The First Quarter La Jolla Economics Industry Selector

Print ArticleEmail Article

Investors who pay attention to government regulations and technological innovations may be able to anticipate industry responses to aggregate demand shocks. In doing so, they may be able to anticipate changes in the alpha and beta coefficients of the various industries and from there implement a successful industry selection strategy. Using historical evidence, we have developed a classification of how each industry performs under different economic conditions. In particular, we have examined the output, employment, and price responses of the different industries over time. Our statistical procedure estimates industry responsiveness to the economic cycle based on the average response over previous cycles. Our estimates are dependent on the availability of quarterly historical data. The shorter and more volatile the earnings history, the less reliable the estimates produced by our model. Given a forecast of the future economic environment, we set out to find previous instances where either similar or directly opposite shocks occurred. The rationale here is that industry groups respond to shocks in the same manner. Put another way, industries behave in a similar way when experiencing comparable shocks. Opposite events also contain information as they provide an inverse ranking of the relative performance expected under a forecast scenario. Our basic stock research focuses on the S&P 1500, so it’s only natural that we will also use the economic sectors and industry classification utilized by S&P. In what follows, we will use the Global Industry Classification Standard (GICS). In particular, we will focus on two parts of the GICS structure: The economic sectors and the industries.

Industries that respond to shocks with above average employment increases and below average profit gains are identified in our classification as elastic industries. In turn, industries that respond to shocks with below average employment increases and above average profitability gains are classified as inelastic. When an industry is inelastic, it is difficult for its companies to adjust production. Inelastic industries have pricing power and an increase in demand means above average profits. Conversely, a decrease in demand means an above average decline in profits. Stocks of companies in inelastic industries are what the financial literature call high beta stocks. In contrast, elastic industries (i.e. industries with beta in the neighborhood of 1) experience little or no change in profitability relative to the market. Most of the adjustment of elastic industries to changes in aggregate demand will occur in the form of changes in employment levels.

LJE generates three different estimates of general market conditions: The likelihood that equities will outperform fixed income and the likelihood that each size and style will outperform or underperform the market. The strategy also needs to figure out the nature of the supply response (i.e., elastic or inelastic). To this end, we generate two economic sector or industry specific forecasts: The likelihood that a sector/industry will appreciate and the likelihood that the sector/industry group will outperform the S&P 1500. From all this, a common sense strategy emerges: Buy inelastic (high beta) industries undergoing positive shocks and short inelastic (high beta) industries undergoing negative shocks. The methodology for industry selection is fairly straightforward (a description of the LJE Industry Valuation Model is presented in the appendix). A visual representation of what we have in mind can be found in Table 1. The sectors/industries highlighted in green represent the overweighted positions while the red highlighted sectors/industries represent the underweighted positions in the overall portfolio. Since a 50% ranking suggests that an industry has an equal chance of outperforming as it has of underperforming the index, we take the position that a 50% ranking denotes a neutral or market weight, and those sectors/industries will be highlighted in yellow. A ranking in excess of 53% suggests the industry has a better than 50/50 chance of outperforming and thus merits an over-weighted position in the portfolio. Similarly, a ranking below 50% suggests that the industry has a less than 50% chance of outperforming the index and thus merits an underweighted position in the portfolio. The amount of overweighing and underweighting is proportional to the deviation of the overall rankings from the neutral 50% ranking.

Applications: The Tilt Strategy

A summary of the performance of the benchmark sector portfolios during the fourth quarter are reported in Table 2. Information on the cap-weighted returns for the 10 economic sectors is reported in column 1, row 1. Since the cap-weighted average of the 10 sectors make up the S&P 1500 index, it follows that the information presented in row 1, column 1 is nothing more than the 5.83% return delivered by the S&P1500 during the quarter.

The 10 rows in-between report the returns of the economic sectors posted during the quarter. For example, the second row indicates that during the second quarter, the Consumer Discretionary sector posted a 7.88% gain.

There are a number of possible applications to the economic sector and industry forecasts. One is a natural extension of the process used to estimate the likelihood that an industry or sector will outperform the benchmark. We have chosen to build a cap-weighted portfolio that overweights and underweights the individual sector groups in direct proportion to their rankings. For example, tilting the economic sectors in direct proportion to their rankings produced a 5.67% gain during the fourth quarter (Table 2, Row 1, Column 2). The tilted portfolio underperformed the S&P 1500 by 16 basis points during the quarter.

A natural extension is to apply the strategy to industries within a specific sector. Examples of this would be to overweight industry groups expected to outperform their economic sector and underweight industry groups expected to underperform their economic sector. The second line of Table 2 reports the returns of the Consumer Discretionary sector. The first column shows that the Consumer Discretionary sector posted a 7.88% gain, while tilting the industry groups within the Consumer Discretionary sector produced a 6.36% increase. Tilting the industry groups within the Consumer Discretionary sector produced a 152 basis point underperformance.

The sixth row shows the performance of the Health Care sector. During the quarter, the sector gained 8.69%. Tilting the industry groups within the Consumer Staples sector added to the performance. The tilted portfolio gained 9.76% beating the Consumer Staples sector’s performance by 107 basis points.

Obviously the LJE industry tilting strategy adds value in some of the sectors while it does not for other sectors. In order to facilitate a visual analysis of the possible strategies we have bolded the tilted portfolios that outperformed their benchmark. Whether the strategy adds value on balance or not can be determined by comparing the performance of the strategy that tilts all the industries in the index in direct proportion to their rankings. This result can be found in the last row of Table 2. The industry tilting strategy gained 5.80% while the S&P 1500 posted a 5.83% gain. During the fourth quarter the industry tilting strategy subtracted 3 basis points.

This section outlines two alternative tilting strategies. One focuses solely on tilting the economics sectors, the first row, while the other strategy tilts the industries within the sectors. The net effect of the industry tilting strategies can be found in the last row of the table. The information presented indicates that economic sector and industry tilts gained 5.67% and 5.80% during the quarter. Both strategies were behind of the benchmark which gained 5.83% during the fourth quarter (Table 2).

On a year-to-date basis the tilting of the economic sectors has gained 18.80% and is ahead of the 18.60% posted by the S&P 1500 (Table 3). In contrast, the industry tilting strategy with a 17.88% year-to-date gain is lagging the S&P 1500. So far the data suggests that the broader forecast has proven to be more accurate than the narrower industry specific forecast.

Application: The Buy, Hold and Sell Portfolios

Another application is to simply use the rankings to identify winners and losers. One simple strategy that comes out of this is to classify the groups into buys, holds and sells. There are a number of potential problems with this approach. First it does not take into consideration the strength of the rankings, i.e. how far above or below average the ranking for a particular industry or sector is. A second issue is the weighting scheme. One way to calculate the performance of the three portfolios is by equally weighting the industry groups within each category. In columns 3, 4 and 5 of Table 2, we report the performance of buy or above average, hold or average and sell or below average recommendations of the industry rankings reported in our last industry publication. However, for the year to date numbers we only include the cells for which we have a fill in for each of the quarters in question. In order to facilitate the visualization, we have bolded the buy cells that outperformed the benchmark and the sell cells that underperformed their benchmark.

Looking at Table 2 it is apparent that the buy, hold and sell classification worked to perfection when applied to the industries in the Health Care and Information Technology sectors. Since not all the cells are bolded, it is fairly apparent that the classification is not perfect.

Other Applications

There are many other possible strategies that could be developed using the sectors/industries forecast:

• The overweight/underweight could be converted into long/short industry portfolios.

• Since our forecasts are logically consistent with each other by construction, a weighted average of the forecasts adds up to the next level index, and we could use this approach in several ways. In addition to selecting individual industries likely to outperform, we could also use the individual industry forecasts and compare them to similar industries in the same economic sector and make sector neutral bets.

• Another possible application is to go further upstream to S&P/Citigroup indices and focus on the industry groups relative to the benchmarks and make market neutral bets.

• Going downstream, we could use the industry forecast and individual stock forecast generated by LJE to develop long/short industry neutral stock portfolios.

Appendix: The LJE Industry Valuation Model

LJE has developed a comprehensive industry valuation approach that overlays macro-, or market-level, forecasts and then narrows the process to end up at the industry level. The process uses multiple valuation screens to provide a reliable gauge of investment opportunities for industries and other economic sectors. When factors in the model reinforce one another LJE is able to make recommendations that have proven historically to be extremely accurate. Alternatively, when factors contradict one another the level of confidence in an industry decreases, leading to a more neutral recommendation from LJE.

The LJE scoring process begins with a measure intended to capture general aggregate shocks to the economy and markets in general. The Stock Market Appreciation Potential is an estimate of the cumulative bullishness or bearishness of the stock market. It incorporates general market effects that may add to, or subtract from, the returns of individual industry groups. In other words, a rising tide will lift all boats, just as an ebb tide will draw most boats down. For each industry group, the LJE process also identifies all the stocks in the different earnings-growth categories (value or growth) to generate an industry “style effect,” as well as the different market capitalization categories (large-cap, mid-cap and or small-cap) to produce an “industry size effect.” The idea is that similar stocks and or industries will in general move together: that “neighborhood effects” are important in the industry-selection process.

Thus, if we are bullish about, say, large-cap growth stocks, that bullishness would contribute to our assessment of any particular industry.

Next the LJE scoring process moves down to a narrower focus, the individual industry groups’ appreciation potential. This measure uses past earnings in order to capture the earnings momentum of an industry. The relationship between the current and past values of earnings and interest rates also helps capture the effect of overall economic conditions on an industry’s earnings. For example, the expectation of higher interest rates induces people to accelerate purchases of consumer durables and other items. This in turn leads to higher profits. Over time, that profit surge is offset by below-average purchases.

One tool that is valuable in assessing an industry’s earnings-sensitivity to changes in interest rates is the change in the slope of the “yield curve”. The changes in the slope of the yield curve (e.g. upward-sloping, flat, and inverted) signal the current and future state of the business cycle thereby allowing us to make inferences and forecasts about the future health of the economy and the stock market. In addition to the industry appreciation potential, the LJE process also identifies the industry appreciation potential relative to the market. Doing so removes any trend effects meaning there is the potential for an industry to outperform the market – even perform well above the market – irrespective of general market conditions.

Table A reports a summary of the rankings of the economic sectors for the coming quarter. Three signs reinforce the buy rating for Auto Components. The overall market appreciation potential, the industry group appreciation potential relative to the sector and the industry appreciation potential relative to the market all indicate that the sector is undervalued and poised to rise. Auto Components resides in the overweight or bullish range.

Victor A. Canto
Andy Wiese
La Jolla Economics
www.lajollaeconomics.com