NYSE 200-day Market Breadth & Decline Probability

sjds

I was intrigued by the preponderance of on-line news stories regarding Tom Lee's "near-perfect" indicator for 4% or greater market declines in the S&P 500 over the coming month--a decline in the number of NYSE stocks above their 200-day moving average to 50% of the total NYSE stocks.

https://www.cnbc.com/2017/08/28/an-indicator-with-a-nearly-perfect-track-record-is-predicting-a-stock-market-pullback.html


It looks like this number is tabulated differently on stockcharts.com than on barcharts.com. But any thoughts on this indicator?

I'm looking for someone with a quantitative bent to backtest what market-breadth might predict about the S&P 500 (whether certain levels are associated with veritable probabilities of declines. Of note, CXO Advisory recently posted an article indicating these appears to be a near-futile enterprise, but the 200-day of all NYSE stocks was not part of their examination so far as I can tell.

markd

Well, whoever does it, it should probably be done with with carefully compiled and inspected historical data. Companies come and go. If you use historical data for existing stocks, they may not have existed during some portion of the back test. Also, there may be no data at all for companies that have ceased to exist or have merged. You may also want to research further whether this was done for cited study. I've seen some sloppy academic stuff on line. It's easy to manipulate data, but it takes some care to understand what the data actually represents. Also, when they say "track record", do they mean real-time track record, e.g. predictions made on a certain date for a certain time period looking forward, or do they mean an after the fact, found in the calculated data, track record. The first is better.