Expanding AIER’s Business-Cycle Research

Thursday, August 4, 2016
Chart 1. Over 144 years, the U.S. equity index shows many ups and downs but also long periods when prices trended flat or up.
Chart 2. U.S. equity price cycles, 1948 –2015
Chart 3. U.S. earnings per share, 1948 –2015
Chart 4. Can the BCC model help investors in making portfolio decisions? Our first step is to look at how the Leaders compare with the S&P 500.
Our new model for monitoring the U.S. economy could be a valuable tool for investors as we adapt it to global economies and develop guidelines for how investors can apply it

In January AIER introduced our updated Business-Cycle Conditions model. The purpose of AIER’s model is to identify turning points—peaks and troughs—in the economy. It combines a set of economic indicators in a way that anticipates these turning points and enables us to statistically analyze their significance. 

AIER’s Business-Cycle Conditions original model was introduced in 1950. Changes in the economy and changes in data since then necessitate periodic reviews of the model and the indicators we have chosen. The new model maintains our commitment to scientific economic analysis and the overall structure of the original. But we have added new economic indicators and dropped those that were no longer effective. We tested the new mix of indicators for more than a year to see how well they performed.

Now that we have used the new BCC model for several months and demonstrated its effectiveness in telling us what is happening in the economy, we are launching two new areas of business-cycle research: applying the new model to investment decisions and adapting it to analyze economic conditions on other countries.

Going global
Business-cycle research at AIER has so far focused on the U.S. economy, but that is changing. Over the eight decades since AIER’s founding, the world has become more interconnected. The exchange of information, trade in goods and services, the flow of capital, and the movement of people have all expanded exponentially. Moreover, investments in global equities are now part of many investors’ portfolios.

Though the U.S. still dominates equity markets, there are numerous other markets available to investors. The S&P Global Broad Market Index, for example, has more than 11,000 stocks from 25 developed and 22 emerging markets. While U.S. stocks account for about 3,000 of these and 51.8 percent of total market capitalization, that leaves more than 8,000 securities and 48.2 percent of market capitalization outside of the U.S.

By considering global markets, investors can further diversify their portfolios. We believe that expanding our business-cycle research and applying our model methodology to monitor more of the world’s major economies will provide investors with insights that will allow them to make informed decisions about global investing. We have already begun working on models that apply to other countries, and we plan to introduce them over the next few years.

A fresh look at business cycles and investing
The 1950 BCC model (or Statistical Indicators, as they were known) aided AIER researchers in calling the 1953 recession about one month before it began. That was followed by AIER warning of a recession in January 1957, seven months before it began. Since then, AIER researchers, guided by the BCC model, have amassed an enviable track record by identifying eight more turning points in economic activity and alerting investors and consumers about impending recessions months ahead. Most recently AIER sounded an alarm in April 2007, eight months before the Great Recession began.

For much of its history, AIER has held that investors who are knowledgeable about business cycles and their impact on equity markets could protect their wealth by adjusting their portfolio allocations in anticipation of peaks and troughs. We have recommended that investors reduce their portfolio holdings of cyclically-sensitive securities and increase their holdings of cash and gold before an expected recession. This approach then calls for increasing exposure to cyclically-sensitive securities again during a recession when they are more favorably priced.

In preliminary “back-testing” of our new model, we are using theoretical portfolios and developing decision rules, based on the results shown by the model, for increasing and decreasing equity exposure at different points in the business cycle. If we find in our testing that using our new BCC model can help boost returns, it would help investors make decisions to adjust their portfolios during long periods of zero-trend growth in U.S. equities, for example—those periods when taking no action could increase the risk that they might not reach their long-term investing goals.

Sign up here to learn how small businesses meet the challenge—and succeed—because of rather than in spite of their size.

Testing underway
In one test, we began with a theoretical portfolio of $10,000 in cash and $90,000 in stocks. We established a rule to reduce the portfolio’s stock exposure by 10 percent when our BCC Leaders index fell below 34 and was declining (the 50 level in our model is neutral—when most of the Leading indicators are above that, we are in an economic expansion, and when they fall below that, the economy is considered to be contracting or softening). In the same experiment, our rule called for decreasing cash exposure and increasing equities by 20 percent of available funds when the Leaders went above 8 and were rising. After allowing a one-month interval for a hypothetical investor to implement the new positions called for by the rules, and allowing for transaction costs, we found that the theoretical portfolio beat the performance of the Standard & Poor’s Composite Index of 500 Stocks by 6 percent. This illustrates the power of using the BCC model for making decisions. These rules, however, require more testing.

Further testing will provide more information that will allow us to refine the rules and develop guidelines that we can share with investors, providing them with a useful tool to help keep their portfolios on track. The timely results available with the new model will make this practical. The old procedure involved bringing all of AIER’s researchers together to individually evaluate each indicator once a month. The results were printed and mailed to members, a process that could take weeks, making use of the model for investing less effective. The new process is completely quantitative. We can run the model anytime, as often as desired, and disseminate the results quickly with web-based tools.

Historic stock market trends and how they affect investing
An initial review of the long-term behavior of stock prices in the U.S. reveals a number of interesting characteristics. Between 1871 and 2015 nominal equity prices increased 46,163 percent, or a 4.3 percent annual return excluding dividends. While the overall performance has been positive, equity prices have registered several periods with distinctly different trends (Chart 1). We have identified eight of these over the past 144+ years. Each period is bookended with the date when the U.S. equity index reached a particular value for the first and last time. For example, the 41-year period that began in April 1901 and ended in April 1942 is identifiable because 1901 was the first time the U.S. equity index crossed the value of 7.84, and 1942 marked the last time the index recorded that value. Certainly the index increased and declined in the years that include the 1929 crash and the Great Depression, but it has never again fallen to 7.84.

Of these eight periods, four reflect positive growth (1896–1901; 1942–1965; 1978–1996; and 2009–2016) and four reflect zero growth (1871–1896; 1901–1942; 1965–1978; and 1996–2009). Before World War II, 66 out of 71 years saw overall flat trends for U.S. equities. Since World War II, 49 out of 73 years have shown overall positive trends. It is notable that stock performance in the post-World War II era has been more favorable compared with the pre-war era.

Why is all of this important? A strategy of simply buying and holding a broad market index of stocks doesn’t always produce positive returns. Periods of flat equity markets have been shorter in recent decades, but whether that pattern will continue is unknown. The prudent use of business-cycle research could be particularly important in boosting returns during long periods of flat growth in equity markets.

Business cycles correlate to stock and earnings cycles
Most historical economic indicators can be analyzed in terms of either a time trend or a cyclical component. We can more easily identify just the cyclical component of equity prices (Chart 2) by removing the upward trends in equity prices from the positive-trend periods. Since 1871, dozens of equity-market cycles have occurred, and 29 business cycles have been identified by the National Bureau of Economic Research, the organization that officially dates the beginning and end of U.S. recessions and expansions. Nearly all the economic cycles are closely associated with a significant equity-price cycle. For example, during the 2007 through 2009 recession, as Chart 2 shows, the equity-price cycle index peaked slightly before the recession began (before the gray shading) and bottomed slightly after it ended. Around the 2001 recession, the cyclical index is shown peaking several months before the recession began and bottoming several months after it ended.

Over the past several decades, the exact timing of peaks and troughs in the equity-price cycle have varied in relation to the business cycle—they may not always coincide exactly, or at all, as the chart shows, making it difficult to adjust investment portfolios quickly enough to take advantage of declines in equity prices. That is where the new model promises to offers an advantage. While it is not perfect, it anticipates changes in the economy, particularly major ones, with enough advance warning that investors could make portfolio adjustments based on its information.

The link between business cycles and equity-price cycles is through corporate earnings. Modern finance theory holds that current equity prices reflect expected earnings, so changes in equity prices over time should also be related to earnings expectations. The key question for our review of business cycles is: Do economic conditions, and business cycles in particular, affect expected earnings?

To explore this question, we examined the performance of actual earnings per share over the past 144 years. (Chart 3 shows nearly 68 years of that period.) For the 29 business cycles since 1871, almost all have been associated with a significant decline and recovery in actual earnings per share. This suggests that business cycles do have an impact on earnings per share. However, as with equity- price cycles, earnings-cycle peaks and troughs do not always exactly coincide with business cycles.

In the context of adjusting investment portfolios to take full advantage of knowing how economic fundamentals are affecting the macro economy, we need to further explore how effective the revised BCC model is for helping investors anticipate business cycles. Our first step is studying the performance of the BCC Leaders index vs. the S&P 500 (Chart 4). In future research we will test a wide range of guidelines that may help investors adjust their portfolios based on economic conditions.

In addition to the back-testing being done with our Leaders index and U.S. equity prices, AIER researchers are analyzing and testing possible relationships between the Leaders and the overall earnings cycle. Another initiative is to test our Leaders index against the price behavior and earnings cycles for particular economic sectors, using equity-price indices reflecting stock performance of the health-care or the financial sectors, for example.

Finally, AIER will work on testing BCC models for other countries against the equity markets and earnings cycles for those countries.

Looking ahead
Business cycles affect fiscal and monetary policy, business and financial regulation, and wealth and living standards. Our current research is about whether our newly updated BCC model could help protect investors from economic downturns and whether it could potentially boost returns during long periods of flat-trending equity prices. Now that we have updated our BCC model, prudent behavior warrants a thorough review of its potential use. Our preliminary results are encouraging, but more work needs to be done.

Chart 1. Over 144 years, the U.S. equity index shows many ups and downs but also long periods when prices trended flat or up.
Chart 2. U.S. equity price cycles, 1948 –2015
Chart 3. U.S. earnings per share, 1948 –2015
Chart 4. Can the BCC model help investors in making portfolio decisions? Our first step is to look at how the Leaders compare with the S&P 500.