Why This is Not *the* Bottom

Has the bear bottomed? That was the question I heard with surprising frequency after Thursday's rally. My answer has been, "No. It takes a lot more weakness to make for a cyclical bear market bottom."

Let's look at the historical record:

1966 cyclical bottom: 56% of all stocks make 52-week lows on August 29th.

1970 cyclical bottom: 58% of all stocks make 52-week lows on May 26th.

1974 cyclical bottom: 37% of all stocks make 52-week lows on September 13th.

1978 cyclical bottom: 30% of all stocks make 52-week lows on October 30th.

1980 cyclical bottom: 38% of all stocks make 52-week lows on March 27th.

1981 cyclical bottom: 31% of all stocks make 52-week lows on September 28th.

1987 cyclical bottom: 57% of all stocks make 52-week lows on October 20th.

1990 cyclical bottom: 35% of all stocks make 52-week lows on August 23rd.

1994 cyclical bottom: 23% of all stocks make 52-week lows on April 4th.

1998 cyclical bottom: 33% of all stocks make 52-week lows on August 31st.

2002 cyclical bottom: 27% of all stocks make 52-week lows on July 24th.

2006 cyclical bottom??: 8% of all stocks make 52-week lows on June 13th.

I don't think so.

Notice the average spacing of the cyclical bottoms. Those bottoms don't occur until bearishness is rampant, and that typically takes a large proportion of issues making fresh lows.

Notice something else very important: The point at which we hit maximum new lows is rarely the eventual price low for the cyclical decline. For instance, we had 27% of stocks making new lows in July, 2002, but we saw further price lows in October of that year and March of 2003. Similarly, we saw 31% of stocks making new lows in September, 1981, but didn't see actual price lows until August of 1982.

So what we're seeing in the current market is either a normal correction in a bull market (like we had in October, 2005 and May, 2004), or it is just the *start* of a full-fledged cyclical bear market that has much further to go. The quality of the next intermediate-term rally should go a long way toward identifying which scenario we're in.