is generated when a one week reading exceeds 58%, or when the average of the last four weeks exceeds 55%
• the upper 40's is now the warning range because there haven't been readings over 50 in a long time
• occasionally, an extremely low reading appears (i.e. below 30)

• when these sophisticated investors are borrowing money heavily to buy into the market, the future market trend will likely be up
• conversely, when these individuals are liquidating stocks and reducing their margin debt, it is a good indicator that the market will soon be moving lower
• on the last trading day of each month, the NYSE calculates the total amount of debt owed to member firms by their customers
• the information is not released for two to three weeks after month end in Barron's
• can use a 12 month SMA of the ROC to generate buy and sell signals