|  OCT 1992

Seasonal Variations In A Semiconductor Stock by Jack Karczewski

Seasonal Variations In A Semiconductor Stock by Jack Karczewski Do seasonal variations exist in the heavily traded and researched stock market? Recently, a securities analyst was quoted as saying that semiconductor stocks declined an average of 40% from their seasonal highs to their seasonal lows. Jack Karczewski presents this primer on the analytical procedure of investigating seasonal fluctuations in the stock market using a semiconductor stock as an example. Predictable seasonal variations in commodities are both well known and documented. Whether the same kind of patterns can be detected elsewhere, most notably in the stock market, remains to be seen, however. Thus, to analyze cyclical phenomena in stocks, I chose the ratio to moving average method to illustrate the seasonal tendency in a semiconductor stock. This method has been in existence for many years and preceded the advent of computers — including mainframes. It was the method that the U.S. Census Bureau used prior to the adoption of the current Census X-11 program. Now, of course, most traders with access to a personal computer and spreadsheet software such as Microsoft Excel can easily construct a seasonal index on literally any promising time series. For the sake of simplicity, I chose only one stock in an industry rather than a basket of stocks or an index. I will show that a seasonal pattern exists in the stock selected, in this case ca Advanced Micro Devices (AMD), and also illustrate a method that might give a clue to using the seasonal variation to anticipate seasonal price moves. I have presented the material in a way that illustrates the procedure and components for the construction of the seasonal index. Sophisticated programs that will construct these indices are available, but I am presenting a learning exercise in addition to a means of understanding the seasonal nature of a particular stock.

by Jack Karczewski