Conflicting Commentary on Technical Analysis

WALL STREET SMARTS, THE BLOG, IS NOW WALL STREET SMARTS, THE BOOK.  FULLY EDITED AND REVISED WITH NEW MATERIAL ON AMAZON

You may recall Burton Malkiel's discussion of some research done on technical investing in last week's blog.  He said that the results showed little, if any, correlation between past and future prices of stock.  So, we know that Professor Malkiel does not hold much store in technical analysis.  A proponent of technical analysis might, rightly, ask for more details about the scope of the tests.

Norman G. Fosback, a well respected investment adviser and co-founder of The Institute of Econometric Research, disagrees with him.  Fosback's enduring work, Stock Market Logic, A Sophisticated Approach to Profits on Wall Street, was first published in 1976.  The twenty-second edition was printed in 1993.  In his book, Mr. Fosback took random walkers, such as Professor Malkiel, to task.  He criticized the academic studies on which their conclusions about technical analysis were based; writing that the research conducted by them was superficial and not done in sufficient depth to support their conclusion that market timing was of no value.  His research led to a different conclusion.  In referring to a table of daily price/volume relationships for the period 1965-1972 depicted in his book, he wrote:

The most important conclusions to be gleaned from the table are: (1) rising price is a relatively bullish portent for future price changes and falling price is relatively bearish; (2) rising volume is even more bullish and falling volume is relatively bearish; (3) rising price accompanied by rising volume is the most bullish of all and falling price accompanied by falling volume is consistently bearish.  Almost all price/volume tables created with alternative averaging and change parameters,  be they in terms of days or weeks, and the author has created dozens of them, show similar results and lead to identical conclusions.*

Based on his research, Mr. Fosback concluded that there was a correlation between historical price and volume levels with future price movement.

There are two types of research: deductive and inductive. The researcher either starts with a hypothesis and finds supporting facts (deductive) or starts with facts and finds the resulting hypothesis (inductive).  They are mirror opposites. Mark Twain popularized the statement, "There are three kinds of lies: lies, damn lies and statistics."  Depending on your level of cynicism, this may be one of the truest statements ever made.

I suspect that you can find research results to support just about anything you want.  As individual investors, we must decide what form of investing works best of each of us and stick to it.   Consistent application of an investor's chosen strategy is the only way to improve the results of his or her portfolio.

We will continue looking at technical analysis in the next blog.

* Excerpt from Stock Market Logic, A Sophisticated Approach to Profits on Wall Street, Norman G. Fosback, copyright 1976, 1993, The Institute for Econometric Research, page 148

As always, comments are welcome.