The forex market is a complex, evolving, and a non-linear dynamical system, and its forecast is difficult due to high data intensity, noise/outliers, unstructured data and high degree of uncertainty. However, the exchange rate of a currency is often found surprisingly similar to the history or the variation of an alternative currency, which implies that correlation knowledge is valuable for forex market trend analysis. In this research, we propose a computational correlation analysis for the intelligent correlation extraction from all available economic data. The proposed correlation is a synthesis of channel and weighted Pearson's correlation, where the channel correlation traces the trend similarity of time series, and the weighted Pearson's correlation filters noise in correlation extraction. In the forex market analysis, we consider 3 particular aspects of correlation knowledge: (1) historical correlation, correlation to previous market data; (2) cross-currency correlation, correlation to relevant currencies, and (3) macro correlation, correlation to macroeconomic variables. While evaluating the validity of extracted correlation knowledge, we conduct a comparison of Support Vector Regression (SVR) against the correlation aided SVR (cSVR) for forex time series prediction, where correlation in addition to the observed forex time series data is used for the training of SVR. The experiments are carried out on 5 futures contracts (NZD/AUD, NZD/EUD, NZD/GBP, NZD/JPY and NZD/USD) within the period from January 2007 to December 2008. The comparison results show that the proposed correlation is computationally significant for forex market analysis in that the cSVR is performing consistently better than purely SVR on all 5 contracts exchange rate prediction, in terms of error functions MSE, RMSE, NMSE, MAE and MAPE. However, the cSVR prediction is found occasionally differing significantly from the actual price, which suggests that despite the significance of the proposed correlation, how to use correlation knowledge for market trend analysis remains a very challenging difficulty that prevents in practice further understanding of the forex market. In addition, the selection of macroeconomic factors and the determination of time period for analysis are two computationally essential points worth addressing further for future forex market correlation analysis.

The forex market is a complex, evolving, and a non-linear dynamical system, and its forecast is difficult due to high data intensity, noise/outliers, unstructured data and high degree of uncertainty. However, the exchange rate of a currency is often found surprisingly similar to the history or the variation of an alternative currency, which implies that correlation knowledge is valuable for forex market trend analysis. In this research, we propose a computational correlation analysis for the intelligent correlation extraction from all available economic data. The proposed correlation is a synthesis of channel and weighted Pearson's correlation, where the channel correlation traces the trend similarity of time series, and the weighted Pearson's correlation filters noise in correlation extraction. In the forex market analysis, we consider 3 particular aspects of correlation knowledge: (1) historical correlation, correlation to previous market data; (2) cross-currency correlation, correlation to relevant currencies, and (3) macro correlation, correlation to macroeconomic variables. While evaluating the validity of extracted correlation knowledge, we conduct a comparison of Support Vector Regression (SVR) against the correlation aided SVR (cSVR) for forex time series prediction, where correlation in addition to the observed forex time series data is used for the training of SVR. The experiments are carried out on 5 futures contracts (NZD/AUD, NZD/EUD, NZD/GBP, NZD/JPY and NZD/USD) within the period from January 2007 to December 2008. The comparison results show that the proposed correlation is computationally significant for forex market analysis in that the cSVR is performing consistently better than purely SVR on all 5 contracts exchange rate prediction, in terms of error functions MSE, RMSE, NMSE, MAE and MAPE. However, the cSVR prediction is found occasionally differing significantly from the actual price, which suggests that despite the significance of the proposed correlation, how to use correlation knowledge for market trend analysis remains a very challenging difficulty that prevents in practice further understanding of the forex market. In addition, the selection of macroeconomic factors and the determination of time period for analysis are two computationally essential points worth addressing further for future forex market correlation analysis.

In this thesis, we study the return and volatility spillovers between forex and stock markets in Scandinavian countries employing recently developed method- ology of spillover indices. Those measures are based on forecast error variance decomposition of generalized vector autoregressive (GVAR) model. This allows us to estimate both total and directional spillovers. Moreover, frequency connect- edness analysis is conducted by decomposing the spillover indices into frequency bands, corresponding to short-, medium- and long-run connectedness. We used daily data for major stock market indices and exchange rates of domestic cur- rency towards US dollar for Norway, Sweden, Denmark and Finland. Our data spans from February 2002 till July 2018 that covers turmoil periods of global fi- nancial crisis in 2007-2009, European sovereign debt crisis 2010-2013 and Brexit referendum in mid 2016. Our empirical analysis reveals that Norwegian financial markets do not contribute much to both return and volatility spillovers. On the other hand, euro and Danish FX market perform very similarly, by exhibiting the highest spillover contributions for both returns and volatility. Furthermore, distinct increasing trends in spillovers are revealed during the turmoil periods for most of the markets. From frequency...

Les marchés financiers sont sujets quotidiennement à la diffusion de statistiques économiques ainsi que leurs prévisions par des institutions publiques et privées. Ces annonces sont prévues ou non prévues. Les annonces prévues sont organisées selon un calendrier connu à l’avance par tous les opérateurs. Ces annonces telles que les statistiques d'activité, d’exportation ou de sentiments, sont publiées une fois par mois par des agences spécialisées telles que Bloomberg. La diffusion d’une statistique économique ou financière est toujours précédée par la publication de sa prévision calculée comme la médiane de toutes les prévisions individuelles fournies par les agents. Cette médiane est un proxy de la vision commune des opérateurs et aide à la construction d'une représentation collective de l'environnement économique. Le premier chapitre de ma thèse a pour objectif d'analyser l'hétérogénéité dans la prévision des annonces macroéconomiques est testée grâce à des données mensuelles de prévisions issues d'enquêtes conduites par Bloomberg, sur une série d'indicateurs macroéconomiques. S’ensuit alors une deuxième problématique. Quels sont aux yeux des investisseurs, les critères qui permettent de considérer qu’une annonce est plus importante qu’une autre ? L’analyse du processus par lequel une information est incorporée dans les cours, nous a éclairés sur l’existence d’une forte rotation dans les statistiques considérées comme importantes (Market Mover indicators). Le deuxième chapitre tente donc de répondre à cette problématique. Dans un dernier chapitre je m’interroge sur la dynamique des prix post-publications d’annonces macroéconomiques et financières. Des connections sont réalisées entre les Jumps sur les cours des actifs et les annonces macroéconomiques, financières mais aussi imprévues. / Financial markets are subjected daily to the diffusion of economic indicators and their forecasts by public institutions and even private ones. These annoncements can be scheduled or unscheduled. The scheduled announcements are organized according to a specific calendar and known in advance by all operators. These news such as activity indicators, credit, export or sentiments’ surveys, are published monthly or quarterly by specialized agencies to all operators in real time. Our thesis contributes to diferent literatures and aims to thoroughly analyze the three phases of the diffusion process of new information on financial markets : anticipation of the announcement before its publication, interest that arouse its publication and impact of its publication on market dynamics. The aim of the first chapter is to investigate heterogeneity in macroeconomic news forecasts using disaggregate data of monthly expectation surveys conducted by Bloomberg on macroeconomic indicators from January 1999 to February 2013. The second chapter examines the impact of surprises associated with monthly macroeconomic news releases on Treasury-bond returns, by paying particular attention to the moment at which the information is published in the month. In the third chapter we examine the intraday effects of surprises from scheduled and unscheduled announcements on six major exchange rate returns (jumps) using an extension of the standard Tobit model with heteroskedastic and asymmetric errors.