Copyrighted by ISS EVA and reprinted with permission by ISS EVA. PDF available here and webinar here.
.
Introduction
.
Creating business value is about generating profits. More specifically, equity investors gain when profits are greater than they require for the risk borne. Then the firm produces economic profits or adds economic value (EVA is economic value added), which drives up firm value so that market value added (MVA is the firm’s value less the book value of contributed capital) is created.
.
Investors care about EVA. The higher the EVA Margin (EVA / sales), or profitability, the higher the MVA Margin (MVA / sales), or valuation (figure 1). The median EVA Margin is 0.9 percent for the ISS EVA US universe of companies with market capitalization of $250 million or greater, which means that, on balance, firms create value for shareholders, and this is rewarded with a positive MVA Margin (the median is 123 percent of capital).
.
However, not all companies generate economic profit. The median small firm (approximately the smallest third of the sample as measured by market capitalization) destroys value (-1.1 percent in EVA Margin). Small companies are also punished for this value destruction with a low MVA Margin (3 percent). The gap between large and small caps is wide, as is the gap between high and low EVA Margin companies in general.
.
Defensive sectors including FactSet’s sectors for communications, consumer non-durables, health services, and utilities, have a higher EVA Margin than cyclicals, but they trade at a lower valuation, perhaps because these sectors generally produce lower growth. However, defensive companies may also exhibit less risk, which should be rewarded with a higher value.
.
EVA drives value, so corporations should strive to add EVA. Plus, investors may gain by identifying opportunities of disconnects between profits and valuation – high EVA Margin / low MVA Margin stocks outperform.
.
Current Profit Levels Drives Value, Especially for Mid Profitability Stocks
.
EVA Margin explains 42 percent of the variation in MVA Margin for median companies across sectors (figure 2), excluding FactSet sectors for miscellaneous, technology services, and health technology. An R-squared figure of 42 percent is very strong, especially since EVA Margin is just one of three drivers of MVA Margin. Risk and growth also drive valuation (see equation 4 of box 1). MVA Margin = EVA Margin * (1+g) / (WACC – g). There are only a few companies in the miscellaneous sector, so this explains its lack of reliability and exclusion from the analysis. Investors’ high expectations for growth from health technology and technology services explains these outliers (they have negative EVA Margins, but very high MVA Margins).
.
.
Firms with mid-level EVA Margin are most closely correlated with MVA Margin (R-squared equals 54 percent), and those in the top and middle thirds have a combined R-squared of 40 percent (figure 2). The bottom-third EVA Margin stocks have no correlation with their MVA Margins. Investors give these firms the benefit of the doubt – while they should trade at negative MVA Margins based on their current EVA Margin, their MVA Margin is essentially zero. For companies in the bottom third of the sectors, the average MVA Margin is -1 percent, or about book value of invested capital, while median EVA Margin equals -9 percent. These companies are probably associated with higher risk, so investors are paying up for expected growth. It is later shown that those expectations are often too high, as expensive low EVA Margin firms tend to underperform. High EVA Margin firms are also uncorrelated with MVA Margin (R-squared of 4 percent). The market apparently expects profitability to deteriorate, which could be due to cyclical forces or due to advantages that generated the high EVA Margin being competed away. It is also possible that some of these companies are mispriced and cheap, and their stocks may outperform (as is shown later).
.
For additional insights on the relationship between EVA Margin and MVA Margin, see EVA, not EBITDA: A Better Measure of Investment Value.
.
Change in EVA Margin Drives Change in MVA Margin
.
Business news is dominated with stories about large companies such as Alphabet Inc. (parent of Google), Apple Inc., and Amazon.com Inc. Not only are they popular, but large firms have been leading the stock market higher (the S&P 500 has outperformed the Russell 2000 since about 2011). If sustainable, maybe this outperformance is justified as the gap between the EVA Margin of large and smaller companies has widened (figure 4), driving up the MVA Margin spread between large and small companies to its widest level since at least 1996 (figure 5).
.
.
Winners gaining on the losers is not limited to large versus small. In aggregate, high-profitability companies have been improving at the expense of the low-EVA Margin firms across sectors, and the dispersion of the MVA multiples has grown (figures 6 and 7). The widening EVA Margin gap is mostly due to declining EVA Margin for the bottom-third of companies; although, the top third have also improved. This situation may be temporary or perhaps permanent due to one or more reasons.
.
- Over the last several years, many small and lagging companies have been able to keep afloat, where they may have died off before, because of ample liquidity. Low rates have driven investors up the risk spectrum, and they may be providing these firms with inexpensive and extra funding they need to survive. In late 2018, more than 35 percent of small cap companies were unprofitable, up about 10 percent points from historic averages.
- The large and/or more profitable companies may be becoming more monopolistic and could be driving the losers out of business. The large companies may have scale to invest in technology and spread the cost over more units, and the more profitable firms may have funds to invest in technology to increase their advantages over time.
.
.
Figure 8 shows that company turnover between EVA Margin fractiles (high (1), mid (2), and low (3)) is declining. Turnover is calculated as the number of new companies entering a fractile during a quarter plus the number of old firms exiting a fractile divided by total number of prior period companies. Lower turnover means reversion, which is driven by cyclical and competitive forces, appears to be declining. Turnover has fallen by almost 10 percent from year 2000 for the top-third firms in terms of EVA Margin. For the bottom-third of companies in terms of EVA Margin, a company only has approximately a 20-percent chance of moving out of the doldrums compared to about 35 percent in 2000. Since the turnover figures here include new and old companies entering and leaving, one could also state turnover as half of these amounts.
.
Turnover appears to decline just before or during recessions, such as in the early 2000s and in 2006-2007, and rises during recessions (notice the high turnover in 2008-2010). Thus, perhaps part of the recent decline in turnover is cyclical. Turnover may rise if high-EVA Margin companies overinvest leading up to recessions (profit growth tends to lead capital spending growth as shown in figure 10 of The Expectations Clock: A Model for Cycles and Sentiment). Then those companies may pay the penalty during a recession. Overinvestment is probably not limited to high-EVA Margin companies. However, turnover also rose for low-EVA Margin firms during the Great Financial Crisis, which means companies exited the bottom third as well. The high turnover could occur if the bottom third firms, based on their prior rankings, improved or stayed relatively steady as the mid- and top-third of the stocks deteriorated.
.
During recessions, turnover jumps more for the bottom two-thirds of companies in terms of EVA Margins (see figure 8) than those in the top third of firms (dark blue line in figure 8). Importantly, reversion among top-third firms is the lowest followed by the bottom third of companies. While performance may move higher and lower as companies improve and decline, performance does not change enough, at least over short periods, to move the overall level of performance (high or low). Thus, the overall level of performance is “sticky” and the glue has been getting stronger because turnover has been declining (for more on performance stickiness, see figure 4 of The Expectations Clock: A Model for Cycles and Sentiment).
.
High EVA Margin and Low MVA Margin Companies are Winners
.
Screening for quality, based on EVA Margin, produces alpha as does buying cheap companies, based on MVA Margin.
.
Figures 9-11 show 12-month return and risk metrics for stocks sorted based on EVA Margin and MVA Margin. Both cheap and the profitable companies significantly outperform expensive and low profitability companies, respectively. High-EVA Margin companies only have two-thirds the volatility of low-EVA Margin stocks and a much lower beta. While low-MVA Margin stocks have higher risk, they have average to above-average risk-adjusted return ratios.
.
.
Furthermore, excess returns grow with the holding period. Figures 12-13 show that the return spread of high margin over low-EVA Margin grows from 2-3 percent for 12-month holding periods to over 6 percent for 36-month periods. The returns also rise with the holding period for low-MVA Margin companies (versus high-MVA Margin). Outperformance is about 3 percent for one year and rises to around 9 percent for three years.
.
.
High-EVA Margin companies also outperform low-EVA Margin companies in 67 percent of the rolling one-year periods (68 percent for three years), and low-MVA Margin stocks better high-MVA Margin stocks during 61 percent of the 12-month periods (70 percent for three years).
.
Reversion provides the fundamental basis for low MVA Margin company outperformance. Competitive pressures force changes for losers (generally low-EVA Margin firms), which drives up their EVA Margin and therefore MVA multiples and performance. Or, for a cyclical sector, the higher-risk, lower-EVA Margin and lower-MVA Margin company with greater operating and financial leverage may fluctuate more with the economy – rising more during a recovery and declining more during a recession. Like tides in the ocean cause all ships to rise and fall, the economic cycle influences both strong and weak companies. However, the old fishing boat (low-EVA Margin / low-MVA Margin company) rises and falls more with the tide than the aircraft carrier (high-EVA Margin / high-MVA Margin firm).
.
High-EVA Margin companies may outperform if they have a moat and are able to maintain their margins (figure 8 shows that high EVA Margin companies have lower turnover) and can reinvest the high profits in new good ventures. By nature, these companies have high return on capital (at least versus their cost of capital), so they do not need to invest as much capital to generate profits. Cash flow may be high for these firms, which they can use to pay dividends, buy back shares, or invest for growth, each adding to returns.
.
Therefore, the results may be explained by both reversion for the cheap companies and moats which protect strong margin firms. However, the return spread for high- to low-EVA Margin companies is lower than it is for low- to high-MVA Margin companies, which means the forces of reversion for the weak is greater than staying power for the strong. It is also possible that poor companies are overlooked, so MVA Margin is too low for these low-profitability companies, and high-EVA Margin firms are popular and too expensive, and inefficiencies in pricing are eventually corrected.
.
Combining High EVA Margin with Low MVA Margin is Even Better
.
Forming portfolios with either high EVA Margin or low MVA Margin, quality or value, produced alpha over time (as shown above); however, alpha rises when the two screens are combined.
.
Panel A of figure 14 shows that high-EVA Margin / low-MVA Margin companies produced more than 13 percent annual returns since 1996, whereas low-EVA Margin / high-MVA Margin stocks were up by less than 7 percent. The spread is nearly 7 percent per year. However, the top companies – high margin and cheap – are uncommon. They only make up 3 percent of the dataset so it may be more difficult to have a large diversified portfolio of these stocks. The bottom firms are somewhat more common (9 percent).
.
The most northwest cell in the data of panel A in figure 14 depicts the best hunting ground for investors. This is likely especially true if the high EVA Margin is sustainable – one gets a high EVA Margin for a low MVA Margin price. However, look closely at the top three northwest combinations. The return for high-EVA Margin / low-MVA Margin companies is essentially the same return as for mid-EVA Margin / low-MVA Margin companies and high-EVA Margin / mid-MVA Margin firms. In some sectors, highly profitable companies are cheap for a reason and do not produce the highest returns. In these cases, the high EVA Margin firms may be under attack by competitors or could be riding high on the rising wave of the economy just before it crests. On these occasions, either or both the mid-EVA Margin / low-MVA Margin firms and high-EVA Margin / mid-MVA Margin companies perform best. As an example, in the energy minerals sector, the high-EVA Margin / low-MVA Margin stocks could be priced low because investors expect EVA Margin to deteriorate. They earned just 0.7 percent annual returns, but the mid-EVA Margin / low-MVA Margin firms earned 8.1 percent and the high-EVA Margin / mid-MVA Margin stocks were up by 7.3 percent.
.
The top left of panel B of figure 14 shows the combined return (13.3 percent) of top to mid profitability firms that are lowly valued to moderately priced, and the bottom right quadrant shows the returns (7.6 percent) of the low to mid profitability companies that are moderately to expensively valued. The return spread is 5.7 percent, which is still good compared to the 6.6 percent spread of going long the high-EVA Margin / low-MVA Margin stocks and short the low-EVA Margin / high-MVA Margin securities. Yet the northwest three cells represent 27 percent of the stocks (the three southeast cells include 23 percent of the stocks) versus only 3 percent for the most northwest cell, so the northwest three cells are a sufficiently large group of stocks to consider for many to most portfolios.
.
About 50 percent of the stocks fall closer to the EVA Margin / MVA Margin regression line (like in figures 1-3). Those along the line may or may not be appropriately priced and those on either side may not be incorrectly priced, but, on average, it appears they are. Otherwise, one would not see the approximately 6-percent spread in annual returns. The most northwest quadrant is generally an area of stocks where investors believe EVA Margin is at risk of a greater decline than materializes and/or have lower EVA growth than occurs. The bottom-right quadrant represents the opposite expectations. Investors are overestimating growth (for low to moderate margin companies) and too little risk, which is typical for high-growth / lower-profitability stocks. Also, mispricing seems to be more pronounced for small companies than large companies, probably because large stocks are more widely followed. The small cap return spread is 9.7 percent between the three northwest cells and the three in the southeast versus 3.1 percent for large.
.
Returns are Not Driven by Outliers
.
A good stock picking filter should not be influenced by outliers. More stocks that screen well should outperform than stocks that do not, the filter should work in most sectors, and it should perform well in most environments. If it does not meet one of these criteria, then the strategy is still decent if the periods and areas of underperformance are predictable. The EVA Margin and MVA Margin screens meet the good stock picking criteria.
.
Randomly Selecting High-EVA Margin and Low-MVA Margin Stocks Within Sectors Works
.
On average, only 46 percent of stocks outperform the FactSet universe (figure 15). That means there are more losers than winners. One can improve the odds of outperforming by buying the high-EVA Margin and low-MVA Margin stocks. High-EVA Margin companies have a 5.5 percent better chance of outperforming the universe than low-EVA Margin stocks, and the probability is greater than 50 percent (51.4 percent). The enhanced odds of success are a little bit lower for the MVA Margin screen, but there is still a positive probability spread of 3.3 percent for low- versus high-MVA Margin companies, and low-MVA Margin companies are still more likely (51.0 percent) than not to outperform the universe.
.
While the spreads and the probabilities are not very high, they are still an improvement over random stock picking across all stocks. An investor would likely outperform by throwing many “darts” at either the high EVA Margin or low MVA Margin subsets and owning a diversified portfolio.
.
It Works Across Almost Every Sector
.
The top-screening stocks outperform the bottom stocks in most sectors (figures 16-17). The only exception is miscellaneous which has very few companies (only 223 observations out of 148,008 since 1996). Top firms are those with high EVA Margin and low MVA Margin (1-1s), high EVA Margin and moderate MVA Margin (1-2s), and moderate EVA Margin and low MVA Margin (2-1s). Bottom firms include companies with low EVA Margin and high MVA Margin (3-3s), low EVA Margin and moderate MVA Margin (3-2s), and moderate EVA Margin and high MVA Margin (2-3s).
.
.
The screen works best in small caps, which may mean they are more mispriced than large stocks as large companies are more widely followed. Also, defensive sectors (communications, consumer non-durables, health services, and utilities) tend to perform better than cyclical sectors (includes everything but defensives), where one could receive a false cheap signal. A stock is cheap for a reason if profitability is cyclical and peaking. For instance, returning to the earlier example, sectors such as energy minerals probably have low valuations on peak profits. Despite the low valuation, they underperform when profits turn down. However, utilities, a defensive sector, also fails to deliver as well as other areas. Utilities are regulated, and if a firm’s EVA Margin is high this could mean its return on equity, which is set by regulators, is high and is due for an adjustment down. A low MVA Margin for utilities may accurately reflect this change.
.
Figures 18-19 show that performance within sectors is not being driven by – at least not too much – by irregular period specific outcomes. For every sector except miscellaneous, the top firms perform better than the bottom stocks the majority of the time.
.
It Works Most of the Time, and When It Doesn’t, the Performance is Predictable Over the Cycle
.
As noted above, the strategy of buying high-EVA Margin and low-MVA Margin companies works most of the time. Plus, when the strategy does not work, performance is predictable based on the economic and market cycle. High-EVA Margin companies outperform counter to the cycle, and low MVA Margin companies outperform with the cycle.
.
Factor efficacy is influenced by financial market and economic cycles, just like asset classes and sectors of the bond market (credit versus government) and the equity market (cyclicals versus defensives and small versus large). A good EVA Margin is equivalent to good “quality.” Good quality has a cost (i.e., higher MVA Margin). Quality tends to lag during risk-on markets and performs better during risk-off markets as people pay up (higher MVA Margin) for safety. The blue shaded area of figure 20 shows how a strategy of going long the top third of EVA Margin stocks (within FactSet sectors) and shorting the bottom third EVA Margin stocks performed over time. The long-short return is negatively correlated with overall stock returns (red line).
.
Conversely, sorting companies by MVA Margin produces the opposite results (figure 21). Low MVA companies have been highly correlated with overall stock market returns since the early 2000s. The exception was the internet bubble and burst when valuation as a screen performed terrible on the way up and then incredible on the way down as markets became more rational. Low-MVA Margin companies may be considered higher risk, so in rising risk-on markets they outperform.
.
EVA Margin and MVA Margin are positively correlated, so it makes sense that EVA and MVA Margin have negative relationships with overall stock returns. High-MVA Margin companies tend to be high-EVA Margin firms (see figures 1-5 and 7). Thus, when high-EVA Margin stocks outperform in low-return markets, one would expect high-MVA Margin stocks to also outperform since many of these firms may be the same company, and vice versa.
.
Value has been out of favor for several years (figure 21), whereas quality has been outperforming (figure 20). Given market returns and the historical relationships presented in these figures, this trend is exactly what should have been expected. As returns have come down and investors have become worried about the economy, they have been more willing to pay up for quality, and high-EVA Margin / high-MVA Margin stocks have outperformed (for more on the value and growth cycle, see When Do Value and Growth Outperform?). As shown in figure 7, the gap between high- and low-EVA Margin companies has widened, so part of the outperformance and increasing MVA Margin gap is justified, but this is only if we assume the widened EVA Margin spread is sustainable. Remember, turnover rose substantially during the last recession (figure 8).
.
Combining EVA Margin and MVA Margin Results in More Consistent Alpha
.
.
Both EVA Margin and MVA Margin are components in ISS-EVA’s PRVit model. Ignoring other variables, high-EVA Margin / low-MVA Margin companies are rated better than low-EVA Margin / high-MVA Margin firms. It’s a great deal to find a high performance (high-EVA Margin) stock for a good price (low-MVA Margin). While profits could decline to justify the lower valuation, they also may not. Conversely, expensive (high-MVA Margin) stocks with low-EVA Margin are high-expectations stocks; investors are betting on and already paying for improving fundamentals which may or may not materialize.
.
Combining a screen for EVA Margin and MVA Margin produces more consistent alpha, as shown in figure 22. For much of the period since 1996, quality, reasonably valued stocks (top third stocks) have produced relatively consistent positive alpha versus the expensive, high expectation securities (bottom third stocks). One exception was the Internet bubble (1998-9), when many good companies were mispriced with low valuations in favor of placing high values on businesses that would ultimately fail. Other periods of more minor negative alpha are consistent with the factor cycle.
.
The minor setbacks included the peak to the beginning of the burst of financial bubble and high returns markets starting in mid-2002 and late-2009. In mid-2002 through 2003, economic conditions, as gauged by the ISM PMI (Purchasing Managers Index), were improving off a weak economy (figure 23), and this environment favors lower-quality value. In the period beginning in late 2009, the economy was strong and slightly accelerated (the PMI was high and modestly rising for the majority of the period), which typically favors growth/momentum.
.
Trailing 12-month returns of the long-short top third-bottom third trade has been negative since September 2018 (figure 22), which may mean that markets are getting lofty. Today’s market environment is somewhat reminiscent of the Internet bubble; there is a flood of liquidity to high-flying money losing IPOs, growth in private equity and venture capital, and a rise in the percentage of money-losing small companies. In addition, corporate leverage levels are high. While negative alpha is never fun, note that alpha of the long-short trade was quite high following bubble periods. Moreover, investors should probably consider factors beyond EVA Margin and MVA Margin, such as EVA growth, risk considerations, and other valuation and fundamental variables (ISS EVA’s PRVit model includes many variables).
.
Finally, in figure 24, notice how a long-short strategy with EVA Margin and a long-short with MVA Margin produced about the same returns over time (up 2-3X since 1996); however, combining the two is much more powerful (up 8X since 1996). Over the entire period, the 1-1s (high-EVA Margin / low-MVA Margin) were up 39X, whereas the 3-3s (low-EVA Margin / high-MVA Margin) were only up 2X (figure 25). The long-short value strategy faltered recently (the aqua return line is declining), whereas quality has been in style (the grey line is rising). The combination of the two (the dark blue line) is somewhat flat over the last two years as markets rotated first to momentum and then to quality (but not low priced quality), as it typically does late in the cycle on the way up (2017) and then during the initial slow down (since mid-2018).
.
