Dissertation: Small investors' Google searches are linked to subsequent stock market returns
– For example an increase in the volume of Google searches related to a decline in share prices predicts negative stock market returns for the following week, says Klemola, who will be defending his doctoral thesis at the University of Vaasa on Thursday 21 November.
Therefore, it pays to be cautious if terms such as “bear market,” reflecting negative market sentiment, suddenly become more common in Google searches.
The study did not, on the other hand, find as strong a relationship between searches related to an upward market movement and stock market returns for the forthcoming week as with searches related to market decline.
It can be assumed that investors will more readily notice news related to market decline in the media or public debate than positive news about the markets. Investors are thus more likely to end up conducting Google searches that are related to negative market development. They will then form their investor sentiment, or view, on future market development based of these searches.
The smaller the company, the bigger the “Google effect”
The more uncertainty there is in the pricing of a share, the bigger the impact any changes in investors' Google search volumes tend to have on future profits from these shares. According to Klemola's study, this is the reason why, e.g., small companies and value companies are particularly sensitive to changes in Google searches related to market downturn or upturn.
Conventional finance theory assumes that investors only make calculated and carefully considered investment decisions, i.e., they act rationally. Behavioural finance theory, on the other hand, assumes that at least some investors allow their personal feelings affect their investment decisions, i.e., they act irrationally.
– In general, small investors can be considered to represent irrationally behaving investors. If rational investors do not wish to - or are not in a position to - fully counter the irrational investors' trading on the stock exchange, the irrationally behaving investors will be able to influence share prices. In such a case, the share price will deviate from its actual fact-based value - a situation which may even lead to a price bubble, says Klemola.
A modern alternative for traditional sentiment measurement
According to Klemola, Google searches can be utilised to measure small investors' market sentiment and expectations in an alternative and a modern way. Search-based sentiment measurement also has other advantages over more traditional sentiment measurements.
– Many current sentiment measurements are based on information derived from market prices, such VIX, known as the investors' fear factor, which uses option prices to measure investors' expectations about future market volatility. The challenge in such sentiment measurements is that they can be affected by various other economic factors. Therefore they do not entirely reflect true investor sentiment, says Klemola.
According to Klemola, investor sentiment can also be measured directly through various surveys.
– The problem with surveys is that they are most likely not conducted frequently enough, perhaps only once a week or once a month. Additionally, people may not always answer survey questions truthfully. Google searches reflect people's real intentions better.
For example, people may not always say they are looking for a new job but they might still be actively searching for open positions online. Another good example of utilising Google searches is the way people are likely to start searching for information about how to apply for unemployment benefits if they know they are going to be made redundant. This is immediately reflected in increased search volumes, whereas official unemployment statistics are updated with a delay.
Google search volumes in Finland are insufficient for the assessment of sentiment
Klemola's study analyses the linkage between investor sentiment revealed by Google searches and future U.S. stock market returns. Thus far, the volume of Google searches in Finland has not reached the required levels to make a similar study on the Finnish stock market reliable.
In his doctoral thesis, Klemola has also studied the pricing of power derivatives in the Nordic electricity exchange. The study showed evidence of seasonality in option pricing. Seasonal overpricing or underpricing of options may indicate that market participants are overconfident and place too much faith in their own forecasts.
The public examination of M.Sc. Antti Klemola’s doctoral dissertation ” Essays on irrational investors’ behavioral biases and pricing efficiency ” will be held on Thursday 21 November at noon in K218 Nissi auditorium, Tritonia (University of Vaasa). The field of dissertation is Accounting and Finance.
Professor Eero Pätäri (Lappeenranta University of Technology) will act as opponent and professor Jussi Nikkinen as custos. The examination will be held in Finnish.
Antti Klemola, tel +358 44 345 6544, email: antti.klemola(at)uwasa.fi
Klemola, Antti (2019) Essays on irrational investors’ behavioral biases and pricing efficiency. Acta Wasaensia 432, Doctoral thesis, University of Vaasa, Vaasan yliopisto.
Publication pdf: http://urn.fi/URN:ISBN:978-952-476-886-3