Spectral Forecasting And The Financial Markets
by Denis Ridley, PhD.
Ever considered spectral wave analysis to determine cycles in the stock market? Denis Ridley tells you
As a rule, the financial markets are not easy to forecast. While forecasting every future value of the
index is not a reasonable objective, predicting sharp changes in very short-term trends is useful for
trading, and that is really all that should matter to the trader. While we cannot and will not prove or
disprove the random walk theory, we will look at some predictable elements that do exist in the DJIA and
are of interest to the trader. When a variable is observed over a period, the series of observations is
referred to as a time series. Many financial time series have been classified as random walks, in which
case forecasting may be impossible. What sort of cyclical content exists in the Dow Jones Industrial
Average (DJIA)? Are there any predictable elements?
INFORMATION EMBEDDED IN CYCLES
The spectral approach offers a better way of estimating cyclical components. When recombined by the
moving window spectral (MWS) method, the effect of cycles aligning will produce a forecast of major
turning points, which in turn represent changes in the short-term trend. The information needed to make
predictions are embedded as component cycles.