Für neue Autoren:
kostenlos, einfach und schnell
Für bereits registrierte Autoren
39 Seiten, Note: 1
2. Literature review
2.1. Developed countries
2.2. Emerging countries
3.1. Country-by-country regressions
3.2. Speed of adjustment measurement
3.3. Market efficiency measurement
3.4. Size distribution measurement
3.5. Pooled regression
5.1. Country-by-country regressions
5.2. Pooled regression
Contrary to the efficient market hypothesis, there is irrefutable support that short-term horizon stock returns are predictable. Advocates of this paradigm suggest that stock prices do not follow a random walk as implied by Fama (1970) since significant deviations can be observed in the form of autocorrelation and cross-serial correlation of short-term portfolio returns (Lo and MacKinlay, 1988). Closely related to the predictability of short-term stock prices, numerous studies have examined whether portfolios formed on specific characteristics could be ascribed forecasting power over other portfolios.
Amongst the abovementioned studies, the present paper focuses on predictable patterns of stock returns determined by size. This study analyzes size-driven lead-lag effects whereby movements in large capitalization stocks lead in time changes of small capitalization stocks. Returns of large companies hence gain forecasting power over returns of small companies, giving rise to movements that could be anticipated and exploited. Additionally, we explore new causal factors of the size-related lead-lag effect, namely size dispersion and market efficiency.
Starting with the seminal work of Lo and MacKinlay (1990), extensive literature has focused on studying the driving forces of lead-lag effects. Amongst these, lead-lag relationships are documented between portfolios sorted on size, volume, institutional ownership and analyst coverage. In an approximately universal manner size has proved to be an underlying factor of short-term portfolio predictability. However, most studies are investigating developed countries, while less emphasis is put on emerging markets. Having considered this research gap, the purpose of this study is to examine size-driven lead-lag patterns in both emerging and developed European markets under a common methodology, following the guidelines of Brennan, Jegadeesh and Swaminathan (1993).
In addition, our objective is to find new explanations of lead-lag effects. Assuming the level of stock market development determines distinct market characteristics, we investigate whether the forecasting power of large capitalization firms over small capitalization firms differs between developing and advanced economies. Further, we employ a measure for the lead-lag effect similar to that used by McQueen, Pinegar and Thorley (1996) and analyze the underlying factors that determine the slower adjustment of small companies returns to large companies returns. These factors have never been considered by previous literature and refer to market efficiency and size distribution.
The relevance of the present study refers to identifying another form of market inefficiency and its underlying causes. Having considered the core determinants, measures can be taken to improve stock markets efficiency. If large companies returns can predict small companies returns, then this market efficiency anomaly could be exploited. On the one hand, an investment strategy consisting of a long position in large companies and a short position in small ones would mark arbitrage profits. On the other hand, the economy could benefit from inducing a more efficient market infrastructure in the form of higher investments level, more efficient capital allocation, attracting foreign investors and boosting overall economic growth (see, e.g., Barro (1990), Levine and Zervos (1996), Christopoulos and Efthymios (2004), Boubakari and Jin (2010)). The study is also significant in concluding if extensive measures should be taken in emerging countries compared to developed ones in order to promote an efficient market structure.
This research contributes to prior literature by providing a common method for identifying lead-lag patterns in emerging and developed markets. Under this universal approach, differences in the forecasting ability of large companies returns based on the level of market development can be detected. Additionally, we formally test two hypotheses concerning the fundamental determinants of distinct sizes of lead-lag effects in emerging and advanced economies. This methodology includes two unique factors never considered by previous research: size distribution and market efficiency. Additionally, our method enables us to distinguish if size is the actual driver of lead-lag patterns by using a distinct approach than sorting portfolios on different variables and then controlling for size.
Based on assumptions further developed in Section 2, we expect that market inefficiencies broadens the lead-lag patterns in emerging markets, while heterogeneous size distributions in developed markets cause higher lags between small and large companies. However, country-by-country regressions demonstrate that sluggish adjustment of small companies returns is not universal and no clear distinction can be made between emerging and developed countries. Our pooled regression provides evidence that supports the hypothesis that size is the actual driver of lead-lag patterns, whereas there is only limited proof to validate that market inefficiency can determine a more pronounced leading role of large firms returns. Additionally, size homogeneity narrows the size of the lead-lag effect in emerging countries.
The remainder of this paper is structured as follows. Section 2 provides an outline of the relevant literature concerning lead-lag effects, followed by a presentation of the employed main methodology and additional robustness checks in Section 3. Portfolio sorting and formation and descriptive statistics are presented in Section 4. We show our results and inference in Section 5, while Section 6 concludes.
The existence and significance of lead-lag relationships between size-sorted portfolios have been attested in a universal manner in developed countries (see Lo and MacKinlay (1990), Martikainen, Perttunen, Puttonen (1995), McQueen et al. (1996), Yu and Wu (2001), Mills and Jordanov (2001), Hameed and Kusnadi (2003), Kanas (2004), Kanas and Kouretas (2005), Hou (2007)). The conclusion of extensive literature regarding lead-lag patterns suggests that the returns of large capitalization stocks lead the returns of small capitalization stocks, creating an asymmetric cross-autocorrelation structure based on size and assigning large companies the power to predict the price movements of small companies.
The discovered relationship between contemporaneous returns of small companies and lagged returns of large ones is proven to be robust to various measurements and methods. Both simpler models such as ordinary least squares or variations exploiting autoregressive conditional heteroskedasticity (McQueen et al., 1996, Iwaisako, 2003) and more complex methods such as Granger causality, vector autoregressive or vector error correction framework, simultaneous equations and cointegration (Mills and Jordanov, 2001, Hameed and Kusnadi, 2003, Kanas and Kouretas, 2005, Karmakar, 2010) establish the underlying consensual finding that returns of large companies lead those of small ones.
The rationales under the size-based lead-lag effect distinguish between four different explanations. The broadest reason for the existence of size cross effects refers to slower adjustment of small stocks to market-wide information. Additionally, a number of studies argue that small firms lag large ones partly due to less transparent financial activities and less efficient information disclosure. Contrary to the abovementioned propositions, another group of scholars claim that markets process information rationally and the presence of the lead-lag pattern is a result of market imperfections. Lastly, size cross-autocorrelations are regarded as spurious as size is not in itself an information transmission mechanism; cross effects are thus explained by either correlations between size and other firm characteristics or by individual assets’ own autocorrelations.
To begin with, the different information adjustment speed is assumed to be the principal cause of large companies leading small ones (see McQueen et al. (1996), Mills and Jordanov (2001), Kanas (2004), Hou (2007)). Small capitalization stocks have a delayed reaction to common news and respond slower to new information in the market. This sluggish adjustment to market-wide information falls under three related hypotheses relying on information transmission.
First, the trading cost hypothesis presumes that information-based trades are executed first where they produce the highest net profit (Fleming, Ostdiek, Whaley, 1996). Given the assumption that trading costs are lower for large companies than for small ones, they will be favored in transactions. This postulation also implies that there is a differential between the liquidity of large and small capitalization stocks given that trading volumes are lower and trading costs (mainly composed of bid-ask spreads) are higher for small stocks. Hence, price discovery occurs first in lowest cost stocks – namely large stocks, as the profit from information-based trading is lower for small capitalization stocks. Another valid hypothesis to explain the lead-lag structure is the set-up cost associated with information processing. Stocks are considered and analyzed only if the added value exceeds the set-up cost; hence, investors focus on a subset of stocks, which, as argued by Merton (1987), tend to belong to large capitalization firms. The third hypothesis deems that informational shocks to large capitalization stocks have significant systematic information content (Badrinath, Kale and Noe (1995)). News regarding large stocks often contains information about the future prospects of small stocks, which will only further be incorporated in their price. Under market imperfections, this systematic component is thus first impounded in the price of large stocks, and only later used as a signal for small stocks, giving rise to a lead-lag relationship.
In addition to slower adjustment to market-wide information, small companies are subject to adverse selection. As reported by Buzby (1975), information disclosure for small companies is less adequate than compared to large companies, where information disclosure is defined as the informational needs of professional financial analysts that assess the investment value of common equities. Therefore, a higher level of effective information disclosure for large capitalization stocks and the higher quality of their information signals might lead to faster adjustment to information. Also, Yu and Wu (2001) document that different quality in cash-flow information between size-sorted portfolios is a fundamental determinant of large stocks leading small ones.
The third explanation attributes cross effects to market frictions (Lo and MacKinlay, 1990, Boudoukh, Richardson and Whitelaw, 1994). These market imperfections, mostly observed in the form of nonsynchronous trading, thin trading or discontinuous and non-synchronized trading periods, introduce data errors in terms of securities prices mistakenly assumed to be sampled simultaneously. These errors further induce cross-autocorrelations in the sampled stock returns and lead-lag effects arise. Distinct from the aforementioned hypotheses, this presumption does not violate market efficiency as stocks incorporate all available information, although this information in itself contains timing errors.
Lastly, a group of researchers regard size-related cross-autocorrelations as a coincidental result. On the one hand, lead-lag patterns are attributed to individual assets’ own autocorrelations (Conrad, Gultekin and Kaul, 1991). On the other hand, it has been argued that size is not itself a relevant variable in generating lead-lag patterns, but it is positively correlated to other significant factors that are underlying determinants of large companies leading small ones. Badrinath et al. (1995) and Sias and Startks (1997) report that portfolios with high institutional ownership lead portfolios with low institutional ownership as institutional trades contain market-wide information, which is only later mimicked by individual traders and impounded in other stocks as well. Size is also a proxy for the magnitude and quality of information produced, which increases with the number of following analysts (see Brennan et al. (1993), Chan and Hameed (2006)). They conclude that high analyst coverage portfolios reflect information more rapidly than low analyst coverage portfolios. In addition, Chordia and Swaminathan (2000) conclude that size is positively related to the trading volume. The rationale behind this implication suggests that the speed of adjustment to market information differs based on the trading costs that stocks employ. As large portfolios are traded more and at lower costs, a lead for large capitalization stocks is created.
Regarding emerging countries, the lead-lag patterns are not always unidirectional, similar for all countries or as clear as for advanced economies. If evidence from developed countries suggests that small companies lag large ones, results for emerging economies do not consistently agree with this established pattern. Researchers mainly attribute these differences to a more diverse range of market structures, security laws enforcement and investor protection. Also, the above stated hypotheses to explain the existence of a size-based lead-lag structure do not apply in the same manner as for developed countries as stock prices in emerging economies are driven more by behavioral aspects and sentiment rather than by information.
The only explanation of cross-autocorrelations that applies consistently refers to capital markets laws and imperfections. For instance, Kang, Liu and Ni (2002) reveal that on the Chinese market the leading portfolio varies based on the holding period, result which is ascribed to Chinese governmental intervention and predominant individual ownership. Both Chiao, Hung and Lee (2004) and Chiao, Hung and Srivastava (2004) support the idea that the absence of a lead-lag structure on the Taiwanese market is a direct consequence of the restricted price variation imposed by the government, high proportion of domestic institutional investors and proportional transaction costs across size-sorted groups of securities. Rehman and Rehman (2010) document a bidirectional flow between small and large capitalization stocks on the Pakistani stock exchange, which is presumably justified by indices favoring large companies and a strong home bias within national asset selection.
For more solid and unrestricted emerging stock exchanges, the existence of a size-based cross effect is not rejected (see Gebka (2008) for Poland, Karmakar (2010) for India). This finding is explained by the information adjustment hypothesis as presented for developed countries. Large companies impound new information more rapidly and act as proxies for macroeconomic news, having predictive power over the dynamics of small firms returns. Furthermore, Chang, McQueen and Pinegar (1999) report that although lead-lag structures are not US unique or developed countries unique, they are strongest in the US. The potential explanation states that greater heterogeneity in liquidity, volume and market restrictions determines stronger cross-autocorrelations (Boudoukh et al. (1994)), hence more powerful lead-lag effects on the US market.
When analyzing lead-lag effects in developed and emerging markets, the abovementioned studies present disagreements in findings. Therefore, we use the differences between developing and advanced markets in order to shed light over the existence and validity of the explanations that justify the lead-lag pattern. Four distinct presumptions are to be taken into account in order to investigate if cross effects develop differently depending on stock market growth. Hence, the overall purpose is to examine if there is a difference in the significance and size of the lead-lag effects in emerging and developed countries and to explain potential dissimilarities.
The first underlying origin of a distinct lead-lag pattern in developed and emerging countries refers to different degrees of investor protection and ownership concentration. Developing countries are highly associated with less investor protection in the form of security laws, enforcement and transparency. In response to poorer protection of emerging markets, ownership concentration arises as a substitute meant to ensure improved company monitoring and to solve principal-agent conflicts (see, e.g., La Porta, Lopez-de-Silanes, Schleifer and Vishny (1998), Himmelberg, Hubbard and Love (2002)). Additionally, ownership concentration is observed mostly in large capitalization companies (La Porta et al.,1998, Pedersen and Thomsen, 1999). In turn, stocks liquidity is negatively related to block holdings (Glosten, 1985, Rubin, 2007), which is consistent with the trading cost hypothesis. Hence, large companies are less liquid and given that price discovery occurs first in lowest cost environments, large stocks do not have an advantage over small ones. Therefore, emerging markets with fewer investor protection measures are dominated by block holding of large companies, reducing large stocks liquidity and increasing their trading costs. Following this line of reasoning, the lead-lag effect should be smaller in emerging economies as trading occurs on an informed basis and the trading costs place large companies at a disadvantage. Figure 1 presents the arguments above in a graphical means.
illustration not visible in this excerpt
Figure 1: The line of reasoning under the first hypothesis concerning different sizes of lead-lag effects in emerging and developed markets.
A second supporting argument of distinct sizes in lead-lag effects based on country development refers to market and indices structures. Stock market structure in emerging countries is based on distinct composition and consistency than that of advanced countries in a manner that makes developed markets prone to larger lead-lag structures. These differences fall under the market imperfections explanation of size-based cross effects. Emerging market indices usually favor large companies leading to a homogeneous size distribution within an index (Rehman and Rehman, 2010) and bid-ask spreads are proportional between large and small companies in developing markets (Chang et al., 1999, Chiao et al., 2004). Homogeneous size and spreads distributions imply more similarities between companies, hence less differences that could cause predictable patterns. The above stated distinctions support the idea that – given the homogeneity size and bid-ask spreads in emerging economies, the magnitude of the lead-lag effect should be smaller in comparison to advanced economies. This hypothesis directly refers to the ability of size itself to determine lead-lag structures and relies on relative size distribution to explain if emerging markets have less pronounced predictable patterns driven by size than advanced economies. This hypothesis is synthesized graphically by Figure 2.
illustration not visible in this excerpt
Figure 2: The line of reasoning under the second hypothesis concerning different sizes of lead-lag effects in emerging and developed markets.
Opposing these views, a third assumption lies on the information disclosure hypothesis explaining lead-lag patterns. Corporate transparency and information quality is less adequate for small companies in comparison to large ones in both emerging and developed countries (as reported by Buzby (1975)). However, analyst coverage and the information and analyses they produce is influenced by country-specific variables – such as average firm size, the size of the stock market or the accounting disclosure, in a way that emerging stock markets are less favored by analysts and the accuracy of their forecasts is less precise (Chang, Khanna and Palepu, 2000). In addition, firms in advanced economies have a higher level of effective information disclosure when compared to firms in emerging countries (see Salter (2002), Chan and Hameed (2006)). Considering these informational disadvantages of emerging markets, it is natural to assume greater cross effects in developing markets. This is an indirect outcome of different signaling power and speed of adjustment to information between developed and emerging countries. Figure 3 presents the abovementioned arguments.
illustration not visible in this excerpt
Figure 3: The line of reasoning under the third hypothesis concerning different sizes of lead-lag effects in emerging and developed markets.
Bachelorarbeit, 17 Seiten
Diplomarbeit, 62 Seiten
Masterarbeit, 93 Seiten
Diplomarbeit, 73 Seiten
Bachelorarbeit, 36 Seiten
Diplomarbeit, 75 Seiten
Bachelorarbeit, 38 Seiten
Diplomarbeit, 78 Seiten
Diplomarbeit, 78 Seiten
Der GRIN Verlag hat sich seit 1998 auf die Veröffentlichung akademischer eBooks und Bücher spezialisiert. Der GRIN Verlag steht damit als erstes Unternehmen für User Generated Quality Content. Die Verlagsseiten GRIN.com, Hausarbeiten.de und Diplomarbeiten24 bieten für Hochschullehrer, Absolventen und Studenten die ideale Plattform, wissenschaftliche Texte wie Hausarbeiten, Referate, Bachelorarbeiten, Masterarbeiten, Diplomarbeiten, Dissertationen und wissenschaftliche Aufsätze einem breiten Publikum zu präsentieren.
Kostenfreie Veröffentlichung: Hausarbeit, Bachelorarbeit, Diplomarbeit, Dissertation, Masterarbeit, Interpretation oder Referat jetzt veröffentlichen!