Measuring at what degree one variable moves in relation to another is a very difficult and imprecise exercise but it could be useful in predicting the risk factor in markets, in general. Establishing a “logical link” between such variables is the business of the CNBC talking heads and is by definition loose talk. In reality the more distant and apparently “unlinked” they are the more interesting the correlation, if it holds FOR SOMETIME.
Chart 1: In bear markets technologies sometimes display a “bullish flat”. Such a case was recently reflected in the moves of the Qs. As previously posted , I noticed (since 2000) that at the end of such periods, biotechnologies have a run, during which BTK (yellow) outperforms the Qs (green), as shown below. When this “mini BTK rally” ends, its drop under the relative value of the Qs (see cross-over line) would amplify for a while the bearish sentiment for the whole teks market. Then, the beginning of a countertrend is marked by QQQQ moving UP faster than BTK. This event can be used as an early BUY QQQQ signal. Then both move up together confirming the solidity of the trend.
Chart 2: Can a simple correlation between two variables, be reflected in the moves of a third one? GLD = green; VIX = purple; SPX = white. In this case, the divergence of VIX and GLD was used for the generation of four valid SELL signals for SPX and one short lived and questionable (03/19/03). This chart also shows how difficult it was to “swing trade” around March OE, when SPX was subjected to a whipsaw. I am curious about what might happen now. I am willing to assume that if GLD and VIX will cross as indicated by the arrows, a BUY SPX signal will confirm the new uptrend.
Market Data 
Markets 
AI
