Online big data-driven oil consumption forecasting with Google trends

Yu, Lean; Zhao, Yue; Tang, Ling; Yang, Zebin

doi:10.1016/j.ijforecast.2017.11.005

Cited by 185 publications

(85 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…To ensure that the Google Search volume (GSV) of a word is a true representative of at least some oil‐related events, we have to verify that the GSV led to changes in volatility. A Granger causality test is conducted to identify which words do not contain information that help predict crude oil volatility (Yu, Zhao, Tang, & Yang, ). We select the 17 significant keywords and use these keywords to develop an event‐triggered predictor (Table reports the F ‐values of hypothesis testing for Granger causality from words to volatility).…”

Section: Constructing the Predictorsmentioning

confidence: 99%

Oil financialization and volatility forecast: Evidence from multidimensional predictors

Pan

2019

Journal of Forecasting

129

View full text Add to dashboard Cite

Using the generalized dynamic factor model, this study constructs three predictors of crude oil price volatility: a fundamental (physical) predictor, a financial predictor, and a macroeconomic uncertainty predictor. Moreover, an event‐triggered predictor is constructed using data extracted from Google Trends. We construct GARCH‐MIDAS (generalized autoregressive conditional heteroskedasticity–mixed‐data sampling) models combining realized volatility with the predictors to predict oil price volatility at different forecasting horizons. We then identify the predictive power of the realized volatility and the predictors by the model confidence set (MCS) test. The findings show that, among the four indexes, the financial predictor has the most predictive power for crude oil volatility, which provides strong evidence that financialization has been the key determinant of crude oil price behavior since the 2008 global financial crisis. In addition, the fundamental predictor, followed by the financial predictor, effectively forecasts crude oil price volatility in the long‐run forecasting horizons. Our findings indicate that the different predictors can provide distinct predictive information at the different horizons given the specific market situation. These findings have useful implications for market traders in terms of managing crude oil price risk.

show abstract

Section: Constructing the Predictorsmentioning

confidence: 99%

Oil financialization and volatility forecast: Evidence from multidimensional predictors

Pan

2019

Journal of Forecasting

129

View full text Add to dashboard Cite

show abstract

“…In the emerging digital society, search engines have become a significant tool for acquiring the latest relevant news about a target term. Google search engine is essential for a wide range of online searching tasks [35]. In particular, Google Trends is a free and accessible online portal that analyzes a portion of billions of daily Google searches, generating data on geographical and temporal patterns according to specified keywords [36].…”

Section: Google Trends Datamentioning

confidence: 99%

Exploring crime patterns in Mexico City

Piña-García

Ramírez

2019

J Big Data

View full text Add to dashboard Cite

show abstract

“…Varian and Choi (2012), Vozlyublennaia (2014), Bijl et al(2016) and D'Amuri and Marcucci (2017) do not perform any differencing or detrending of the series, which suggests that the Google Trends they use are stationary. Yu et al (2018) use an ADF test on three Google Trends queries: "oil inventory", "oil consumption" and "oil price" and find evidence of stationarity at the 5% level (10% level) in "oil inventory" ("oil consumption"), but are not able to reject the null of a unit root for "oil price". Da et al (2014) take log-differences on the series.…”

Section: Datamentioning

confidence: 99%

In Search of a Job: Forecasting Employment Growth in the US Using Google Trends

Schütte

2018

SSRN Journal

View full text Add to dashboard Cite

We show that Google search activity on relevant terms is a strong out-of-sample predictor of future employment growth in the US and that it greatly outperforms benchmark predictive models based on macroeconomic, financial, and sentiment variables. Using a subset of ten keywords, we construct a panel with 211 variables using Google's own algorithms to find related search queries. We use Elastic Net variable selection in combination with Partial Least Squares to extract the most important information from a large set of search terms. Our forecasting model, which can be constructed in real time and is free from revisions, delivers an out-of-sample R 2 statistic of 65% to 88% for horizons between one month and one year ahead over the period 2008-2017, which compares to between roughly 30% and 60% for the benchmark models.

show abstract

Online big data-driven oil consumption forecasting with Google trends

Cited by 185 publications

References 51 publications

Oil financialization and volatility forecast: Evidence from multidimensional predictors

Oil financialization and volatility forecast: Evidence from multidimensional predictors

Exploring crime patterns in Mexico City

In Search of a Job: Forecasting Employment Growth in the US Using Google Trends

Contact Info

Product

Resources

About