AUGUST AND SEPTEMBER ARE USUALLY WEAK MONTHS -- THAT WOULD BE A LOGICAL TIME FOR MARKET CORRECTION TO RESUME -- CLARIFYING MOVING AVERAGE LINES AND SIGNALS-- MARKET RALLY FADES AS 50-DAY AVERAGES BECOME NEW RESISTANCE

AUGUST THROUGH OCTOBER IS MOST DANGEROUS TIME... In my weekend analysis, I closed with the comment that I thought the market could suffer a 10% correction "between now and the autumn". I chose that time period for a very good reason. By the end of July, any summer rally has usually run its course. That's followed by three of the worst months of the year -- August through October. Most people are aware that the period from September to October is generally the weakest of the year. What they may not know is that August is dangerous as well. According to the 2007 Stock Trader's Almanac, August has become the weakest S&P 500 month in the last nineteen years. There's some good news further out, however. That weak period is usually followed by bottoms being formed during October. Along those lines, I've gotten requests to look at some previous summer downturns starting with 1998. I've included downside corrections in 1997 and 1990 as well. What makes 1998 somewhat similar to today is that it took place while a major hedge fund (Long Term Capital Management) was collapsing. Some have drawn comparisons to the current subprime problems. Chart 1 shows that the S&P peaked in mid-July 1998 and fell 22% before turning back up in October. Chart 3 shows a similar pattern in 1990 with the S&P falling 10% from mid-July to October before turning back up again. The situation in 1997 was a bit different, but still shows the same seasonal tendency (Chart 2). The S&P peaked in mid-August 1997 and lost 7% during that month; it reached a marginal new high in early October before falling 13%. In total, the S&P lost 11% from its August peak to its October low. In all three instances, the market resumed its long-term uptrend after an October bottom.

Chart 1

Chart 2

Chart 3

WHAT ABOUT OTHER YEARS... Two other market peaks -- in 1994 and 2000 -- are very different from the current situation. 1994 was a stealth bear market where the stock market traded sideways for a year before turning back up again. Even so, the S&P fell 6% from the end of August (Labor Day) to October before rebounding. The final bottom took place during December 1994, which was 7% from its late August peak. The year 2000 was the only year where the market didn't turn back up in October. That's because it was the start of a major bear market. During 2000, the S&P peaked in March. It then rallied in late August before rolling over again. From its late August peak to its October bottom, the S&P lost 15%. Unfortunately, it kept dropping until October 2002. I'm showing 1987 with some hesitation. Please don't get the impression that I'm expecting anything like a repetition of that year. My focus here is in on the "seasonal" pattern, not the size of the downturn. After a strong rally into the summer of 1987, the S&P 500 peaked during August and dropped 8%. A modest rebound into early October was followed by another 31% plunge. From mid-August to October 1987, the S&P lost 33%. After treading water for awhile, the market turn back up again the following year. Five of the last six major downturns since 1987 resulted in the market turning back up again during the fourth quarter (usually in October). That may be the case again next year. My main message here is that the period from August to October is the most likely to witness a market correction. If an October bottom occurs again this year, I suspect it will be at least 10% below the July peak.

Chart 4

ONLY DAILY 13-34 EMAS HAVE TURNED DOWN ... I mentioned on Friday evening that the 13-34 EMA system had given the first short-term sell signal in four months. I've been asked to clarily which lines I'm taking about. I was referring to the "daily" 13-34 exponential moving averages which turned down last week (the 13 day line fell below the 34 day). The weekly EMAs are still in an uptrend. As I pointed out on July 24, however, the spread between those two weekly lines (top of Chart 5) has started to weaken from its last major peak formed in early 2004. I take that as a potential warning that the major trend is weakening. The only signal given so far, however, has been "short-term" in nature.

Chart 5

WHY AN 80 WEEK MOVING AVERAGE?... I've also been asked why I plotted an 80-week moving average line on the weekly S&P 500 chart on Saturday. Here's why. Chart 6 is a monthly S&P chart with Bollinger bands applied. [The bands are the two outer lines]. The middle (dashed) line is the 20-month moving average. [20 is the value used on daily, weekly, or monthly Bollinger bands]. You'll see that the middle line acts as a great support level during bull markets. That was the case from 1997 to 2000 and again from 2003 to the present. [The breaking of the 20-month line in 2000 signalled a bear market. The move above it on 2003 signalled a new bull market]. Every downside correction since 2003 has bounced off the 20-month average. I simply took that number (20 months) and converted it to 80 weeks to use on a weekly chart. The 20-month (or 80 week) m.a. now sits about 12% below the July S&P high. I was trying to show that the market could still drop 10%-12% from its July peak without interrupting its long-term moving average uptrend, and also to suggest a potential downside target if and when a correction materializes.

Chart 6

MARKET RALLY FADES BELOW RESISTANCE LEVELS... The modest market bounce that started on Monday faded badly today. The hourly bars in Chart 7 show how modest the rally was for the S&P 500. The flat lines show Fibonacci retracement levels measured from the July peak to Monday's low and represent potential resistance barriers. So far, the S&P 500 bounce has fallen well short of the 38% retracement line (lowest line). The chart also shows the inability of the S&P to get back above its late June low at 1484. That's a classic case of previous support becoming new resistance. The daily bars in Chart 8 show the S&P bearing down once again on its 200-day moving average. Did anyone (besides Wall Street and TV commentators) really think that last week's carnage could be repaired so easily? At the top of this story, I wrote that August had become the weakest month of the year for the S&P 500. Unfortunately, that dangerous month starts tomorrow.

Chart 7

Chart 8

50 DAY AVERAGES BECOME RESISTANCE... The 50-day moving average is another example of a previous support line (below the market) becoming new resistance (over the market) after it's broken to the downside. Charts 9 and 10 show the Dow and the Nasdaq Composite meeting new selling just below their 50-day (blue) lines today. The Dow is threatening to break its June lows. The Nasdaq Composite has just undercut its late June low at 2560 and appears headed toward its 200-day average. The Nasdaq market took another turn for the worse today. The Nasdaq 100 Trust (QQQQ) closed below its 50-day line today (Chart 11). And it did so on rising volume. Rather on than "buying on dips" (as Wall Street strategists on TV have been suggesting), astute investors who can tell up from down, have been "selling on strength". Given the damage done to the market last week, that seems the wiser strategy at this point.

Chart 9

Chart 10

Chart 11