One of the things you'll note immediately when viewing the chart is that the P/C data tends to reach tops and bottoms before the volatility data. Secondly, the volatility data often tends to overshoot the P/C data at peaks and troughs. One of the reasons for this is that trend changes aren't generally seen for what they are until well into the process. For instance, after a prolonged period of selling, people approach the market more cautiously, and that bearishness takes longer to work itself off in Put pricing. Of course, the reverse is also true, and it's one of the reasons why many contrarians believe that the VIX and VXN are a good measure of complacency in the markets.
The bottom line is that the P/C Ratio tends to be more of a "real-time" indicator, but doesn't provide enough of a benchmark for persistence of trend. At the same time, the VIX and VXN are too slow for timing purposes. When studying this chart, therefore, we want to look at divergences between the data sets, which are more apt to occur during periods of complacency, and at convergences, and in particular in-synch movement, to confirm the persistence of trend, and as an aid in marking sentiment extremes. These extremes, of course, are always followed by a reversal of the trend.
www.market-harmonics.com/free-charts/sentiment/pcvi.htm
