Country News Sentiment Factors Predict Forex Prices

Executive Summary:

We develop a cross-over strategy in the FX market based on short to long-term news sentiment inflection points, but this time we extend the study beyond the EURUSD to other major currency pairs.

The indexes, which represent the world’s major economies, are proven to predict short-term price moves in major currency pairs for up to several hours after a sentiment inflection point.

Inflection Points in RavenPack Country Sentiment Indexes Predict Price Moves for Major Currencies. Trading an equal-weighted portfolio of 7 major currency pairs:

• a cross-over strategy between short (1-week) and long-term (3-month) U.S. sentiment, using a 10-hour holding period, generates an annualized Information Ratio of 1.1 .
• a cross-over strategy between short (1-day) and long-term (3-month) non-U.S. sentiment, with a 9-hour holding period, generates an annualized Information Ratio of 2.2 .
• a cross-over strategy between short (1-day) and long-term (3-month) for both U.S. and non-U.S. sentiment, using a 9-hour holding period, generates an annualized Information Ratio of 1.9 .

Introduction

In a previous study , we introduced a short-term foreign exchange trading strategy that considered the principles of technical analysis to create buy or sell signals based on data derived from fundamental news. More specifically, a sell signal was generated on EURUSD whenever short-term U.S. sentiment became more bullish over the short-term as compared to the longer-term sentiment trend.

Similarly, a buy signal was generated when the opposite occurred. In this study, we move beyond U.S. sentiment and extend the cross-over strategy to 7 major currency pairs including GBPUSD, EURUSD, AUDUSD, NZDUSD, USDJPY, USDCAD, and USDCHF. Unlike the previous study that only examines the price impact of U.S. macro news, we also investigate the impact of non-U.S. news including the U.K., Eurozone, Australia, New Zealand, Japan, Canada, and Switzerland. The back-testing period is further extended to capture the second half of 2012, hence including the period January 2010 through December 2012.

To measure macro-level sentiment, we rely on the global macro package of RavenPack News Analytics that tracks and analyzes information on over 138,000 key geographical entities, more than 2,200 government organizations, and all major currencies and traded commodities. For any news record that can be matched with an event category, RavenPack generates an Event Sentiment Score (ESS) signaling its potential impact on any given economy or financial market (see Appendix A for details). RavenPack covers over 1,200 events of which 895 relate to unscheduled news such as political events, natural disasters, war and conflict, etc., as well as scheduled news such as the release of important macroeconomic indicators.

Similar to our previous study, the strategy detects and trades based on sentiment inflection points. That is, we take a "bullish” view on a currency when the short-term sentiment trend crosses above the long-term sentiment trend and a bearish view when the short-term sentiment trend crosses below the long-term sentiment trend.

In Section 2, we introduce the methodology to construct the macro sentiment index for a given economy. In Section 3, we provide a brief review of the cross-over strategy. In Sections 4 and 5, we investigate the price impact of U.S. and non-U.S. sentiment across all 7 currency pairs, respectively. In Section 6, we test the combined effects of U.S. and non-U.S. sentiment, while in Section 7, we separately examine the predictive power of bearish and bullish trading signals. Finally, in Section 8 we present our conclusions.