Last week, I addressed the divergence that was evident between the share price of QQQ and its equal-weighted cousin, QQEW. Both of those ETFs are designed to track the stocks in the Nasdaq 100 Index (NDX). This same divergence is evident in this week’s chart, which shows the daily A-D Line for the component stocks which make up the Nasdaq 100 Index.

An A-D Line is a cumulative running total of all daily Advances and Declines. It changes each day by the value for the daily breadth (A-D). Because it is describing the behaviors of a population of stocks, it is in a category of indicators known as a “diffusion index.”

Years ago, I started a project to compile data on all of the component stocks in the NDX, just because I thought it might be interesting to track the A-D data and other data. It was a lot of work fetching price data on those 100 stocks, plus going back as far as I could to track changes in the components of that index and assemble the price data on the departed issues. It remains an ongoing task to keep up with splits and index component changes.

The A-D data for the NDX stocks turned out to be surprisingly not as useful as I had hoped. Most of the time, it does whatever the NDX itself is doing. The whole point of looking at an A-D Line is to get different answers from what prices are saying, and hopefully some useful divergences at important turning points. So if a particular A-D Line just does whatever prices are doing, there is not much value in studying it. One can still get useful information from the A-D data in other ways, such as looking at the acceleration taking place. Because of that, the McClellan Oscillator based on these A-D data is quite useful, and I feature it regularly in my Daily Edition.

Something different is happening now, and we are seeing a rare divergence between the NDX A-D Line and the NDX itself. This almost never happens, which makes it all the more noteworthy, and the message is that there is an unusual disagreement right now between the big cap stocks like Apple (AAPL) and Microsoft (MSFT), which drive the price level of the NDX, versus the behavior of the other index components.

An A-D Line is useful because it helps us detect the health of the liquidity stream, which affects all stocks. But good or bad liquidity affects the small and least-deserving ones first. So when illiquidity comes around, it picks off the weak first, while the strong can still muscle their way in to get a drink. In the A-D data, every stock gets an equal vote, the weak ones the same as the strong ones. That is why the messages one can get from A-D data can sometimes be so useful.

This point cannot be made, however, for the message from the A-D Line for the overall Nasdaq market. It has such a tremendously bearish bias that it makes any messages one might take from it unreliable. I like to stump fellow analysts sometimes by asking, “When was the last time that the Nasdaq’s A-D Line made a new all-time high?” This is a trick question, because the Nasdaq’s A-D Line started downward from the beginning of the data in 1972 and it has never made it back to that level. So it has not ever made a new all-time high.

This negative bias stems from the fact that the Nasdaq has looser listing standards than the NYSE (the A-D Line for which is very useful). If a stock is going to come public and then go broke, it is more likely to do that on the Nasdaq, and every down day from IPO to delisting contributes to the Declines column.

It can still be useful sometimes to track the acceleration taking place in the Nasdaq’s A-D data, and so a McClellan Oscillator for those data can be useful. But the raw A-D Line is so unreliable in the messages it gives as to be functionally unusable. With that said, I will note that the overall Nasdaq’s A-D Line this week made another new all-time low. So it is definitely NOT saying that there is hidden strength in the market.