"If they insist on trying to time their participation in equities, they should try to be fearful when others are greedy and greedy only when others are fearful." -- Warren Buffet, Chairman's Letter, 2004
CNN Money has an interesting take on attempting to quantify the current fear and greed in the market. It combines highs/lows, bond spreads, put/call ratio, bond vs. stock returns, S&P 500 movement, volume summation, and volatility into a single number from 0 (extreme fear) to 100 (extreme greed). The number is not very interesting itself, reflecting what you probably already know - for example, the market is way up right now.
What is interesting is that periods of extreme greed (index >80) have reliably been followed by market drops for the last three years (the viewable lifetime of the index). In April 2010, the index hovered over 75 for a month, just reaching 80 and in July 2010 SPY was down over 15% from the April high. In October 2010, the index briefly went over 80 followed by a 3% drop in SPY during November. In February-March 2012, the index was over 80 for some time before SPY dropped about 8% in May. In September 2012, the index went over 80 for two weeks, stayed near 80 for two more weeks, and by early November had dropped 7.5% from the September high.
The method predictor is not perfect - the index stayed below 80 for all of 2011 and thus missed the August drop. At the very end of 2011 the index flirted with 80, but the top didn't occur until February and the drop was only 4%. If less than 100% accurate is not acceptable to you, I suggest you get out of the market and ask Congress for higher social security payments.
If you feel that one false positive (counting 2011 as a false signal due to being very early) out of six signals is a good record, then the signal point and the timing still need to be worked out. In early 2010, the index stayed above 75 for some time before only once hitting 80. Additionally, the index tends to hit high values in the middle of a rally - not a good exit time. What I have observed is that the index first hits 80 about one to six weeks before the top, and four to 10 weeks before the market bottom. Thus, roughly speaking, three weeks of "extreme greed" is a time to start being fearful. Six weeks of "extreme greed" is a time to definitely be fearful.
How I would play this: The majority of my investments are long-term oriented. I don't try to jump in and out of most positions. Instead, once the overly greedy signal starts going off, I cut back on purchases. Then sometime after the three-week signal (depending on whether the rally seems strong or weak to me based on other observations), I put on a small bearish option spread. Personally, I keep this low risk; my goal is just to cushion the dip. This worked out well in October 2012. I recently put this on again. The index first hit 80, exactly, on Jan. 3. It has consistently stayed above that level for six weeks now.
My biggest concern is what a long bull market will do to this indicator. Since limited data is available on the CNN Money site, it's not possible to check this idea during a tech bubble or a housing bubble or other sustained bullish run. My secondary concern is a long sideways movement that makes the signal a wash. For a current contrary indicator, Philip Mause has an interesting article on the dividend yield of the S&P 500 - essentially, it says that yields are not exceptionally low by post-2009 standards. So if prices are (somewhat) range-bound by yield, then there is room for still more price gains before a top.