Literature Review of Stock Market Anomalies
Disclaimer: This work has been submitted by a student. This is not an example of the work written by our professional academic writers. You can view samples of our professional work here.
Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of UK Essays.
Published: Wed, 14 Mar 2018
The history of fundamental analysis as a trading mechanism began with Benjamin Graham in 1928. Graham published his first book, Security Analysis in 1934. This book defined the framework of Value Investment and is now in its fifth edition.
Since that time, a great deal of research focused on specific fundamental measures as key determinants of a securities future price. The concept of efficiency is central to finance. Mainly, in an efficient financial market an asset price should be the best possible estimate of its economic value.
A financial market is efficient when widely available information to participant is reflected in stock prices.
1.3 FORMS OF EFFICIENT MARKET
There are three forms of market efficiency; the weak form, semi-strong form and strong form.
Weak Form: Stock prices reflect all of the information contained in past stock prices. The inference drawn from this is that you cannot consistently profit by spotting trends and patterns in stock prices.
Semi-strong Form: Stock prices all of publicly available information (including company accounts, industry data e.t.c). This simply implies you cannot consistently profit by analysing company accounts and other public data.
Strong Form: Stock prices reflect all information (including information known only to company insiders. The insinuation is that no one can earn excess returns by stock picking, even people with inside information.
However, this dissertation will concentrate more on using fundamental analysis to disprove the efficient market hypothesis which asserts that stock markets impounds all publicly available information about a company into stock prices in an instantaneous and unbiased manner. The implication of this assertion is that publicly available information such as financial statement figures cannot be used to detect mispriced securities; any investment strategy designed on the basis of published financial information should not prove profitable. However, in contrast to this argument, fundamental analysts believe that the markets may misprice securities and that it is possible to make an informed financial projections using financial statement information to earn an abnormal returns.
According to Fama (1970) he established the efficient market model after it has been tested and found supported by different wider markets and this became widely in use by financial communities, applied economist and financial economist. The weak forms of the efficient market hypothesis is strongly supported by evidence and the result follow and consistent with random walk model, while the strong form test is strongly supported by the efficient market hypothesis Fama, Fisher, Jensen and Roll (1969) found that the real time information such as the time of stock split, the future information on future dividend is fully reflected in this price
Grossman and Stigliz 1980 argues that if the market efficiency theory hypothesis govern the way price respond to information ,then Institutional investors and security analyst who engage in equity research will not arrive at any meaningful conclusion and the whole research will be a complete waste of time and resources. To mention but a few other argument includes the market anomalies: such as the post earning drifts, value versus growth and small cap stocks anomalies.
According to Ball and Brown (1968), even after earnings are announced, cumulative abnormal returns for stocks with unexpected positive (negative) earnings surprise continue to drift upwards (downwards). This anomaly is known as Post Earnings
Announcement Drift (PEAD). Foster, Olson and Shevlin (1984) noted that sixty days prior to an earnings announcement, combining a long position in stocks with unexpected earnings in the highest decile with a short position in stocks in the lowest decile generates an abnormal return of 25% p.a. before accounting for transaction cost, which confirms the existence of the PEAD anomaly.
Many other researchers who have argued in favour of fundamentals analysis includes: Basu (1977) studied the relationship between price-earning (P/E) ratios and excess returns, and was the first to uncover evidence that appeared to oppose the efficient market hypothesis (EMH). Basu concluded that there was an information content present in publically available P/E ratios, and portfolios built from low P/E stocks earned excess returns even after adjusting for risk. Banz in 1981,did analysis on monthly returns over the period 1931- 1975 on listed shares on the New York stock exchange. Over this interval, he observed that the fifty smallest outperformed the largest by an average of one percentage point per month. He observed a size effect, and concluded that there was a relationship between market capitalization of a firm, and its returns, even after adjusting for risk. Also In 1981, Reinganum confirmed that data on firm size could be used to create portfolios that earn excess returns.
Further fundamental anomalies were discovered, such as the book-to-market effect described by Rosenberg et al. (1984), which found that stocks with a high book-to-market value yielded higher long-term returns. Fama and French (1992) surveyed the above styles of anomaly detection, and concluded that if asset pricing is rational, then size and ratio of book to market value must be proxies for risk.
Lakonishok et al in 1994 evidence in the US by considering a time horizon up to five years found a wide range of value strategies (based on sales growth, book-to-market, cash flow, earnings, etc) all produced higher returns, and refuted Fama and French’s claims that these value strategies are fundamentally riskier. In 1995, Fama and French responded to Lakonishok by stating that size and book-to-market equity are proxies for sensitivity to risk factors in returns. Their results also suggest that there is a size factor in fundamentals that might lead to a size related factor in returns. Later, Fama and French (1998) studied returns on market, value and growth portfolios for the US and 12 major EAFE countries (Europe, Australia and the Far East). They found that value stocks tend to have higher returns than growth stocks, and conclude that these returns are explained by a one-state variable Intertemporal Capital and Asset Pricing Model (ICAPM) (or a two-factor APT), that explains returns with the global market return and a risk factor for relative distress.
Frankel and Lee also in 1998 estimate firms fundamental values (V), using I/B/E/S consensus forecasts, and a residual income model. They find V is highly correlated with stock price, and that V/P is a good predictor of long term returns. Piotroski focused on high book to market securities, and shows that the mean return earned by a high book-to-market investor can achieve a return of at least 7.5% annually. Piotroski 2000 also studied a number of different fundamental ratios and criteria with similar outcomes, and notes that returns are concentrated in small and medium size companies, companies with low share turnover, and firms with low analyst following.
In 2001 Aby et al. focus on using fundamentals to screen stocks for value. Also in 2001 Aby et al. concentrated on four fundamental conditions, namely, single valued P/E’s; Market price less than Book Value; established track record of return (established by ROE), and dividend payout ratio. The authors conclude that when the four criteria are used to screen stocks, quality investments seem to result. It is interesting to note that in earlier work, the authors had simply focused on shares with low P/E and a market price below book value, and concluded this filtering method did not produce satisfactory returns.
Fundamental Analysis provides a framework for modelling the financial technicalities of a company. Primarily, it aids in the development of company or industry specific models, and provides a means of evaluating the performance of a given company in terms of those models. A considerable contribution of the fundamental models is that they provide for the calculation of a number of financial ratios. These ratios are then used to assess the financial health of a company, and to compare directly to the ratios for different companies for alternative investment decision.
There is a long established tradition of attempting to use these fundamental ratios as predictors of a companies future share price. Primarily, this started with the work of Benjamin Graham in 1928, and forms the heart of an investment philosophy known as ‘Value Investing’. Value investors believe that the market does not price securities accurately, and that the true price of a security, its ‘intrinsic’ value, only rarely coincides with the market price. The trading manner of value investors is to determine the intrinsic value of a security, and acquire the security as long as the intrinsic value is above the price the market will sell at but according to the conventional definition, a value investor is one who invest in low price book value or low price-earnings ratios stocks while the generic definition defines it as one who pays a price which is less than the value of the assets in place of a firm. For example firms with high E/P, D/P or B/P ratios earns positive capital asset pricing model (CAPM) – adjusted abnormal returns.
However, on the contrast growth investor focuses on the potential of a company with much less emphasis on its present price. They invest on company that are trading higher than their current intrinsic worth in anticipation that the company intrinsic worth will grow and exceed the current valuation
One important issue, the subject of this dissertation, is to apply fundamental analysis to two UK listed insurance companies namely : Beazley Group Plc and Hardy Underwriting as a case study to investigate how detailed the financial statement information can be used to implement an active investment strategy with the hope of making abnormal returns. The remainder of the dissertation proceeds as follows: Section II provides a brief review of the relevant literature on various market anomalies such as post earning drift, value and growth anomalies and small cap anomalies. Section III describes the methodology by explaining relevant descriptive financial ratios and valuation dynamics procedure. The analysis of the result is explained in Section IV. The last section offers the conclusion.
2.0 Literature Review and Theoretical Frame Work
The semi-strong form efficient market hypothesis developed by Fama in 1970 assumes that security prices respond quickly to new information such as earnings announcement. But research has observed an apparent delay in share price response to earnings information (Richard, 2004). According to Ball and Brown (1968), even after earnings are announced, cumulative abnormal returns for stocks with unexpected positive (negative) earnings surprise continue to drift upwards (downwards). This anomaly is known as Post Earnings Announcement Drift (PEAD). Foster, Olson and Shevlin (1984) noted that sixty days prior to an earnings announcement, combining a long position in stocks with unexpected earnings in the highest decile with a short position in stocks in the lowest decile generates an abnormal return of 25% p.a. before accounting for transaction cost, which confirms the existence of the PEAD anomaly.
Various competing hypotheses try to explain the PEAD phenomenon. One school of thought attributes the phenomenon to ‘delayed response’ due to investors’ failure to appreciate the full implication of earnings announcement and high transaction costs that restrain immediate response to earnings news. Another school of thought suggests that there is misspecification in the Capital Asset Pricing Model (CAPM) because the model fails to adjust returns for risk and thus, the abnormal return is a compensation for bearing risk not captured in the CAPM (Ball et al 1988 in Bernard and Thomas, 1989).
2.1.0 Evidence of Post Earnings Announcement Drift
The most widely accepted study of PEAD is that of Bernard and Thomas (1989). They empirically test competing explanations of PEAD in the US using a sample of quarterly earnings announcements for firms listed on NYSE/AMEX, covering the period from 1974 to 1986. The sampling criteria and the methodology are consistent with the models used by Foster et al. (1984) which are based on standardised unexpected earnings (SUE) i.e. unexpected earnings scaled by their standard deviation. The SUE model allows for a better comparison among firms because it accounts for earnings announcements at different points in time. Additionally, this model prevents firms with more volatile earnings from unduly biasing the result. Firms are ranked in deciles according to SUE compared to their distribution in the previous quarter and assigned to ten portfolios where the tenth portfolio consists of firms with the highest ranking of SUE. The results of Bernard and Thomas (1989) show that PEAD exists for all stocks and it increases monotonically from the lowest to the highest SUE decile. The effect of PEAD is more significant for smaller firms.
Finally, most of the drift appears within the first sixty days, almost all of the effect happens within nine/six months for small/large firms and the largest part occur during the first five days. Bernard and Thomas reject the hypothesis of CAPM misspecification but find evidence consistent to delayed price response as an explanation for PEAD. They conclude that the 12 existence of PEAD is due to investors’ failure to realise the serial correlation of quarterly earnings. Ball and Bartov (1996) however suggest that investors understand this implication but underestimate the degree of serial correlation between current and future earnings by 50%. However, the UK evidence Liu, Strong and Xu (2003) are consistent with that of Bernard and Thomas (1989) their result shows evidence of PEAD in the UK. The PEAD portfolio (highest decile minus lowest decile) yields significant profit of 2.9%, 5.2%, 8.2%, and 10.8% over 3-, 6-, 9-, and 12 months investment horizon respectively. After controlling for risk and market microstructure effect, the price based earnings surprise measure gives the strongest drift effect. Other evidences by Forner et al. (2009) in the Spanish stock market, Truong (2010) in New Zealand, Booth et al (1996) in Finland. The overall results justify the presence of PEAD in their respective stock markets.
2.1.1 Price Momentum and Post-Earnings Announcement Drift
Price momentum is the predictability of future returns from past returns. High (low) stock prices continue to drift up (down) in the same direction so that on average, past winners continue to outperform past losers over a three to twelve month period (Jegadeesh and Titman, 1993). Chan et al (1996) attributes the existence of price momentum to market’s under reaction to earnings news. They analyse the returns of different earnings momentum strategies which differ based on how earnings surprises are measured. They find that earnings momentum strategy that buys stocks in the top decile over the previous six months and sell stocks in the bottom decile over the previous six months generates the highest abnormal return of 8.8% when held for six months. Chan et al (1996) conclude that drifts in future returns for the next six to twelve months are predictable from previous news about earnings
and a stock’s returns.
Since PEAD is observed in several countries, it is evident that investors do not respond efficiently to public information like earnings announcements, thereby questioning the efficient market hypothesis. A delayed response to information is accepted as an explanation by a few scholars, while models misspecification is less convincing due to insufficient evidence. A new area of research into a possible explanation of PEAD is the ‘Inflation Illusion Hypothesis’ where investors fail to consider the inflation effect while predicting future earnings growth rate. This may partly explain the delay in stock price response ‘to information as visible and freely available as publicly announced earnings’ (Chordia and Shivakumar, 2005).
2.2.0 Investment Performance of common stocks in relation to their Price –
In an efficient capital market, security prices rapidly respond to available information in an impartial way and thus provide an unbiased estimate of the underlying values. Although there is significant empirical evidence supporting the efficient market hypothesis but many still question its validity. As part of the anomalies of the efficient market hypothesis is that price earnings ratios (P/E) are indicators of future investment performance of a security. The price earnings ratio (P/E) hypothesis proposes that low P/E securities will tend to outperform high P/E stocks. Although, security prices reflect some degree of biasness and the P/E ratio is an indicator of this. Empirical research by BASU(1977) to determine whether investment performance of common stocks is related to their P/E ratios and findings reveals that returns on stocks with low P/E ratios tends to be larger than reasonable by the underlying risks, even after adjusting for transaction costs, and differential taxes which is inconsistent with the efficient market hypothesis.
In summary, the behaviour of security prices over the 14 –year period studied is, possibly, not fully described by the efficient market hypothesis to the extent that low P/E portfolios actually earn higher returns on a risk-adjusted basis which validate the price-ratio hypothesis on the relationship between investment performance and their P/E ratios. Other recent researchers includes Ahmed and Nanda (2001) use the growth in earnings-per-share (EPS) to capture growth in the US markets from 1982 to 1997. They show that strategies focusing on investing in stocks that have the dual characteristics of a high E/P ratio (ie value stocks) and a high growth in EPS outperform a high E/P alone. Bird and Whitaker (2003, 2004) focus on seven European markets (United Kingdom, France, Germany, Italy, Switzerland, the Netherlands, and Spain) from January 1990 to June 2002. They indicate that a combination of value and earnings momentum results in a small improvement when compared to using the sole book-to-market criterion.
2.3 Relationship between Return and Market Value of Common Stocks
The single-period capital asset pricing model (CAPM) postulates that there exist a linear relationship between market risk of a security and expected returns. However, Banz study examines the empirical relationship between the return and the total market value of NYSE common stocks. But direct test on this have not been concluded. Additional factor which is relevant for asset pricing is recently suggested. In 1979 Litzenberger and Ramaswany establish a significant positive relationship between return of common stocks and dividend yield for the period of 1936-1977. Although Basu in 1977 summarised his findings that P/E ratios and risk adjusted returns are related and concluded to interpret his findings as evidence of market ineffiency. But in 1978 Ball points out, some of the anomalies that have been responsible for the lack of market efficiency is a result of a misspecification of the pricing model. Between 1936-1975 Banz study also contributes another piece of emerging puzzle by examining the relationship between the total market value of common stocks of a firm and its returns.
Over a forty year period, The analysis revealed that CAPM is misspecified on average, small NYSE firms had had significantly larger risk adjusted return than large NYSE firms. Although the size effect is not in any way linear in the market proportion but it is most pronounced for the smallest firms in the sample.
In summary, the size effect exists but the reason still remains questionable until we find an answer to it. As a result of this extra caution is required during interpretation
2.4 EQUITY VALUATION M
3.0 Data and Research Methodology
3. 1 Data
The experimental analysis of this dissertation employ data from the Beazley Group Plc and Hardy Underwriting Plc annual reports for the financial year ended 31st December 2009, in other to make an informed decision by calculating relevant financial ratios. The annual report was obtained from the official website of the company.
3.2 Overview of the Company (Hardy Underwriting)
Hardy Underwriting Bermuda (HUB) is a management and holding company. The company’s major activity is acting as a managing agent at Lloyds’s for syndicate 382 by providing capacity and specialist underwriting expertise amongst wider range of insurance and re-insurance classes on a global basis. It’s line of business operation is divided into four segments: marine and aviation, specialty lines, non marine property and property treaty. The company carries out its operation through its subsidiaries namely Hardy Re Limited, a Bermuda-based reinsurance company and Hardy Bermuda Limited, a coverholder company, which assumes risk on behalf of syndicate 382 via a binding authority. It operates in four segments: Marine and aviation, Specialty lines, Non-marine property, and Property treaty. Its subsidiaries include Hardy Underwriting Group Plc, Hardy Re Limited, Hardy Bermuda Limited, Hardy Underwriting Limited and Hardy Names Limited. During the year ended December 31, 2009, it acquired 50% interest in a Bahrain-based company, HAIM Limited.
3.3 Business Description.
Hardy is a fully integrated insurance company, engaged in underwriting a wide range of insurance and reinsurance products. The company’s principal activities include managing agents and underwriters at Lloyd’s for syndicates 382. It underwrites through its wholly owned limited liability subsidiary, Hardy Underwriting Limited.
Through its Syndicate 382, the company underwrites commercial business except motor and liability sectors. The company is involved in reinsurance to other underwriters against the risks in non-marine catastrophes such as hurricanes and earthquake. The major coverage of the syndicate includes non-marine insurance lines such as direct property, accident and medical cover and financial institutions, property treaty, political risks and conveyancing. Syndicate 382 is specialized in airline and general aviation business, mainly focused on helicopters insurance. In addition, the company also has a business insuring ships and their cargo. The company is involved in a wide range of non-marine underwriting, which includes direct insurance and reinsurance.
The company conducts its business operations through four reportable segments: Marine and Aviation, Specialty Lines, Non-Marine Property and Property Treaty. The Marine and Aviation is a company’s core business segment. It principally focuses on risk coverage for general aviation, marine, and cargo and specie. In aviation insurance, the company is a market leader in risk coverage to rotor wing aircraft. The company underwrites range of marine risks, including fishing vessels, harbour craft and loss of hire. The company’s cargo and specie insurance products include coverage for jewelers block, fine art, small motor, classic cars and other niche areas. The company is organised into four business operating divisions: Management service, Adjusting service, Insurance support service and Insurance Company’s run-off.
The Specialty Lines segment provides insurance for financial institutions, political risks, terrorism and conveyancing and other miscellaneous niche lines and schemes. To financial institutions, the company offers bankers bond and professional indemnity coverages. Political insurance includes contract frustration, confiscation, trade credit, war on land and overseas terrorism. The conveyancing insurance portfolio include defective title, restrictive covenant, chancel repair, search and title risk coverage; offered to lenders, purchasers, sellers, owners and conveyancers of residential and commercial property. The Non-Marine segment provides risk coverages for direct & facultative insurance and direct property insurance. In Property Treaty, the company mainly offers Cat XL.
The company generated net insurance premium revenue of GBP 61.8 million from Marine and Aviation segment in 2009 as compare to GBP 43.862 million in 2008, followed by Specialty Lines generated GBP 33.968 millions in 2009 as compare to GBP 29.493 millions in 2008, Non-Marine Property segment generated GBP 37.468 millions in 2009 as compare to GBP 16.777 million in 2008, Property Treaty generated GBP 51.050 millions in 2009 as compare to GBP 33.469 millions in 2008.
The company carries out its operations through seven wholly owned subsidiaries: Hardy Underwriting Group Plc, Hardy Re Limited, Hardy Bermuda Limited, Hardy Underwriting Limited, Hardy Names Limited, Hardy (Underwriting Agencies) Limited and Hardy Insurance Services Limited (HIS).
Beazley Company Overview
Beazley Group plc is the holding company for the Beazley group, which engaged in the underwriting of specialist risk insurance and reinsurance business, through its managed syndicates 2623, 6107, 3623, 3622 and 623 at Lloyd’s in the United Kingdom and Beazley Insurance Company, Inc. Their underwriting business range for clients worldwide can be summarised into five segments: marine, which underwrites a range of marine classes, including hull, energy and cargo; political risks and contingency, which underwrites terrorism, political violence and credit risks; property, which underwrites commercial, homeowners and engineering property insurance; reinsurance, which specializes in writing property per risk, aggregate excess of loss and pro-rata business, and specialty lines, which underwrites professional lines, employment practices liability, specialty treaty, directors and officers liability and healthcare. In June 2009, Beazley plc acquired Beazley Group plc. Beazley Group plc was renamed Beazley Group Limited. On June 2, 2009, it acquired Beazley Re Limited. The group primarily operates in the US, the UK, Singapore, Hong Kong and other European countries.
Beazley is one the ultimate holding company for the Beazley Group, a global specialist risk insurance and reinsurance business. The company operates through Lloyd’s syndicate 2623 and 623 in the UK and BICI, a US-admitted carrier in the US. In the US, the underwriters of the company concentrate on writing the specialist insurance products in the admitted market, backed by Beazley Insurance Company, Inc., an admitted property/casualty carrier in all 50 states and the surplus lines risks are backed by the Beazley syndicates at Lloyd’s. The participation in the Lloyd’s market gives the company an opportunity to access to the brokered business all around the world and also enables the company to underwrite the admitted business in the US.
The company operations are divided into four reportable segments namely Marine, Property, Reinsurance and Specialty Lines.
The company through its Marine segment offers long term insurance solutions to the maritime related industries. It underwrites all marine classes. The clients include ship-owners, chatterers, port authorities, cargo owners and energy companies. It provides specialist coverage for specie & cargo, hull, war and energy risks as well as other risks exposed to catastrophes. In addition, it includes coverage to voyage and towage risks, total loss only, building risks, port risks,yachts/pleasure craft, mortgages interest and offshore and onshore exposures of production and exploration companies.
The company helps to insure 10% of the world’s oceangoing tonnage and cover 35% of the top 200 oil and gas companies.
The Property segment of the company comprised of insures engineering property, commercial property and high-value homeowners. The clients include companies in mining, steel, utilities, retail, commercial real estate, homeowners, jewelers, art dealers, private collectors and construction. The company operates its underwriting business in the US, Singapore and London. The company underwrites the commercial property risks in the US for large risks, small risks, US both admitted
and non-admitted and Binding Authorities; engineering and construction risks for Contractors All Risks (CAR), Erection All Risks (EAR), Machinery Breakdown (MB), Machinery Loss of Profits (MLOP), Electronic Equipment (EEI) and Plant and Equipment All Risks (P.A.R) in Singapore; and homeowners for high value homeowners risks in the UK & overseas, the US high value and Binding authorities. The company’s clients range from Fortune 1000 companies to homeowners.
The company through its Reinsurance segment provides reinsurance products and services to general insurers. It concentrates in the areas of casualty catastrophe, aggregate excess of loss, property catastrophe, pro-rata business and property per risk. The company is specialized in writing the property catastrophe excess of loss on a worldwide basis. The diverse client base includes regional cedants, large nationwide carriers and multi regional companies. Under the casualty
clash risk the company’s portfolio mainly comprises of programmes protecting client’s auto and workers compensation accounts. The segment covers the areas of Australasia, the US, UK, Canada, Continental Europe, West Indies and Japan.
Under its Specialty Lines segment, the company underwrites specialty insurance and reinsurance, including professional liability, directors’ and officers’ liability, employment practices’ liability, healthcare, terrorism, contingency and political risks. The segment even underwrites management liability business on both a primary and excess basis, from Europe, North America and around the world. The US clients of the company are served by underwriters at Lloyd’s and by private enterprise team who focus on smaller scale clients. The clients of the professional liability include Lawyers, Healthcare, Architects & Engineers, Programmes, Technology, Media & Business Services, Specialty treaty and Programmes. The clients of the management liability include: Public company D&O, Employment practices liability, Non-profit organizations, Private company management liability, Crime and Fiduciary.
he subsidiaries of the company are Tasman Corporate Limited, Beazley Furlonge Holdings Limited, Beazley Furlonge Limited and Beazley Pte. Limited, BFHH Limited, Beazley Investments Limited, Beazley Corporate Member No. 2, Beazley Corporate Member Limited, Global Two Limited, Beazley Underwriting Limited, Beazley Management Limited, Beazley Staff Underwriting Limited, Beazley Solutions Limited, Beazley Corporate Member No. 3, Beazley USA Services, Inc., Beazley Holdings, Inc., Beazley Insurance Company, Inc., Beazley Limited Hong Kong, Beazley Group (USA) General Partnership and Beazley Dedicated No.2 Limited.
During the fiscal year 2007, the company generated 49.12% of total revenue from Specialty lines segment, 24.75% from Property, 7.96% from Reinsurance and 18.15% from Marine segment.
The management service includes the following: mutual management, investment management, underwriting services, captive management, and risk management
The mutual management is engaged in the development and management of mutual insurance associations. It provides the full range of management skills including underwriting, risk assessment, accounting, documentation, claims negotiation and paying, investment and regulatory and compliance functions.
The captive management is engaged in the identification, evaluation and implementation of a captive insurance solution. Its activities are carried out principally in Bermuda through CTC Allegro Insurance and Risk Manageme
Cite This Work
To export a reference to this article please select a referencing stye below: