Japan's Golden Week Is Almost Upon Us

Back when I was working for Argonaut Capital in the 90s, Japanese stocks were part of my coverage. It was four or five years into the bear market that followed the 80s boom. The Japanese authorities had noticed that doing almost anything was better than opening the exchange for trade, so they multiplied meaningless holidays and started taking any excuse to close. Or so I remember the manner in which Golden Week became almost an entire week off from the end of April into the first week of May.

I noticed that one of the most powerful seasonal tendencies in the financial markets was for the Nikkei stock index to break shortly before or after Golden Week and decline meaningfully in percentage and time terms. "Sell in May and Go Away" is an adage quoted by stock traders everywhere, but counter to that there is also the market lore of the summer rally, which endures because it works sometimes. But not, it seemed, in Japan, where the very name "Golden Week" seemed a black joke. Leaden Week for financial markets was more like it.

This observation was bankable. In every year of the decade of the 90s, the Nikkei dropped substantially from its pre-Golden Week highs; the biggest drop was 39%, the smallest 9%, the average about 20%.

In 2000 I presented original research on the effect in the late lamented worldlyinvestor.com (hey there Jeremy Pink! Lay off the Dim Sum, will you?) I forecast the same thing to happen that year, and it did: a 20% drop a few weeks after Golden Week, a 30% drop within a few months.

In 2001, the Nikkei peaked in May and plunged.

In 2002, it rose in May but collapsed later that summer.

In 2003,the pattern failed for the first year in fourteen as the index rose 10% in May and kept on trucking.

In 2004, there was a modest 10% loss.

In 2005, there was again no playable decline around Golden Week. However in 2006 there was a rapid loss of 2500 Nikkei points.

Then in 2007, the peak did not arrive until the first week in July, and in 2008 the market rallied in May but began the collapse from which it is still suffering in early June.

One might conclude that the pattern is no longer reliable. Possibly as the effect became more widely known it has been arbitraged away through the action of traders. I always felt things changed since 2000 with the inclusion of a number of key technology shares in the Nikkei that year at very high weightings and the near disappearance of financial sector weightings. Over time the Nikkei has become more like a Nasdaq proxy, and the Nasdaq is having a pretty good year in 2009, all things considered.

I'm positioning for the thing to work again this year, even though at time of writing the Nikkei is only at 8686. That's still 1900 points above its low for the year, at a time that the economy is contracting at rates approaching 10% per annum, world markets may have run out of puff, and there are few redeeming factors in sight.