Tip of the Week #111                   Tip Index       Categorized Tip Catalog

Go to the Prior Tip  Critical Chain Project Management by Lawrence P. Leach
Go to the Next Tip  "Forecasting Shareholder Value: The Missing Objective in Balanced Scorecards"
Return to MaxValue Home Page

Ahead of the Curve: A Commonsense Guide to Forecasting Business and Market Cycles

by Joseph H. Ellis
2005, Harvard Business School Press, hardcover, 276 pages, US$29.95 list  ISBN 1-59139-691-3

Based upon a glowing review in Strategic Finance and the timely topic, I rushed out to buy this book. It has been worthwhile reading.

Joseph Ellis was a partner (and remains an advisory director) of Goldman Sachs and was ranked Wall Street's #1 retail industry analyst by Institutional Investor for 18 consecutive years. His foremost job at Goldman was advising retail clients with forecasts for the economy, their industry, and specific firms within the industry. He claims success with both the general US economy model (later compared with other countries) and with several Goldman Sachs' models for retail store sales.

As with all forecasting books that I've read—hoping to find a magical technique—the evidence isn't convincing as the author would have you believe. Yet, Ellis's story is well-written and persuasive. He supports his hypotheses with logic and historical data.

Key findings from over 40 years of experience:

• Key cause-and-effect relationships are evident in most business cycles.
• Many of the popular indicators lag rather than lead the economy. (For prediction, we are interested in only leading indicators.)
• While there are many interrelated elements, yet what drives the economy primarily are:
  1. Consumer spending is two-thirds of the US economy (GDP) and dominates.
  2. Personal Consumption Expenditures (PCE, a.k.a. Real Consumer Spending) is the best predictor of Consumer Spending
  2. Consumers' real* hourly earnings is the best leading indicator of Personal Consumption Expenditures. Interest rates are also significant.
• The two dominant, popular consumer sentiment surveys have little predictive value. The same with employment.
• Ellis maintains that charts are the best way to understand the data. I believe him. If we are to believe the identified cause-and-effect relationships in the business cycles, one needs to appreciate the highly-variable delays in the system dynamics. This confounds attempts to quantify the relationships. For example, the time for changes in Individuals' real* hourly earnings to affect changes in Real Consumer Spending is 0-9 months.
• He uses rates of change in the most-useful leading indicators to predict key inflection points—changes in the direction in trend. For the rate of change calculation, he uses the ratio of the current month (quarter) over the corresponding period one year earlier. If the data are still too-noisy, then average-in the prior two months.
• Most companies are not doing an adequate job in forecasting their businesses and industries. [This book should encourage more people to study and apply economics and time-series statistics.]

        * "Real" means that inflation is adjusted-out.

Ellis offers an outlook forecast as he finishes the book in early 2005. He is concerned that the trade and budget deficits and a weak US dollar cause the possibility of sharply rising interest rates.

A great feature of the book is a companion Web site: http://www.aheadofthecurve-thebook.com.  There, Ellis (and presumably with helpers) is maintaining the key charts in the book (and links to data sources). With data through 2005 there are worrisome trends: Decreasing Real Average Hourly Earnings (though with a recent uptick) and increasing interest rates. These portend a drop in the relatively lofty Real Personal Consumption Expenditure, Ellis' primary leading indicator for the economy (and stock prices). You may want to sell. It will be interesting to see how this plays out.

For executives in enterprises sensitive to business cycles (which doesn't include petroleum) and for investors, I highly recommend this book. For investing, the priorities for portfolio management are, in order: when (market timing), what sectors (industry concentration), and  which companies. This book should provide ideas for better forecasting at each level.

—John Schuyler, Mar. 2006.

Copyright © 2006 by John R. Schuyler. All rights reserved. Permission to copy with reproduction of this notice.