Investor attention in Borsa Istanbul: analysis of the effect of Google Trends and Youtube views on initial public offering performances by Random Forest Method

Yunus Emre Akdoğan

Makale detayları

Investor attention in Borsa Istanbul: analysis of the effect of Google Trends and Youtube views on initial public offering performances by Random Forest Method

Yazarlar:

Yunus Emre Akdoğan

Yayıncı:

Hitit University

Ana Başlıklar:

Davranışsal Finans | Finansal Öngörü ve Modelleme | Yatırımlar ve Portföy Yönetimi

Dergi:

Hitit Sosyal Bilimler Dergisi Vol.17 Issue 1, pp.70-90

Belge Tipi:

Makale

Yayın Tarihi:

4/2024

Belge Dili:

Türkçe

Document ID (Org No.):

2001315014

pISSN:

2757-7449

Anahtar kelimeler:

Özet:

 The efficient markets hypothesis, which is one of the asset pricing models of traditional finance theory, is based on the assumption that publicly available information is reflected in prices by rational investors with perfect information, and therefore it is not possible to obtain abnormal returns. On the other hand, models based on the assumption of bounded rationality argue that investors have cognitive constraints and one of these constraints is investor attention. Investor attention is a cognitive constraint that leads investors to focus only on a set of information, thus limiting their access to information. This constraint is used as a signal for stock price movements because it leads investors to buy only those stocks about which they have information. However, there are different views on how to measure investor attention. Approaches that measure investor attention indirectly use indirect proxies such as price, liquidity, return, and advertising expenditures, while approaches that measure investor attention directly either ask investors directly or monitor their behavior. Especially with the advancements in information and communication technologies, the widespread use of social media by investors to access investment ideas offers new tools to directly measure investor attention. Among these tools based on investors information-seeking behaviors are Google and Baidu search volumes, Wikipedia page views, and tweets. Another financial dynamic influenced by investor attention is initial public offerings IPOs. An IPO refers to the process of offering a firms shares to the public in the primary market. Therefore, in an IPO, firms raise capital by offering their shares through the primary market. Predicting the performance of a stock in an IPO is crucial for both firms and investors. In this study, we use Google search volumes of firm names and symbols prior to the IPO as proxies of investor attention and the number of views of YouTube content that provides brief information about the IPO company and examine their impact on the performance of initial public offerings in Borsa Istanbul. The study also investigates which firm-specific characteristics are related to investor attention. In other words, it is also investigated which variables such as IPO size, number of shares, IPO price, reason for IPO, sector of the firm, and financial ratios attract more investor attention. As a result of the research, it was found that YouTube view counts are a better proxy for investor attention than Google search volumes. Within the scope of this study, both instruments used as proxies of investor attention are consistent with the assumption of bounded rationality, where perfect information is not available and cognitive abilities are limited. Additionally, the findings also suggest that investors value a firms fundamental value and internal growth potential. Another aim of this study is to develop narrow artificial intelligence capable of making decisions regarding IPOs or supporting investor decisions within the scope of algorithmic financial transactions. Artificial intelligence is defined as the ability of a computer to perform various activities similar to intelligent beings. Super artificial intelligence refers to those with superhuman abilities; general artificial intelligence is designed to perform any task a human can perform; narrow artificial intelligence refers to artificial intelligence designed to perform a specific task. Therefore, each model developed within the scope of narrow artificial intelligence actually adds a new skill to general artificial intelligence. In this context, the findings obtained from this study provide a significant contribution to the development of financial artificial intelligence applications, especially in training bots engaged in algorithmic financial transactions related to IPOs. In other words, in practice, this study presents a narrow financial artificial intelligence product whose framework is limited to IPO performances, while in theory it reveals the determinants of investor attention.

 The efficient markets hypothesis, which is one of the asset pricing models of traditional finance theory, is based on the assumption that publicly available information is reflected in prices by rational investors with perfect information, and therefore it is not possible to obtain abnormal returns. On the other hand, models based on the assumption of bounded rationality argue that investors have cognitive constraints and one of these constraints is investor attention. Investor attention is a cognitive constraint that leads investors to focus only on a set of information, thus limiting their access to information. This constraint is used as a signal for stock price movements because it leads investors to buy only those stocks about which they have information. However, there are different views on how to measure investor attention. Approaches that measure investor attention indirectly use indirect proxies such as price, liquidity, return, and advertising expenditures, while approaches that measure investor attention directly either ask investors directly or monitor their behavior. Especially with the advancements in information and communication technologies, the widespread use of social media by investors to access investment ideas offers new tools to directly measure investor attention. Among these tools based on investors information-seeking behaviors are Google and Baidu search volumes, Wikipedia page views, and tweets. Another financial dynamic influenced by investor attention is initial public offerings IPOs. An IPO refers to the process of offering a firms shares to the public in the primary market. Therefore, in an IPO, firms raise capital by offering their shares through the primary market. Predicting the performance of a stock in an IPO is crucial for both firms and investors. In this study, we use Google search volumes of firm names and symbols prior to the IPO as proxies of investor attention and the number of views of YouTube content that provides brief information about the IPO company and examine their impact on the performance of initial public offerings in Borsa Istanbul. The study also investigates which firm-specific characteristics are related to investor attention. In other words, it is also investigated which variables such as IPO size, number of shares, IPO price, reason for IPO, sector of the firm, and financial ratios attract more investor attention. As a result of the research, it was found that YouTube view counts are a better proxy for investor attention than Google search volumes. Within the scope of this study, both instruments used as proxies of investor attention are consistent with the assumption of bounded rationality, where perfect information is not available and cognitive abilities are limited. Additionally, the findings also suggest that investors value a firms fundamental value and internal growth potential. Another aim of this study is to develop narrow artificial intelligence capable of making decisions regarding IPOs or supporting investor decisions within the scope of algorithmic financial transactions. Artificial intelligence is defined as the ability of a computer to perform various activities similar to intelligent beings. Super artificial intelligence refers to those with superhuman abilities; general artificial intelligence is designed to perform any task a human can perform; narrow artificial intelligence refers to artificial intelligence designed to perform a specific task. Therefore, each model developed within the scope of narrow artificial intelligence actually adds a new skill to general artificial intelligence. In this context, the findings obtained from this study provide a significant contribution to the development of financial artificial intelligence applications, especially in training bots engaged in algorithmic financial transactions related to IPOs. In other words, in practice, this study presents a narrow financial artificial intelligence product whose framework is limited to IPO performances, while in theory it reveals the determinants of investor attention.

 The efficient markets hypothesis, which is one of the asset pricing models of traditional finance theory, is based on the assumption that publicly available information is reflected in prices by rational investors with perfect information, and therefore it is not possible to obtain abnormal returns. On the other hand, models based on the assumption of bounded rationality argue that investors have cognitive constraints and one of these constraints is investor attention. Investor attention is a cognitive constraint that leads investors to focus only on a set of information, thus limiting their access to information. This constraint is used as a signal for stock price movements because it leads investors to buy only those stocks about which they have information. However, there are different views on how to measure investor attention. Approaches that measure investor attention indirectly use indirect proxies such as price, liquidity, return, and advertising expenditures, while approaches that measure investor attention directly either ask investors directly or monitor their behavior. Especially with the advancements in information and communication technologies, the widespread use of social media by investors to access investment ideas offers new tools to directly measure investor attention. Among these tools based on investors information-seeking behaviors are Google and Baidu search volumes, Wikipedia page views, and tweets. Another financial dynamic influenced by investor attention is initial public offerings IPOs. An IPO refers to the process of offering a firms shares to the public in the primary market. Therefore, in an IPO, firms raise capital by offering their shares through the primary market. Predicting the performance of a stock in an IPO is crucial for both firms and investors. In this study, we use Google search volumes of firm names and symbols prior to the IPO as proxies of investor attention and the number of views of YouTube content that provides brief information about the IPO company and examine their impact on the performance of initial public offerings in Borsa Istanbul. The study also investigates which firm-specific characteristics are related to investor attention. In other words, it is also investigated which variables such as IPO size, number of shares, IPO price, reason for IPO, sector of the firm, and financial ratios attract more investor attention. As a result of the research, it was found that YouTube view counts are a better proxy for investor attention than Google search volumes. Within the scope of this study, both instruments used as proxies of investor attention are consistent with the assumption of bounded rationality, where perfect information is not available and cognitive abilities are limited. Additionally, the findings also suggest that investors value a firms fundamental value and internal growth potential. Another aim of this study is to develop narrow artificial intelligence capable of making decisions regarding IPOs or supporting investor decisions within the scope of algorithmic financial transactions. Artificial intelligence is defined as the ability of a computer to perform various activities similar to intelligent beings. Super artificial intelligence refers to those with superhuman abilities; general artificial intelligence is designed to perform any task a human can perform; narrow artificial intelligence refers to artificial intelligence designed to perform a specific task. Therefore, each model developed within the scope of narrow artificial intelligence actually adds a new skill to general artificial intelligence. In this context, the findings obtained from this study provide a significant contribution to the development of financial artificial intelligence applications, especially in training bots engaged in algorithmic financial transactions related to IPOs. In other words, in practice, this study presents a narrow financial artificial intelligence product whose framework is limited to IPO performances, while in theory it reveals the determinants of investor attention.

 Geleneksel finans teorisinin varlık fiyatlama modellerinden biri olan etkin piyasalar hipotezi, kamuya açık bilginin tam bilgiye sahip olan rasyonel yatırımcılar tarafından fiyatlara yansıtıldığı varsayımına dayanmakta ve dolayısıyla normalüstü getiri elde etmenin mümkün olmadığı görüşünü savunmaktadır. Diğer taraftan sınırlı rasyonalite varsayımına dayanan modellerde ise yatırımcıların bilişsel kısıtlarının olduğu ve bu kısıtlardan birisinin de yatırımcı ilgisi olduğu görüşü hakimdir. Yatırımcı ilgisi, yatırımcıların sadece bir dizi bilgiye odaklanabilmesine neden olan ve dolayısıyla yatırımcıların bilgiye erişimlerini sınırlandıran bilişsel bir kısıttır. Bu kısıt yatırımcıları sadece hakkında bilgi sahibi oldukları hisse senetlerini satın almaya yönlendirdiğinden hisse senedi fiyat hareketleri için bir sinyal olarak kullanılmaktadır. Ancak yatırımcı ilgisinin nasıl ölçüleceği konusunda farklı görüşler söz konusudur. Yatırımcı ilgisini dolaylı olarak ölçen yaklaşımlarda fiyat, likidite, getiri, reklam harcamaları gibi dolaylı temsilciler kullanılmakta, doğrudan ölçen yaklaşımlarda ise ya doğrudan yatırımcıya sorulmakta ya da yatırımcıların davranışları izlenmektedir. Özellikle bilgi ve iletişim teknolojilerindeki gelişmelerle birlikte sosyal medyanın yatırımcılar tarafından yatırım fikirlerine ulaşmak için yaygın bir şekilde kullanımı yatırımcı ilgisini doğrudan ölçmek için yeni araçlar sunmaktadır. Yatırımcıların bilgi arayışlarına dayanan bu araçlar arasında Google ve Baidu arama hacimleri, Wikipedia sayfalarının görüntülenme sayısı ve tweetler sayılabilir. Yatırımcı ilgisinin etkili olduğu bir diğer finansal dinamik ise ilk halka arzlardır. İlk halka arz IPO, bir firmanın hisselerinin birincil piyasada halka arz edilmesi sürecini ifade etmektedir. Dolayısıyla ilk halka arzda firmalar birincil piyasa yoluyla hisse senetlerini arz ederek sermaye elde ederler. İlk halka arzda hisse senedinin performansının öngörülmesi gerek firmalar gerek yatırımcılar açısından oldukça önemlidir. Bu araştırmada, yatırımcı ilgisinin temsilcileri olarak ilk halka arz öncesinde firma isimlerinin ve sembollerinin Googleda aranma hacimleri ile halka arz olacak şirket hakkında kısa bilgilerin verildiği YouTube içeriklerinin görüntülenme sayıları kullanılmakta ve bunların Borsa İstanbulda ilk halka arz performansı üzerindeki etkisi incelenmektedir. Çalışmada ayrıca yatırımcı ilgisinin firmalara özgü hangi niteliklerle ilgili olduğu da araştırılmaktadır. Başka bir ifadeyle halka arz büyüklüğü, hisse sayısı, halka arz fiyatı, halka arz nedeni, firmanın sektörü, finansal oranları gibi değişkenlerden hangilerinin yatırımcı ilgisini daha fazla çektiği de araştırılmaktadır. Araştırma sonucunda YouTube izlenme sayılarının Google arama hacimlerine kıyasla yatırımcı ilgisinin daha iyi bir temsilcisi olduğu bulgusuna ulaşılmıştır. Bu çalışma kapsamında, yatırımcı ilgisinin temsilcisi olarak kullanılan her iki araç da tam bilginin olmadığı ve bilişsel yeteneklerin kısıtlı olduğu sınırlı rasyonalite varsayımıyla örtüşmektedir. Ayrıca bulgular yatırımcıların firmanın temel değerine ve içsel büyüme potansiyeline önem verdiğine de işaret etmektedir. Bu çalışmanın amaçlarından bir diğeri de algoritmik finansal işlemler kapsamında, ilk halka arza ilişkin kararları verebilecek ya da yatırımcıların kararlarına destek olabilecek dar bir yapay zekâ geliştirmektedir. Yapay zekâ, bir bilgisayarın zeki canlılara benzer şekilde çeşitli faaliyetleri gerçekleştirme yeteneği olarak tanımlanmaktadır. Süper yapay zekâ, insanüstü yeteneklere sahip olan; genel yapay zekâ, bir insanın yapabileceği herhangi bir görevi yerine getirmek için tasarlanmış olan; dar yapay zekâ ise belirli bir görevi gerçekleştirmek için tasarlanmış yapay zekâyı ifade etmektedir. Dolayısıyla dar yapay zekâ kapsamında geliştirilen her bir model, esasında genel yapay zekâya yeni bir yetenek kazandırmaktadır. Bu bağlamda bu çalışmadan elde edilen bulgular, özellikle finansal yapay zekâ uygulamalarının geliştirilmesinde ve algoritmik finansal işlemler yapan botların eğitilmesinde ilk halka arz kapsamında önemli bir katkı sunmaktadır. Başka bir ifadeyle bu çalışma, pratikte çerçevesi halka arz performansları ile sınırlı olan finansal dar bir yapay zekâ ürünü ortaya koyarken teoride yatırımcı ilgisinin belirleyicilerini ortaya koymaktadır.

Makale detayları

Investor attention in Borsa Istanbul: analysis of the effect of Google Trends and Youtube views on initial public offering performances by Random Forest Method

Yazarlar:

Yunus Emre Akdoğan

Yayıncı:

Hitit University

Ana Başlıklar:

Davranışsal Finans | Finansal Öngörü ve Modelleme | Yatırımlar ve Portföy Yönetimi

Dergi:

Hitit Sosyal Bilimler Dergisi Vol.17 Issue 1, pp.70-90

Belge Tipi:

Makale

Yayın Tarihi:

4/2024

Belge Dili:

Türkçe

Document ID (Org No.):

2001315014

pISSN:

2757-7449

Initial Public Offering

Investor Attention

Random Forest

Google Trends

Youtubeİlk Halka Arz

Yatırımcı İlgisi

Rassal Orman

Google Trendler

Youtube

Sayı

Makale detayları

Yazarlar: Yunus Emre Akdoğan

Yayıncı: Hitit University

Ana Başlıklar: Davranışsal Finans | Finansal Öngörü ve Modelleme | Yatırımlar ve Portföy Yönetimi

Dergi: Hitit Sosyal Bilimler Dergisi Vol.17 Issue 1, pp.70-90

Belge Tipi: Makale

Yayın Tarihi: 4/2024

Belge Dili: Türkçe

Document ID (Org No.): 2001315014

pISSN: 2757-7449

Sayı

Yazarlar:

Yunus Emre Akdoğan

Yayıncı:

Hitit University

Ana Başlıklar:

Davranışsal Finans | Finansal Öngörü ve Modelleme | Yatırımlar ve Portföy Yönetimi

Dergi:

Hitit Sosyal Bilimler Dergisi Vol.17 Issue 1, pp.70-90

Belge Tipi:

Makale

Yayın Tarihi:

4/2024

Belge Dili:

Türkçe

Document ID (Org No.):

2001315014

pISSN:

2757-7449