October 23rd, 2009
I bet you think the above graph is my EKG as I look at the market’s gyrations. It’s actually an oscillator, a momentum indicator that measures the market’s gyrations, on a rolling basis, in terms of each month’s prior twelve-month return. The chart above covers the past 81 years from January, 1929 to January, 2010 (click on image for larger and sharper view). Prior to 1939, the data is the DJ-30; data after September, 2009 are based on my assumptions (to see the spreadsheet with data, click here).
You often hear or read about Fibonacci lines, arcs or fans that many say tell you that a move in the market or stock is over-extended because it has reached a critical benchmark like a 50% retracement. But there are few measurements that are linked to a time dimension, like 12-month returns; the above chart does.
I think it’s an amazing picture. What I find most amazing is its regularity. Usually, after the market has several back-to-back exceptional 12-month runs (where returns over the previous 12-month exceed 25%) it is followed by several back-to-back months of negative 20% return.
Conversely, when the market has a sharp sell-off, or negative back-to-back 12-month returns of over 30% (see 1932, 1937, 1974, 2002, 2008), there follows a bounce back of 25-30% returns. It’s fairly regular and predictable.
What does this mean for us today? I appended a couple of hypothetical scenaria through June, 2010 to see what the graph would look like (the red lines at the right). Case 1 assumes the Index will remain flat (in practice fluctuate in a narrow range around…) yesterday’s close of 1057; Case 2 assumes the Index will continue to move ahead and peak at 1125 in February and retreat moderately back to 1075 by June, 2010. The result is a spike equaling 1983’s spike out of that recession and the 1997 bounce.
In any event, nothing continues growing forever. Moves much beyond the extent of repair that’s already been achieved is inconsistent with market history going back to the Depression. That’s not to say there will be a significant decline sometime in the near future. The other possibility is that the annual changes moving forward will decline to the point that somewhere in 2010, the market will be no higher and, yes, possibly lower than 1080 (I spelled out two possible forms the correction might take in “Two Market Consolidation Models: 2004 and 1933-35“). But I feel technical evidence is showing that we’re quickly approaching the top of these climb.