Finansal Hizmetler Güven Endeksi ile Borsa Getirileri arasındaki ilişki: Toda-Yamamoto ve Asimetrik Nedensellik Analizi
Yunus Emre Turan, Dinara ZubaidullinaFinansal piyasaların dinamiklerini şekillendirmede beklentilerin yönetilmesi önemli bir husustur. Bu nedenle, iktisadi ajanların beklentilerini yansıtan güven endekslerindeki değişimler ekonomistler tarafından dikkatle takip edilmektedir. Bu çalışmada Türkiye’de Haziran 2012-Mayıs 2024 dönemi için finansal hizmetler güven endeksi ve borsa endeks getirileri arasındaki sektörel bazlı ilişkiler açıklanmaktadır. Çalışmada nedensellik sonuçların üzerinde karşılaştırma yapılabilmesi amacıyla değişkenler arasındaki ilişki Toda-Yamamoto (1995) ve Hatemi-J (2012) nedensellik testleri kullanılarak incelenmiştir. Toda-Yamamoto(1995) nedensellik testi sonuçlarına göre finansal güven endeksi ile Borsa İstanbul (BİST) sektör endeksleri getirileri arasında bir nedensellik ilişkisinin varlığı tespit edilememiştir. Bununla birlikte, değişkenler arasında asimetrik bir nedensellik ilişkisi saptanmıştır. Böylelikle, Toda-Yamamoto (1995) nedensellik testinin tespit edemediği saklı ilişkilerin asimetrik nedensellik testi yardımıyla ortaya çıkarıldığı görülmektedir. Asimetrik nedensellik testi sonuçlarına göre, negatif şoklar için finansal hizmetler güven endeksinden BİST Mali endeks getirilerine doğru nedensellik bulunmaktadır. BİST Elektrik endeksi ve BİST Finansal Kiralama Faktöring endeksi getirilerindeki hem pozitif hem de negatif şoklardan finansal hizmetler güven endeksine doğru bir nedensellik ilişkisi bulgusuna ulaşılmıştır. BİST Bilişim endeksi ve BİST Teknoloji endeksi getirilerinden finansal hizmetler güven endeksine doğru bir nedensellik pozitif bileşenlerde tespit edilmiştir. BİST Hizmetler endeksi ve BİST Sınai Endeksi getirilerinden finansal hizmetler güven endeksine doğru bir nedensellik ilişkisi de negatif şoklar arasında tespit edilmiştir. Çalışmadan elde edilen sonuçlar ışığında, finansal hizmetler güven endeksinin finansal piyasalardaki değişimler açısından dikkate alınması gereken bir değişken olduğunu ifade etmek mümkündür.
Relationship between Financial Services Confidence Index and Stock Market Returns: Toda-Yamamoto and Asymmetric Causality Analysis
Yunus Emre Turan, Dinara ZubaidullinaThe management of expectations is a critical factor in shaping the dynamics of financial markets. Therefore, changes in confidence indices, which reflect the expectations of economic agents, are closely monitored by economists. This study examines the sectoral relationships between the Financial Services Confidence Index and stock market returns in Turkey for the period from June 2012 to May 2024. To compare the causality results, the relationships between the variables were analysed using the Toda-Yamamoto (1995) and Hatemi-J (2012) causality tests. According to the results of the Toda-Yamamoto (1995) causality test, no causality relationship was found between the Financial Services Confidence Index and the Borsa Istanbul (BIST) sector index returns. However, an asymmetric causality relationship was identified between the variables. Thus, it is observed that hidden relationships, which could not be detected by the Toda-Yamamoto (1995) causality test, were revealed through the asymmetric causality test. According to the results of the asymmetric causality test, a causality relationship from the Financial Services Confidence Index to the BIST Financial index returns was found for negative shocks. There is a causality relationship from the BIST Electricity index and the BIST Financial Leasing and Factoring index returns to the Financial Services Confidence Index in both positive and negative shocks. A causality relationship from the BIST Information Technology index and the BIST Technology index returns to the Financial Services Confidence Index was found in positive components. A causality relationship from the BIST Services index and BIST Industrial index returns to the Financial Services Confidence Index was also found among the negative shocks. Considering the results obtained from the study, it can be stated that the Financial Services Confidence Index is a variable that should be considered in terms of changes in financial markets.
Classical finance theories assume that investors are rational and that they consider all information in the market during the decision-making process. The number of studies that include and try to develop behavioural aspects that are often neglected in classical theoretical models is increasing in the literature. The increasing interest in the non-classical approach stems from the need to explain phenomena that are regularly observed in financial markets and that do not comply with the predictions of classical models (Yoshinaga and Castro Junior, 2012).
De Long et al. (1990) divided investors into two categories, considering the effect of financial market sentiment: rational investors and rumour merchants. In the model, investors in both categories determine the prices of assets and their expected returns. The possibility of rational investors profiting from mispricing is limited due to various reasons such as short investment horizon, transaction costs, and risks. These reasons allow prices to deviate from their fundamental values. Sentiment in the context of investment may indicate overly optimistic or pessimistic cash flow estimates or changes in risk tolerance (Edelen et al., 2010). As a result of this situation, the relationship between the net present value of the discounted cash flows of the asset and the asset price may occur differently. Here, investors’ emotional reactions may have an effect.
The link between investor sentiment and stock returns has been examined in many studies. Chowdhury et al. (2024) found a positive effect of investor sentiment on stock returns in the S&P 500 index. Schmeling (2009) found a negative relationship between investor sentiment and stock returns. It is observed that there is a separation in the variables selected as sentiment proxies in the studies conducted in the literature. The variables used in previous studies: confidence indices obtained from investor surveys (Usul et al., 2017; Kandır, 2006); investor mood (Kostopoulos and Meyer, 2018); implied volatility of the option contract (Qadan et al., 2019); closed-end investment fund discount (Keleş and Arat, 2019); investment fund flows (Ben-Rephael et al., 2012); transaction volume or transaction amount (Ding et al., 2014) and the composite sentiment index combining these proxies (Baker and Wurgler, 2006).
The Central Bank of the Republic of Turkey (CBRT) calculates the Financial Services Confidence Index (FSCI) to reveal the assessments and expectations regarding the business situations and demand for services in the financial services sector. A financial service survey was conducted to calculate the FSCI. With the monthly financial services survey, the recent past assessments and future expectations of financial institution managers regarding their business situations are obtained and trends in the financial services sector are determined. The FSCI index calculated from the survey results is calculated as a function of the answers to the questions about the business status of financial institutions, the demand for services in the last three months, and the demand for services in the next three months (TCMB, 2024).
The expectations of economic agents affect macroeconomic and financial indicators. Therefore, examining the effect of confidence indices reflecting expectations on the stock market is an important issue. This study aims to empirically analyse the causality relationship between the financial confidence index and stock returns in Turkey in the period 2012:06-2024:05. In addition to the Toda-Yamamoto causality test, the Hatemi-j (2012) asymmetric causality test was used in the study. According to the TodaYamamoto (1995) causality test results, no causality relationship was determined between the financial services confidence index and the Borsa Istanbul (BIST) sector index returns. However, an asymmetric causality relationship was determined between the variables. Thus, the hidden relationships that the Toda-Yamamoto (1995) causality test could not detect were revealed with the help of the asymmetric causality test.