Thursday, June 19, 2008

Selloff In Financials Incredibly Persistent

One statistic I look at and use in my trading is the amount of time a security trades above or below certain moving averages. One moving average I look at is the 10-day moving average. Right now there are 4 stocks in the S&P 100 that have closed below their 10-day moving average for at least 30 days in a row. They are BAC, RF, GM and WB. Using the current S&P 100, I ran some scans back over the last 10 years to find out how unusual this is.

In the last 10 years there have been only 65 occurrences of an S&P 100 stock closing below its 10-day moving average for 30 days in a row. Twice there have been 4 or more on at once. The first time was in late August of 1998 when there was up to 5 and the 2nd time was early March of 2000 with 4.

From a historical perspective, some of the recent selling, especially in banks and financials, has been persistent to a degree that has rarely been seen.


r said...


In the time frame you referenced there was only one persistent bear market (2000-2003). How many occurrences would you find if you went back to the 1970s?


Anonymous said...

What would be interesting is to find any time frame that nasdaq and dow have such a big difference... and what happen afterward... I mean the difference bewteen nasdaq and dow is just huge!

Anonymous said...


last huge difference between the Dow and Nasdaq i can remembered was the late stage of the internet bubble leading up to early March 2000. remember Warren Buffit being on the hot seat for not owning tech stocks? yeah, just curious why the difference. could be that techs not as exposed to credit meltdown but not all Dow techs are exposed too. very strange.


Anonymous said...

i meant...

could be that techs not as exposed to credit meltdown but not all Dow STOCKS are exposed too.


Rob Hanna said...


Good question but difficult to answer. The issue is I would need to rebuild the S&P 100 over the years and use data from extinct securities. It's possible to do, but more work and expense than I'm willing to take on to answer the question. My shortcut was to simply look back 10 years using the current S&P 100. Obviously the further back I look the less accurate the results are since additional index changes would be excluded.

Anon & Jimmy,

Good idea. I did play around with this last night. I looked at it a number of ways and found no compelling evidence for either the bullish or bearish case. The degree of the divergence between the Dow and Nasdaq IS unusual, but it doesn't appear to be consistently predictive of direction over the times frames I looked at.