Saturday, May 14, 2005

Mirror Images of the US Dollar

In a recent posts (here and here) I explained some reasons why I am long term bullish on the US dollar. The three charts you see above will help me explain another reason for being bullish on the dollar.

As you probably know I am a big fan of the late George Lindsay. Lindsay said that the principal method he used to compile his extraordinary stock market forecasting record from 1953 through 1970 was the mirror image chart.

Now you have to remember that Lindsay started out as a graphic artist. His artistic sense led him to look at price charts with the eye of an artist. In particular he loved to find balance and symmetry. The common idea found in all his methods was the idea of a time symmetry around a central point. I've tried to illustrate one of his ideas, the foldback chart, in previous posts (here, here , here , and here). The idea behind a foldback chart is that price highs can be found equidistant in time from a central date, the foldback point. The same goes for lows.

The idea behind a mirror image chart is much less obvious and much more daring. As with the foldback chart, one first must identify a central date, the mirror date. Typically this is the date of a major high or low or of a test of that high or low. Then the mirror inmage chart predicts that each important price extreme after the mirror date will be the same number of days after the mirror date as an equally important extreme was before the mirror date. So far this is the same rule as in the foldback chart.

Now comes the surprise. If top occurs some number of days before the mirror date, the mirror chart predicts that a low (the mirror image of the correponding top!) will occur the same number of days after the mirror date. Similarly, a low some number of days before the mirror date predicts a top the same number of days after the mirror date.

Why should this ever work? Beats me. But one thing is for sure, once you have identified a mirror date the chart can continue making quite accurate predictions for years into the future. And remember, Lindsay himself credited his outstanding forecast record to mirror image charts.

The hard work in constructing a mirror image forecast always occurs when trying to locate a mirror date in a chart. Often I can' find any mirror date in which case I must turn to other methods. Once I have a candidate for a mirror date I check the accuracy of the first few forecasts. Typically it takes some time for the mirror forecasts to start affecting the market so it is not unusual for the first one or two forecasts to fail. But as soon as a chart's mirror forecasts start working I know I have a valuable guide to future market behavior.

The charts ( 1, 2, 3 )you see above this post are monthly charts of the US Dollar index. The April 1995 low is labeled as point X. This is the mirror date I have chosen. The extremes before the mirror date are labeled with single capital letters: A, C, E, G are lows and B, D, F are highs. The corresponding predicted mirror extremes are labeled with the same letter doubled. For example, the low at C predicts a high at CC.

The first forecast of the mirror image chart was for a high at AA in November 1997. This was wrong. But the next forecast for a low BB in January 1999 was very accurate. The high at CC and the low at DD were similarly good forecasts. The forecast high at EE was a year late. The market had already dropped 10% by that time but would drop another 27% after point EE. Finally, the mirror chart predicted an extend drop from EE to FF in June 2005. I think the June 2005 prediction will turn out to be 6 months late, but again it was within 10% of the low.

What does the mirror chart predict now? An extended bull market in the dollar! The earliest date for the predicted dollar top is point GG (not shown on the chart) in January 2010. I expect the dollar to make it back to 121 during the next five years!

No comments: