The 1929-87 analog is the overlay of the 1929 crash over the 1987 crash. The two crashes are eerily similar - this means that knowing the shape 1929 took would have helped one avoid the 1987 crash, indeed profit from it (see the 1987 market crash page, where I explain the 1929-87 analog model in greater detail). Given the maturity and ripening of the eleven month bull market in stocks, I see some potential for adopting the analog for a short side trade this September.

Why would you want to play the analogy? It seems like such a long shot, doesn't it? There is a tendency around this time of year for paranoia to emerge, market pundits making comparisons to previous crashes like 1929 and 1987. You want to be short equity markets before this paranoia emerges. Fear of crashes and impending declines has a tendency to be self-fulfilling, creating falling prices and on the odd occasion 1987-style crashes. Playing the 1929-1987 analog is the most accurate way to take advantage of the potential emergence of this tendency as it gives you practical entrances and exits.

Need additional proof before putting in a short? The fall season has historically been weak for markets. Look at a seasonal chart of the SP500, you'll see the relative weakness of the market in September and October, and sometimes November. I've gone back to 1895 and found the tendency for fall seasonal weakness to exist through every decade up to the present. There is something about Octobers that affects market psyche, not sure what it is... More evidence? Previous post-collapse markets like Japan in 1990 were only able to sustain rallies of about 10 months from their bottoms (see the Japanese connection ). Given the current length of the rally (about 10 months) it seems better to be a bear than a bull. The CBOE volatility index is ridiculously low, meaning the market has absolutely no fear built into it. When the market is this confident, the trader should be very afraid.

Let me describe the trade a bit more. The next few days are crucial. In 1987, markets peaked on August 25 and in 1929 on September 3. That gives the trader a small and narrowing window to exploit potential for a similar top in September 2003. The Dow and the SP500 clocked in recent peaks on September 4 and September 3 respectively - this lies within the same time zone of the '29 and '87 peaks. Over the last few days, the SP500 near futures contract has been unable to rise above 1030, and the Dow above 9660. Wait for technical confirmation before selling these tops. I'm watching for the S&P near contract to break support at 1021 and the Dow Index to fall below 9500. Only then would I sell these tops. What might one expect from this trade? I'm selling with the expectation of a month, maybe two of weakness. Looking to cover in October.

This trade will become null and void if the Dow and the SP500 break into new highs by the end of today or anytime next week. The Aug 25-Sep 3 zone I wrote of earlier is the span of days in which a peak should have formed. If this is violated, the parallel no longer something I'd feel comfortable trading. Consider this a stop loss, although a time stop loss, not the normal price stop loss you might be used to. Keep in mind this is more of an intellectual exercise than a financial. I am very curious if I am going to be right. The last time the parallel was tradable was in 1999, and prior to that 1997. Both were successful. Let's see what happens in 2003.