Is Exchange Rate Moody? Estimating the Influence of Market Sentiments With Google Trends

This paper proposes a novel method of exchange rate forecasting. We extend the present value model based on observable fundamentals by including three unobserved fundamentals: credit-market, financial-market, and price-market sentiments. We develop a method of sentiments extraction from Google Trends data on searched queries for different markets. Our method is based on evolutionary algorithms of variable selection and principal component analysis (PCA). Our results show that the extended vector autoregressive model (VAR) which includes markets' sentiment, shows better forecasting capabilities than the model based solely on fundamental variables or the random walk model (naïve forecast).

Abstract licence: CC BY

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.