Are Tech Book Sales a Leading Economic Indicator?

The Holy Grail of those working in marketing, strategic, or financial departments of corporate America is to predict the future. If you happen to be in the numbers game, then you watch, collect, and plot as much data as you can get your hands on, hoping that in some of this data there is a hidden pattern that divines the future. These numbers might include billings, purchase orders, inventory, and customer sales, plotted weekly, monthly and annually. In addition to the above, it is always instructive to plot this data against as many economic indicators as possible to see how your business is doing against the general economy. And if by chance the pattern of your data matches some of the well-established leading economic indicators, whose historic behavior is well documented, then you just may have that crystal ball you are searching for.

O'Reilly & Associates is in the technical information business. We publish books, host conferences, and publish articles on emerging technologies on the O'Reilly Network. Much of our past success has been in publishing books on highly technical topics such as Unix system administration, Java and Perl programming, and more recently, peer-to-peer networking and the emerging field of bioinformatics. Our customers for these books are programmers and system administrators who are in the technical trenches. They typically find our books in bookstores across the country and through online vendors such as Amazon.com, Fatbrain, and B&N.com. It should come as no surprise that our three largest domestic retail accounts are the big retailers, Borders, Barnes and Noble, and Amazon.com.

Of particular interest are the sales trends of our technical books at some of these retail accounts. Specifically, when the sales data are plotted against a select group of economic indicators published by various government agencies, some surprising patterns emerge. The goal of this article is to expose some of these patterns, offer possible suggestions as to the cause, and propose that the data might warrant a closer look.

Predicting Earthquakes

I am not an economist, so I am not in a position to make economic forecasts. I do, however, have a background in quantitative observation and in comparing seemingly disparate sets of data in interesting and sometimes surprising ways.

Before coming to O'Reilly, I worked as seismologist with a team that gathered seismic data in California. The data was plotted and analyzed in an attempt to glean patterns that might provide clues towards predicting future behavior. However, earthquake data wasn't the only information our research group gathered. We also gathered data on radon levels in water wells, magnetic field fluctuations, the tilt of the earth's surface in and around fault systems, the behavior of domestic animals, specifically cats and dogs, before and after earthquakes, and the behavior of migratory birds. All of this data was plotted against time, and analyzed for correlations and suggestive patterns. Despite the seemingly disparate nature of the data, there were surprising correlations that have been useful in earthquake prediction.

-- M.S.

Business Cycles and Economic Indicators

Before diving into the data, a little background discussion is in order. Recalling some of the early lessons learned in a macroeconomics class, businesses and the economy follow what is aptly termed "The Business Cycle." The Business Cycle has five phases:

1. The business cycle peak
2. The trough
3. The recovery
4. The expansion
5. Back to a new peak

In plotting the performance of any company over a long enough time period, these cycles can be identified time and again. Tracking an economic indicator that correctly identifies or better yet, predicts, the inflection points of the peaks and the troughs is the Holy Grail that we are seeking and attempting to identify and quantify.

Leading economic indicators are ones that have successfully and repeatedly predicted past economic downturns and recoveries. These indictors are evaluated for their economic significance, conformity to the business cycle, availability, smoothness, cyclical timing, revision cycle, and statistical adequacy. The U.S. Department of Commerce computes and maintains an index of Leading Economic Indicators. The components of the index come from the following groups of cyclical indicators:

• Employment
• Sales, orders, and deliveries
• Prices
• Fixed capital investment
• Personal income
• Money supply and stock prices

One of the easiest economic indicators to watch is no further than the business section in your morning newspaper: the stock market. The stock market is a very sensitive indicator of business peaks and troughs, because stock prices reflect both the historical performance and future expectations of a company's performance three to six months hence. According to the Federal Reserve Board, the stock market is one measure of the current value of the nation's stock of capital and is often viewed as a barometer of business and consumer confidence regarding the future. A high or rising stock market may signal robust growth of business investment and consumer spending in the near future, while a low or falling stock market may signal sluggish spending. Standard and Poor's Index of 500 companies (the S&P 500) is one of the series of indicators used to compute the U.S. government's "Index of Leading Economic Indicators."

As economists astutely point out, the stock market is not a perfect leading indicator. That being said, they also point out that an economic downturn has always been preceded by a stock market decline.¹ And, taking a historical look, on average, changes in the stock market precede changes in the general economy by about six months. According to economist and business cycle expert Geoffrey H. Moore (not to be confused with Geoffrey A. Moore of Crossing the Chasm fame), as a business expansion cycle begins to slow down, production costs rise and profits fall. At the same time, there is an increased demand for capital to cover shortages in cash flow or rising expenses due to rising inflation caused by the business cycle expansion. According to Moore, the increased demand on capital causes interest rates to rise. Rising interest rates exert a negative pressure on profits and stock prices drop even as the business activity continues to expand.²

[1] [2] [3] [4] Next