Historical research on initial public offerings (IPOs) presents strong evidence of underpricing. This study investigates if there is a relationship between underpricing of IPOs and pre-IPO research and development (R&D) activities within a company. According to the literature, R&D activities have characteristics of information asymmetry and uncertainty, which can increase underpricing. This study’s sample consists of 231 Swedish companies listed on Nasdaq Stockholm and Nasdaq First North between January 2010 and December 2020. Sweden has a strong association with innovation activities such as R&D, and the country’s IPO market has snowballed in recent years, making it a suitable context for the study. To investigate the relationship between underpricing and R&D activities, the study uses an OLS regression. The findings indicate that R&D positively affects underpricing, which is in line with previous studies on other markets. In addition, the study finds evidence of Firm Size, Offer Size, Shares Offered, VCPE backed, and Firm Leverage related to underpricing.

Background: The Covid-19 pandemic has been affecting the world in some or another way both when it comes to individuals and the adaption to a new lifestyle as well as economic and financial impacts on the global markets. The outbreak started off in China in the late 2019 and spread across every country in the world within three to four months. Living and working standards has since then changed due to restrictions with more people working from home and with impacts on the daily business of companies. The attention is brought upon the real estate sector because of the uncertainty followed by the new norms in societies which makes it an interesting sector on which to study the financial impact. Purpose: This paper aims to investigate how 13 chosen dates regarding the Covid-19 pandemic from announcements of the Swedish Prime Minister, Stefan Löfven and The Public Health Authority has affected the real estate sector in Sweden by using an event study approach. The dates were divided into three categories namely restrictions, work from home and financial support to see whether these specific announcements have a financial impact on Swedish real estate companies. Method: For this study, a deductive approach was applied along with the philosophy of positivism since this is most suitable for quantitative methods.  The event study methodology was used in order to observe the data on how the stock market in Sweden reacts on the impact of the outbreak concerning Covid-19 and how this might affect the real estate sector with the selected dates from the announcements. Conclusion: The findings from this study indicated that there were two significant results when calculating the abnormal returns for each individual company over all event dates with all others being not significant, since all other t-tests that were applied did not show strong enough results to provide significance concerning the real estate sector and the effects of Covid-19. Although, one should keep in mind that this is a short-term analysis and that in the long-term these results might differentiate, therefore, this thesis could be a guideline for future research when more data and information concerning the pandemic is available.

Researchers have dedicated considerable work to explaining components to excess stock market returns. Recently, Atanasov et al. (2020) managed to explain some of this variance in the US stock markets with a cyclical consumption variable. We have applied their model into the Nordic countries and compared it to a second model containing additional control variables. From the analysis, we find that cyclical consumption is able to explain excess stock market returns across five different h-quarter ahead excess returns. However, results are not consistent across countries. The extended model improves the explanatory capabilities of the model only up to two-year ahead excess returns. The cyclical consumption measure is also able to predict excess returns better than an historical average model. The findings in this paper are robust to out-of-sample predictability analysis and to a different measure of consumption and returns.

We consider the problems commonly encountered in asset management such as optimal execution, portfolio construction, and trading strategy implementation. These problems are generally difficult in practice, in large part due to the uncertainties in financial markets. In this thesis, we develop data-driven approaches via machine learning to better address these problems and improve decision making in financial markets. Machine learning refers to a class of statistical methods that capture patterns in data. Conventional methods, such as regression, have been widely used in finance for many decades. In some cases, these methods have become important building blocks for many fundamental theories in empirical financial studies. However, newer methods such as tree-based models and neural networks remain elusive in financial literature, and their usabilities in finance are still poorly understood. The objective of this thesis is to understand the various tradeoffs these newer machine learning methods bring, and to what extent they can improve a market participant’s utility. In the first part of this thesis, we consider the decision between the use of market orders and limit orders. This is an important question in practical optimal trading problems. A key ingredient in making this decision is understanding the uncertainty of the execution of a limit order, that is, the fill probability or the probability that an order will be executed within a certain time horizon. Equivalently, one can estimate the distribution of the time-to-fill. We propose a data-driven approach based on a recurrent neural network to estimate the distribution of time-to-fill for a limit order conditional on the current market conditions. Using a historical data set, we demonstrate the superiority of this approach to several benchmark techniques. This approach also leads to significant cost reduction while implementing a trading strategy in a prototypical trading problem. In the second part of the thesis, we formulate a high-frequency optimal execution problem as an optimal stopping problem. Through reinforcement learning, we develop a data-driven approach that incorporates price predictabilities and limit order book dynamics. A deep neural network is used to represent continuation values. Our approach outperforms benchmark methods including a supervised learning method based on price prediction. With a historic NASDAQ ITCH data set, we empirically demonstrate a significant cost reduction. Various tradeoffs between Temporal Difference learning and Monte Carlo method are also discussed. Another interesting insight is the existence of a certain universality across stocks — the patterns learned from trading one stock can be generalized to another stock. In the last part of the thesis, we consider the problem of estimating the covariance matrix of high-dimensional asset return. One of the conventional methods is through the use of linear factor models and their principal component analysis estimation. In this chapter, we generalize linear factor models to a general framework of nonlinear factor models using variational autoencoders. We show that linear factor models are equivalent to a class of linear variational autoencoders. Further- more, nonlinear variational autoencoders can be viewed as an extension to linear factor models by relaxing the linearity assumption. An application of covariance estimation is to construct minimum variance portfolio. Through numerical experiments, we demonstrate that variational autoencoder improves upon linear factor models and leads to a more superior minimum variance portfolio.

This paper explores the stock market as a source of funding for innovation by looking at the ability of the stock market to attract, identify and channel funds to innovative firms. We analysed 541 IPOs on the Swedish stock market between the years 2000-2015, using patent applications as a proxy for innovation. Results from an event study and regressions using two control groups show that firms find the stock market an attractive source of funding for innovation and that going public helps firms overcome liquidity restraints. By looking at the long- and short-term performance, measured by stock prices, of innovative firms by conducting OLS regressions, our results suggest; one, that there is an initial demand for innovative companies undergoing an IPO in comparison to non-innovative firms. And two, that investors are able to predict future innovativeness to some extent, but that they have some difficulties in anticipating future performance of innovative firms.

A new paradigm in investing has been created where people have easier access than ever to invest in the stock market from the convenience of their phones. Through zero-commission trading apps, like Robinhood, less starting capital is required. This research is used to investigate the relationship between the frequency of social media mentions on Twitter and a particular stock’s volatility. This will be done using the qualitative data analyzing tool AtlasTi to calculate the frequency in which a particular stock ticker is mentioned on Twitter during trading hours. The volatility of the stock will be calculated using data from Yahoo! Finance. Using a panel data analysis, our evaluation reveals that there is a statistically significant relationship between the number of Tweets both one and two days before and the volatility of the stock based on percent change. Additionally, there is a statistically significant relationship between the number of Tweets the day before and the volatility of the stock based on volume traded. It is intended that our research will aid future investors when making decisions on how to invest in assets heavily mentioned on social media.

No description available.

This thesis deals with comparing two approaches to modelling and predicting time series: a traditional one (the ARIMAX model) and a modern one (gradiently boosted decision trees within the framework of the XGBoost library). In the first part of the thesis we introduce the theoretical framework of supervised learning, the ARIMAX model and gradient boosting in the context of decision trees. In the second part we fit the ARIMAX and XGBoost models which both predict a specific time series, the daily volume of the S&P 500 index, which is a crucial task in many branches. After that we compare the results of the two approaches, we describe the advantages of the XGBoost model, which presumably lead to its better results in this specific simulation study and we show the importance of hyperparameter optimization. Afterwards, we compare the practicality of the methods, especially in regards to their computational demands. In the last part of the thesis, a hybrid model theory is derived and algorithms to get the optimal hybrid model are proposed. These algorithms are then used for the mentioned prediction problem. The optimal hybrid model combines ARIMAX and XGBoost models and performs better than each of the individual models on its own. 1

This project investigates the relationship between social interactions among marketparticipants and the emergence of cyclical trends in stock markets. Two models aredeveloped to capture the interaction dynamics among individuals within the market:a basic model and a model with a price mechanism. By conducting numerical examplesand systematic simulations, we examine the behavior of these models. Ourfindings reveal that the basic model does not generate cyclical trends in the stockmarket. However, the model incorporating the price mechanism demonstrates theability to create such trends.

This dissertation examines the reaction of common stock prices to changes in bond ratings by Moody's Bond Service. The question is whether an announcement of a re-rating by Moody's is new information. There are only two studies of stock price reaction to bond changes and the results are conflicting. Pinches and Singleton (1978) [PS] concluded that any reaction comes well before the re-rating. Griffin and Sanvicente (1982) [GS] found that their portfolio test indicated that rating changes do convey new information. This was particularly true for downgradings. Both studies used monthly data and neither performed a statistical testing of residual reversals. PS provided a graph of the residuals which indicated the presence of a reversal trend. GS provided no information on this topic. This study, using daily data and the cumulative prediction error technique, finds that bond re-ratings offer new information. The results indicate that the market only partially anticipates the bond change. For the downgrades, the excess return on the announcement day is .6% which is statistically significant. The residuals reverse after the announcement day, but are not statistically significant. The upgrades do not have a significant reaction on the announcement day, but have a statistically significant negative reaction from day 1 to 10. The cumulative residual for days 1 to 10 is -2.8% with a test statistic of -3.85. This study finds as PS that there is some anticipation for both upgrades and downgrades. It extends their work by statistically testings the reversals after the announcement date and by testing the announcement day effect. There is significant abnormal return for the downgrades on the announcement day and the upgrades have a significant reversal in their residuals from day 1 to 10. This provides both support and extension of Griffin and Sanvicente's results and suggests that Moody's is offering the market new information.