Using Sentiment Data to Build an Inflation-based Allocation Strategy

High inflation has returned to developed markets after three decades of stability. This makes portfolio construction and rebalancing challenging, because today’s investors have little experience navigating high inflation periods.

In our most recent paper, we describe how to build an asset allocation strategy that combines inflation states with a hedging component. To test our approach, we perform an out-of-sample backtest where we target US monthly CPI inflation to allocate capital among three asset classes: an equity index, a commodity future and a risk-free rate asset. Results show that:

• The proposed strategy can outperform the passive strategies (50/50, 60/40, S&P 500) when relying upon inflation trend forecasts delivered by a machine learning approach that incorporates RavenPack Analytics.
• Our approach delivers a real Sharpe ratio (net of inflation) of 1.04 and 0.72, and 1.22 and 0.88 in gross terms when targeting Headline and Core CPI, respectively.
• The strategy that depends on sentiment data outperforms the same strategy when inflation forecasts merely depend on observed inflation.

Watch the video summary

During high uncertainty periods, data-driven decisions are more important than ever. Download the white paper and see how sentiment insights can help exploit high uncertainty periods by predicting inflation regimes.