A critical evaluation of risk return characteristics of environmentally focused stocks companies

501 ACTA UNIVERSITATIS AGRICULTURAE ET SILVICULTURAE MENDELIANAE BRUNENSIS Volume LXI 57 Number 2, 2013 http://dx.doi.org/10.11118/actaun201361020501 A CRITICAL EVALUATION OF RISK-RETURN CHARACTERISTICS OF ENVIRONMENTALLY FOCUSED STOCK’S COMPANIES Stanislav Škapa, Tomáš Meluzín, Marek Zinecker Received: February 15, 2013 Abstract ŠKAPA STANISLAV, MELUZÍN TOMÁŠ, ZINECKER MAREK: A critical evaluation of risk-return characteristics of environmentally focused stock’s companies. Acta Universitatis Agriculturae et Silviculturae Mendelianae Brunensis, 2013, LXI, No. 2, pp. 501–506 The objective of the paper is to critically evaluate and determine risk-return profi le environmentally focused stock’s companies which are covered by STOXX Global ESG Environmental Leaders Index and whether this index should be taken in as an independent asset class of investments portfolio for its risk-return improvement. This paper gives an empirical view on the ex-post asset classes characteristics focused mainly on risk side of investment. environment, investment, stock’s companies, risk-return An environmental investing is a subset of Socially Responsible Investing (SRI) that concentrates on investing in companies or technologies that have a positive eff ect on the environment, such as businesses making an eff ort to reduce their greenhouse emissions, or those looking for renewable energy sources. For example, clean technology or “cleantech” companies are those that work to improve operations, performance, productivity and effi ciency while lowering their energy consumption, inputs, waste or pollution. Investors are placing money into environmentally oriented companies for a combination of two reasons: to fulfi ll perceived ethical responsibilities by helping to improve the environment, and to ride the next big wave in the stock market. But whether you are an individual or institutional investor, the key to making a profi t is to choose the right instruments in which to invest. The growth of environmentally oriented investment funds, exchange traded funds (ETFs) and certifi cates is on the rise. Investors who are more risk-averse should look to invest in one of these types of funds, which typically provide a good amount of diversifi cation. Environmentally mutual funds invest in a variety of companies, and some will invest in various green industries, such as solar, wind, bio-fuels, or recycling. ETFs track indices and also off er instant diversifi cation. Finally, there are many funds that invest in international green companies, thereby providing global diversifi cation. Those investors willing and able to take on more risk can go for a “pure play” strategy – investing directly in environmentally oriented companies. The defi nition of an environmentally oriented company is a bit subjective, but there are organisations which make a same classifi cation for environmentally oriented companies. One of them is Sustainalytics which supports investors with the development and implementation of responsible investment strategies. Sustainalytics together with STOXX Ltd. off er STOXX® Global ESG Environment Leaders index. The sustainability data is provided by Sustainalytics and the index model is developed by STOXX Ltd.. The indices follow a bottom-up approach and are based on company sustainability ratings. They are based on a very fl exible concept, which allows issuers to combine the indices according to their own requirements. This paper analyses the risk profi le and some other important characteristics (e.g. return, Sharpe ratio, Sortino ration) of STOXX Global ESG Environmental Leaders equity index compared to some well known equity indexes: 502 Stanislav Škapa, Tomáš Meluzín, Marek Zinecker • STOXX Global Select Dividend 100 Net Return is the composite equity index covering countries in the developed markets. Companies are selected on the basis of dividend criteria and the weighting of the companies in the index is purely based on dividends. • MSCI World Net Return Index is the composite equity index covering countries in the developed markets. The index is capitalizations weighted. • CECE Net Return Index is the composite equity index comprising the Czech Republic, Hungary, and Poland It is a capitalization-weighted index consisting of the Czech, Hungarian, and Polish blue chip stocks, which are members of the respective country index: CTX Czech Traded Index, HTX Hungarian Traded Index, and PTX Polish Traded Index. The index is calculated and disseminated by Wiener Börse. • iBOXX € SOVEREIGNS EUROZONE 3–5® represents the 3–5 years maturity Eurozone currency sovereign debt issued by Eurozone governments. The Index covers the Sovereigns Eurozone 3–5 years maturity bucket. Within the Index, each bond is weighted according to its amount outstanding. The Index is calculated and disseminated by Deutsche Börse and sponsored by International Index Company Limited (“IIC”). • DBLC Index™ – Optimum Yield balanced (the “Index”) is intended to refl ect the performance of 14 commodities: WTI Crude Oil, Brent Crude Oil, Heating Oil, RBOB Gasoline, Natural Gas, Aluminium, Copper, Zinc, Gold, Silver, Wheat, Corn, Soybean and Sugar. Recomposition is done using the Optimum Yield methodology that seeks to optimise the roll yield that will maximise the positive roll yield or minimise the negative roll yield. The index is currency hedged on a monthly basis. The index level is hedged by means of a forward contract with a maturity of one month. Index changes during the month are not hedged. All returns are quarterly total returns (with dividends) expressed in Euros. The data are available from 30th September 2001 until 30th September 2012 (44 values per index). This time is not a long period due to the short time series of STOXX ESG Environmental Leaders index. Measurement of risk During historic period, problems of objectifi cations of risk measurement by using a concept of probability and applying statistical analysis were discussed. In 1952, two authors published the ultimate papers for fi nancial industry; the fi rst was H. Markowitz (1952) who identifi ed risk as related to the varying fi nancial outcomes and adopted the standard deviation of the residual assets as the tool for measurement of risks. He also provided a quantitative framework for measuring the portfolio risk. The second one was A. Roy (1952) who introduced the “Safety First” criterion, which meant introduction of a downside risk measurement principle. A few years later, Markowitz (1959) gave a generalized discussion on the risk and introduced alternative measurement tools as semivariance, expected value of loss, expected absolute deviation, probability of loss and the maximum loss. Markowitz also introduced his idea of downsiderisk and suggested two types of measurement of a downside risk: • A semi-variance computed from the mean return or below-mean semi-variance (SVm); • A semi-variance computed from a target return or below-target semi-variance (SVt). Both measures compute a variance using only the returns below the mean return (SVm) or below a target return (SVt). Markowitz called these measures partial or semivariances, because only a subset of the return distribution is used, see (Nawrocki, 1999).  2 1 1max 0, K T T SVm E R K      2 1 1max 0, K T T SVt t R K    , where RT is an asset return during the time period T, K is the number of observations, t is the target rate of return and E is an expected mean return of the asset return. A maximizing function denoted as max, indicates that the formula will square the larger of the two values i.e. 0 and (ERT) or (tRT). A er proposing the semi-variance measure, the classical author stayed with the variance measure because it was computationally simpler. The semi-variance optimization models using a cosemivariance matrix (or semi-covariance if that is your preference) require twice the number of data inputs than the variance model. With the lack of cost-eff ective computer power and the fact that the variance model was already mathematically very complex in these times as it belonged to the class of quadratic programs, this was a dominant consideration in practical applications until the 1980s. It was used with the advent of the microcomputer (Nawrocki, 1999). Markowitz (1991) and this approach was also further developed in order to defi ne a measure of downside risk. According to fi ndings by Kahneman and Tversky’s (1979) loss aversion preferences imply that investors who dislike downside losses will demand greater compensation, in the form of higher expected returns, for holding shares with high downside risk. Sortino and Van der Meer (1991) note that standard deviation has one major drawback. Standard deviations measure uncertainty or variability of returns but in some cases this does not match one’s intuition about risks. Large positive outcomes are treated as equally risky as large negative ones. In practice, however, positive outliers should be A critical evaluation of risk-return characteristics of environmentally focused stock’s companies 503 regarded as a bonus and not as a risk. It is therefore better to look at some measure of downside risk. MATERIALS AND METHODS 44 quarterly data per each index were obtained. It is a relatively small sample to make some strong conclusions. Due to this fact, some parametrical tests were not found suitable. Therefore were used some robust statistical methods and bootstrap method, too. It means that robust statistical methods aim to construct statistical procedures that are stable (robust) even when the underlying model is not perfectly satisfi ed by the available data set. Robust models focus on the statistical properties of the bulk of the data without being distracted by outliers, while in classical models all data equally participate in the analysis. Classical estimators that assign equal importance to all available data are highly sensitive to outliers. Therefore, in the presence of just a few extreme losses, classical analysis can produce arbitrarily large estimates of mean, variance, and other statistics. Bassett et al. (2004) investigate the performance of portfolio return distribution using robust and quantile-based methods, and conclude that the resulting forecasts outperform those under a conventional classical analysis. Perret-Gentil and Victoria-Feser (2005) used robust estimates for mean and the covariance matrix in the meanvariance portfolio selection problem. They showed that the robust portfolio outperforms the classical one, as the outlying observations (that account for 12.5% of the data set) can have serious infl uence on portfolio selection under the classical approach. The same purposes, i.e. the presence of skewed distributions and extreme values, led to useing the interquartile range (by practitioner’s hint for a normal distribution is approximately equal to 1,35*standard deviation). The bootstrap method was used with only on a small number of data. The bootstrap method was originally proposed by Efron (1979) and it is a computationally-intensive method for estimating the distribution. To use the bootstrap or any other statistical methodology eff ectively, one has to be aware of its limitations. The bootstrap is of a value in any situation in which the sample can serve as a surrogate for the population. If the sample is not representative of the population because the sample is too small, biased, or not selected in a random way, or its constituents are not independent, then the bootstrap based techniques fail. Canty et al. (2000) also list data outliers, inconsistency of the bootstrap method, incorrect resampling of a model, wrong or inappropriate choice of statistics, non-pivotal test statistics, nonlinearity of the test statistics, and discreteness of the resample statistic as potential sources of error. The pitfall of using the bootstrap method also shows the Terpstra and McKean (2005) and Salibian-Barrera M., and Zamar R. H. (2002). Therefore for an index analysis there were made 10 times 2 500 bootstrap samples per index that means that each index statistics and characteristic was estimated 10 times. For a “fi nal” enumerating of statistics and characteristics was used trimean. RESULTS AND DISCUSSION Firstly, it was processed an explanatory data analysis of all indices (quarterly data), the results are shown in Tab. I. According to the descriptive data analysis one could say that medians are greater than means except bonds index. STOXX ESG index has higher average return, medians and standard deviations than MSCI World index. STOXX ESG index characteristics are close to STOXX Global Dividend except risk characteristics. STOXX ESG index´s minimum value (loss) is highest of all equity indexes. In another way it means that there is a probability of small loss comparing to another equity indexes. In addition, kurtosis statistics show that the majority of indexes distributions have fatter tails than normally distributed variables. For deeper data analyses the related Box and Whiskers plots were made and results are shown in Fig 1. Result of this analyze corresponds with explanatory data fi nding, ESG index has no extreme values (outliners) but 1st quartile is lowest of all equity indexes. For complex reward/risk comparing of all indexes were calculated annualized summary statistics. I: Quarterly summary statistics of equity indices CECE DBLCI ESG iBOXX MSCI STOXX Count 44 44 44 44 44 44 Average 4,2 3,6 2,1 1,1 0,9 2,1 Medians 4,9 6,0 5,1 1,0 3,1 4,0 Standard deviation 14,9 10,1 9,6 1,6 8,5 9,1 Minimum −38,3 −30,7 −20,3 −2,4 −20,2 −28,7 Maximum 36,2 18,2 18,6 4,9 13,7 22 Range 74,8 48,9 38,9 7,4 33,9 50,7 Stnd. Skewness −1,2 −3,9 −1,9 −0,1 −2,5 −2,4 Stnd. Kurtosis 1,05 3,9 −0,4 −0,2 0,6 3,2 Source: Own calculation 504 Stanislav Škapa, Tomáš Meluzín, Marek Zinecker Annualized returns were enumerated as quarterly return multiply 4-times. Annualized standard deviations were enumerated as quarterly standard deviations multiply 2-times. For complex reward/ risk comparing was used Sharpe ratio (see Tab. II). According to Sharpe ratio has STOXX ESG index better reward/risk profi le than MSCI World index, this profi le is nearly the same as STOXX Global Dividend index. Kurtosis statistics (according summary statistics) show that the majority of indexes distributions have fatter tails than normally distributed variables and concurrently Box and Whiskers plots indicates some extreme values for majority of equity indexes therefore a bootstrap techniques was used for enumerating some selected statistical characteristics wit impact on downside risk (losses). For enumerating of downside risk was used downside deviation, where target rate of return was set to 5% p.a. The bootstrapped results are shown in Tab. III. According bootstrapped results is STOXX ESG index better in complex reward/ downside risk indicator comparing to MSC World (nearly the same standard deviation and a bit smaller downside deviation, but more than doubled expected return). The Sortino ratio is positive for STOXX ESG index opposite to MSCI World Index. A comparison of the STOXX ESG index with the STOXX Global Dividend index is not so favorable for STOXX ESG index, all characteristic are better for STOXX Global Dividends index. For portfolio investors is very important how strongly are indexes correlated each to another therefore correlations were enumerated, due the fact some of equity indices should be a better complement to diversifi ed portfolio then others (see Tab. IV), for example if an investor´s portfolio includes commodities the best equity index is STOXX Global Dividend, not STOXX ESG index. CECE DBLCI ESG iBOXX MSCI SOXX Box-and-Whisker Plot -39 -19 1 21 41 response sample 1: Box and Whiskers plot Source: Own calculation II: Annualized summary statistics of equity indices CECE DBLCI ESG iBOXX MSCI STOXX Expected Return 16.8 14.4 8.4 4.4 3.6 8.4 Standard Deviation 29.8 20.2 19.2 3.2 17 18.2 Sharpe ratio+0.48 0.59 0.31 0.59 0.06 0.32 +Risk free rate is set to 2.5% Source: Own calculation III: Estimated annualized summary statistics of all indices CECE DBLCI ESG iBOXX MSCI STOXX Expected Return 15.3 14.3 7.4 4.3 2.9 8.1 Median 16.3 16.6 8.5 4.3 4.1 9.2 Standard Deviation 28.7 19.1 18.6 3.2 16.4 17.6 Downside Deviation*12.2 7.1 9.3 0.36 10.1 8.4 5th percentile −32.6 −20.1 −24.2 −0.8 −25.3 −22.4 Sortino ratio* 0.84 1.31 0.26 −1.94 −0.21 0.37 *Target return is se to 5% Source: Own calculation A critical evaluation of risk-return characteristics of environmentally focused stock’s companies 505 CONCLUSIONS According obtained result in the process of data analyzing of indexes there were found the following facts. Firstly, STOXX ESG Environmental Leaders index has nearly the same Sharpe ratio as STOXX Global Select Dividend 100 and the statistical characteristics are very similar to STOXX Global Select Dividend 100. Next, among four equity indexes has STOXX ESG Environmental Leaders index the highest median value. Finally a correlation to commodities is relatively high comparing the rest of equity indexes. This index should not be taken in as an independent asset class of investments portfolio for its risk-return improvement. IV: Correlations of all indices CECE DBLCI ESG iBOXX MSCI STOXX CECE 0.3233 (44) 0.4408 (44) 0.1820 (44) −0.0690 (44) 0.2510 (44) DBLCI 0.3233 (44) 0.2188 (44) −0.4107 (44) −0.2871 (44) −0.0692 (44) ESG 0.4408 (44) 0.2188 (44) 0.0576 (44) 0.7026 (44) 0.0795 (44) iBOXX 0.1820 (44) −0.4107 (44) 0.0576 (44) −0.3570 (44) 0.3846 (44) MSCI −0.0690 (44) −0.2871 (44) 0.7026 (44) −0.3570 (44) 0.3846 (44) STOXX 0.2510 (44) −0.0692 (44) 0.0795 (44) 0.1544 (44) 0.3846 (44) Correlation (Sample Size) Source: Own calculation SUMMARY This paper analyses the risk profi le and some other important characteristics (e.g. return, Sharpe ratio, Sortino ration) of STOXX Global ESG Environmental Leaders equity index compared to some well known equity indexes (MSCI World Net Return Index, CECE Net Return Index, iBOXX € SOVEREIGNS EUROZONE 3–5® and DBLC Index™ – Optimum Yield balanced). 44 quarterly data per each index were obtained. Apart from some parametrical tests we used some robust statistical methods and bootstrap method. There were made 10 times 2500 bootstrap samples per index. For an index analysis that means that each index statistics and characteristic was estimated 10 times. For a “fi nal” enumerating of statistics and characteristics was used trimean. According to obtained result in the process of data analysing of indexes there were found the following facts. Firstly, STOXX ESG Environmental Leaders index has nearly the same Sharpe ratio as STOXX Global Select Dividend 100 and the statistical characteristics are very similar to STOXX Global Select Dividend 100. Next, among four equity indexes has STOXX ESG Environmental Leaders index the highest median value. Finally, among all equity indices has STOXX ESG the lowest correlation to bond’s index but a correlation to commodities is relatively high comparing the rest of equity indexes. This index should not be taken in as an independent asset class of investments portfolio for its risk-return improvement. Acknowledgement This paper is supported by the Czech Science Foundation. Name of the Project: IPO Strategy – Specifi c Approaches in the CEE Region. Registration No. 13-38047S. REFERENCES BASSETT, G., GERBER, G., ROCCO, P., 2004: Robust strategies for quantitative investment management. Hubert et al., 27–37. ISBN 3-76437060-2. CANTY, A. J. & DAVISON, A. C. & HINKLEY, D. V. & VENTURA, V., 2006: Bootstrap diagnostics and remedies. Canadian Journal of Statistics, 34: 5–27. ISSN 1708-945X. EFRON, B., 1979: Bootstrap methods: Another look at jackknife. Annals of Statistics, 7, 2: 1–26, ISSN 0090-5364. KAHN, A. E., 1970: The Economics of Regulation: Principles and Institutions. New York: Wiley, 559 s. ISBN 0262610523. KAHNEMAN, D., TVERSKY, A., 1979: Prospect Theory: An analysis of Decision Under Risk. Econometrica, 47, 3: 263–291. ISSN 1468-0262. 506 Stanislav Škapa, Tomáš Meluzín, Marek Zinecker KAYSEN, C., TURNER, D., 1959: Antitrust Policy: An Economic and Legal Analysis. Cambridge: Harvard University Press. KIDOKORO, Y., 1998: Rate of Return Regulation and Rate Base Valuation. Regional Science and Urban Economics, 28, 9: 629–654, ISSN 0166-0462. MARKOWITZ, H. M., 1952: Portfolio Selection. Journal of Finance, 7, 1: 77–91, ISSN 1540-6261. MARKOWITZ, H. M., 1959: Portfolio Selection. New York: John Wiley and Sons, 341 s. ISBN 978-0300-01372-6. MARKOWITZ, H. M., 1991: Portfolio Selection. 2nd ed. Cambridge, MA: Basil Blackwell, ISBN 978-155786-108-5. NAWROCKI, D., 1999: A Brief History of Downside Risk Measures. Journal of Investing, 8, 3: 9–25. ISSN 1068-0896. PERRET-GENTIL, C., VICTORIA-FESER, M., 2005: Robust mean-variance portfolio selection. FAME – International Center for Financial Asset Management and Engineering. 43 s. Avaiable at: http://www.swissfi nanceinstitute.ch/rp140.pdf. PINDYCK, R., RUBINFELD, D., 2001: Microeconomics. 5th ed. Upper Saddle River, N. J.: Prentice-Hall. ISBN 0130165832. POSNER, R. A., 1969: Natural Monopoly and Regulation. Stanford Law Review, 21, 6: 548–643, ISSN 0038-9765. ROY, A., D., 1952: Safety First and the Holding of Assets. Econometrica, 20, 3: 31–449, ISSN 14680262. SALIBIAN-BARRERA, M., ZAMAR, R. H., 2002: Bootstrapping robust estimates of regression. Annals of Statistics, 30: 556–582. ISSN 0090-5364. SORTINO, F., VAN DER MEER, R. 1991: Downside risk, capturing what’s at stake in investing situations. Journal of Portfolio Management, 17, 4: 27– 31. ISSN 0095-4918. SHARKEY, W. W., 1982: The Theory of Natural Monopoly. Cambridge: Cambridge, University Press. TERPSTRA, J. T., MCKEAN, J. W., 2005: Rankbased analyses of linear models using. R. Journal of Statistical So ware, 14, 7. ISBN 1548-7660. Guide to the LPX Equity Indices. Version 2.7. Available at: http://www.lpx-group.com/nmx/fi leadmin/ News/pdf_2007_E/NMX_Guide_to_the_Equity_ Indices.pdf. Address doc. Ing. et Ing. Stanislav Škapa, doc. Ing. Tomáš Meluzín, Ph.D., Department of Economy, doc. Ing. Marek Zinecker, Ph.D., Department of Finance, Faculty of Business and Management, Brno University of Technology, Kolejní 4, 612 00 Brno, Czech Republic, e-mail: [email protected].cz, [email protected].cz, [email protected].cz