Cette étude tente d’améliorer le modèle standard de prévision de l’indice de volatilité VIX à l’aide du Search Volume Index (SVI) rapporté par Google Trends à l’échelle journalière. La volatilité des marchés étant basée sur la peur et la recherche de gains économiques des agents de marché, Google Trends permet un accès direct et virtuellement immédiat aux désirs et aux inquiétudes de ceux-ci. Durant la période de janvier 2010 à décembre 2016, les séries de 15 termes de recherche liés à l’économie sont utilisées pour tenter d’améliorer les modèles ARIMA et SARIMA recommandés dans la prévision du VIX. Contrairement aux études similaires effectuées à l’échelle hebdomadaire, les résultats démontrent qu’il n’y a pas d’amélioration des prévisions causée par l’addition des données Google Trends aux modèles.

Fear of crime is an important topic in research as well as in public opinion. However, data on fear of crime is limited and difficult to collect, being heavily reliant on surveys with different methods of operationalization yielding different results. This paper aims to investigate if an alternative method can be used to estimate fear of crime. Using a large representative unique data on fear of crime from Google Trends, I analyze if fear of crime can be estimated in Sweden during years 2011-2019, using The Swedish Crime Survey as benchmark. The results show that the method is accurate for country-level and the most populated regions of Sweden, but less so for the lesser populated regions. This method can be used to estimate fear of crime in a time- and money-efficient way, producing daily estimates at little to no cost.

This diploma thesis is based on the theory of behavioral attention and examines the effect of the search for negative words in conjunction with the name of the bank on the price and on the yield of the shares of these banks. As a sample, 12 global, publicly traded and significant banks were selected. In this work, the behavioral attention is identified as the level of search on Google. The panel regression with random effects is used in the work, and Bayesian Model Averaging is used to identify suitable variables. The data proves the effect of negative behavioral attention, when an increased level of attention diminishes yield and share price. The results are then subjected to a robustness analysis where the impact of behavioral attention is examined before, during, and after the financial crisis. Furthermore, the effect of regulation and the level of behavioral attention itself is examined. The diploma thesis corresponds to the knowledge of behavioral economics and confirms a certain irrational behavior of investors on the market.

本文主要是透過 Google trends 網站提供的關鍵字搜尋量資料， 探討網路資料是否能夠提供通貨膨脹率的即時資訊。 透過美國消費者物價指數的組成細項作為依據，蒐集美國2004年1月至2015年12月的 Google trends 關鍵字變數，並藉由最小絕對壓縮挑選機制(Least absolute shrinkage and selection operator)、 彈性網絡(Elastic Net)以及主成分分析法(Principal component analysis)等等變數挑選機制，有效地整合大量的關鍵字資料。實證結果發現，透過適當變數挑選後的 Google trends 關鍵字變數確實可改善美國通貨膨脹率的即時預測表現，並為美國通貨膨脹率提供額外有效的資訊。此外，我們透過台灣的關鍵字資料檢驗，也確認Google trends 關鍵字資料可以幫助台灣通貨膨脹率的即時預測。

Den här studien utvärderar potentialen i söktermsdata från Google som ledande indikator för priser på bostadsmarknaden i Stockholm. Det prediktiva innehållet i söktermsdata från Google Trends jämförs mot en mer klassisk prognosmodell byggd på makroekonomiska variabler. Genom att mäta avvikelsen i en pseudo-prognos redovisas respektive datakällas förmåga till riktiga prognoser. Den huvudsakliga slutsatsen är att det finns resultat som styrker tesen om prediktivt innehåll i Googledata, framförallt för prognoser med horisonter upp till sex månader. Genom att använda Googledata skapas prognoser som har en mindre avvikelse från den faktiska tidsserien är vad modellen byggd på makroekonomiska variabler kan leverera. Resultatet visar på användbarheten i söktermsdata från Google som ledande indikator för priser på bostadsmarknaden i Stockholm. / This study evaluates the potential in using Google search term data as a leading indicator of prices on the real estate market in Stockholm. The predictive content of the search term data from Google Trends is contrasted against a more classic forecasting model using macroeconomic variables. The ability of each data source to generate powerful forecasts is demonstrated by measuring the deviation in a pseudo-forecast. The main finding is that the results support the hypothesis on predictive content in Google data, mainly forecasts with up to six months’ horizon. By using Google data, forecasts can be made with less deviation from the actual time series than forecasts built on macroeconomic variables. The results point to the usability of search term data from Google as a leading indicator for prices on the real estate market in Stockholm.

In this thesis, the usefulness of search engine data to nowcast the unemployment rate of Sweden is evaluated. Four different indices from Google Trends based on keywords related  to unemployment are used in the analysis and six different regARIMA models are  estimated and evaluated. The results indicate that the fit is improved for models when data  from Google Trends is included. To evaluate the nowcast ability of models, one-step-ahead  predictions are calculated. Although the prediction error is lower for the models with data  from Google Trends, Diebold-Mariano tests do not indicate that the predictions are  significantly better compared topredictions from a model without data from Google Trends.  It is therefore concluded that one cannot state that data from Google Trends improves  nowcasts of the unemployment rate of Sweden. Additionally, predictions are calculated for  longer forecast horizons. This analysis indicates that Google search data could be useful to  forecast the unemployment rate of longerforecast horizons.

No description available.

Yes / Purpose: We develop new quantitative methods to estimate the level of speculation and long-term sustainability of Bitcoin and Blockchain. Design/Methodology/Approach: We explore the practical application of speculative bubble models to cryptocurrencies. We then show how the approach can be extended to provide estimated brand values using data from Google Trends. Findings: We confirm previous findings of speculative bubbles in cryptocurrency markets. Relatedly, Google searches for cryptocurrencies seem to be primarily driven by recent price rises. Overall results are sufficient to question the long-term sustainability of Bitcoin with the suggestion that Ethereum, Bitcoin Cash and Ripple may all enjoy technical advantages relative to Bitcoin. Our results also demonstrate that Blockchain has a distinct value and identity beyond cryptocurrencies - providing foundational support for the second generation of academic work on Blockchain. However, a relatively low estimated long-term growth rate suggests that the benefi ts of Blockchain may take a long time to be fully realised. Originality/value: We contribute to an emerging academic literature on Blockchain and to a more established literature exploring the use of Google data within business analytics. Our original contribution is to quantify the business value of Blockchain and related technologies using Google Trends.

<p>The purpose of this Bachelor thesis is to analyze if there is a correlation between stock prices and the amount of searches of the companies names on Google. The theories used in the study were Capital Asset Pricing Model (CAPM) and Efficient Market Hypothesis (EMH). Regressions analysis is used as the statistical method to see if there is a significant correlation between the stock prices and the amout of searches of the company name on Google. The data used were the rate of return of three companies (ABB, Oriflame and Sandvik) on the Nasdaq OMX Nordic stock market, the rate of return of the Nasdaq OMX Nordic stock market index (OMX Stockholm_PI) and the Google search frequency from Google Trends on each company. The result showed no significance and the conclusion of the thesis is that there is no significant correlation between the three studied companies and their search frequency on the search engine Google.</p> / <p><strong>Syfte</strong>: Syftet med uppsatsen är att undersöka ifall det finns ett samband mellan företags aktiekurser och sökfrekvens på företagets namn på söktjänsten Google.</p><p><strong>Data: </strong>Daglig avkastning på ABB:s, Oriflames och Sandviks aktier, Aktieindex samt Googels sökfrekvens.</p><p><strong>Teorier: </strong>Capital Asset Pricing Model (CAPM), Effektiva marknadshypotesen (EMH)</p><p><strong>Slutsats: </strong>Det råder inget signifikant samband mellan de undersökta företagens aktiekurser och deras företagsnamns sökfrekvens på söktjänsten Google.</p>

The purpose of this Bachelor thesis is to analyze if there is a correlation between stock prices and the amount of searches of the companies names on Google. The theories used in the study were Capital Asset Pricing Model (CAPM) and Efficient Market Hypothesis (EMH). Regressions analysis is used as the statistical method to see if there is a significant correlation between the stock prices and the amout of searches of the company name on Google. The data used were the rate of return of three companies (ABB, Oriflame and Sandvik) on the Nasdaq OMX Nordic stock market, the rate of return of the Nasdaq OMX Nordic stock market index (OMX Stockholm_PI) and the Google search frequency from Google Trends on each company. The result showed no significance and the conclusion of the thesis is that there is no significant correlation between the three studied companies and their search frequency on the search engine Google. / Syfte: Syftet med uppsatsen är att undersöka ifall det finns ett samband mellan företags aktiekurser och sökfrekvens på företagets namn på söktjänsten Google. Data: Daglig avkastning på ABB:s, Oriflames och Sandviks aktier, Aktieindex samt Googels sökfrekvens. Teorier: Capital Asset Pricing Model (CAPM), Effektiva marknadshypotesen (EMH) Slutsats: Det råder inget signifikant samband mellan de undersökta företagens aktiekurser och deras företagsnamns sökfrekvens på söktjänsten Google.