Saturday, December 20, 2008

Rising Summation Index

I was looking at the McClellan Summation Index both for the NYSE and Nasdaq this week and wondering what the implications are that the summation indexes are rising rapidly. This can be really confusing and it might lead many people to believe that the rally is stronger than it really is.

The summation index can be very useful in that devergences between it and the indexes usually precede reversals and reversals without confirmation from the summation index will usually fail. However, a rising summation index does not mean the market must continue to go up, nor does a falling summation index mean that the market must continue to go down.

For example, the summation index fell from late April 2007 to mid August 2007, but the markets continued higher until mid July. The falling summation index did precede the sharp sell off from July to August 2007 and the summation index reversed coincident with the markets. In this case this was a clue the market advance into July was on weak underpinnings and a sharp reversal was likely.

On the other hand, I looked back 10 years for market behavior (using the QQQQ) after the 12dema crossed the 26dema, i.e. the macd crossed above 0. In all but 5 cases in 10 years the markets continued higher for at least a few days after that event. In most cases, the market continued higher for weeks to months, which is a strong reason for not ignoring such signals. But what about the 5 cases when the market reversed 0 to 3 days after the cross? The summation index continued higher for 2 to 3 weeks in each case.

The conclusion is that if the summation index diverges from the market, we should be on guard for sudden reversals, but the market can reverse without a signal from the summation index. Traders should not be complacent just because the summation index is moving with the market. I am not certain of the reason for this behavior, but in the 5 cases above, and in the current case I suspect it is because the weakest laggards are being bought because they appear "cheap" which keeps the summation index rising while the leaders are being sold. It takes time for the "buy cheap stocks" mentality to dissapate after a bottom even as the markets are rolling over.

To follow up on Friday's post, if the markets do break out Monday, the rally should be anywhere from 3 to 11 days based on the length of wave a, which puts the end of the rally at December 24 to January 7.

1 comment:

dave said...

What are the implications of the upper & lower BB curving towards one another like a lobster claw ?