In the fast-moving world of financial markets, macroeconomic data is a rich but underutilized source of systematic trading signals. One promising development in this space is the use of macro information state changes—timely updates to public economic indicators—as inputs for high-frequency strategies. These point-in-time updates capture shifts in public economic knowledge and can be used to generate daily or weekly trading signals that respond quickly to market-relevant developments.
This post explores how changes in macro information states can be standardized, aggregated, and integrated into systematic fixed-income strategies, with a focus on interest rate swaps across developed and emerging markets. The results show consistent predictive power and impressive simulated performance, especially during periods of rapid economic change.
What Are Macro Information State Changes?
A macro information state is the latest publicly available value of an economic indicator, recorded at the time it is released. For instance, Japan’s core inflation reading reported on September 18, 2024, for July 2024, is a new information state. A change in this value the next day—whether due to new data or revisions—constitutes an information state change.
These updates are not necessarily surprises. They may be partially expected, but their actual release shifts the market’s knowledge base. Because most economic indicators are revised or updated frequently and reflect multiple economic processes, information state changes happen regularly—often daily. Unlike traditional macro data, which is typically analyzed monthly or quarterly, these point-in-time updates offer higher-frequency insight.
Why Information State Changes Matter for Trading
There are two compelling reasons to use macro information updates in trading systems:
- Short-Term Signals
Although some economic changes are noise in the eyes of economists, they can still shape market expectations and price action in the short term. Information state changes capture these fluctuations and allow traders to respond quickly to shifts in perception—even when the broader trend remains unchanged. - Tail Risk and Trend Reinforcement
During economic shocks or escalatory phases—like recessions or inflation surges—information state changes tend to accumulate in one direction. Trading strategies based on these changes can position effectively during such times, providing a layer of protection against systemic risks.
Historically, the challenge has been data access. But platforms like JPMaQS now provide point-in-time data across major economic indicators, allowing for robust backtesting and real-time signal generation.
Constructing Macro Information Change Signals
To evaluate the use of information state changes in systematic trading, a study was conducted across 22 developed and emerging markets using five macroeconomic themes:
- Growth: Real GDP trends, industrial production, and retail activity
- Sentiment: Business and consumer confidence surveys
- Labor: Employment and unemployment data
- Inflation: Headline, core, and producer price indices
- Financing: Credit growth, money supply, and external balance metrics
Each data point is normalized to create comparability across units, frequencies, and countries. A transformation called Normalized Information Change in Annualized units (NICA) is applied, adjusting for expected volatility and data frequency. This ensures that changes in a quarterly GDP release can be directly compared with weekly sentiment surveys.
Aggregating Signals Across Time and Indicators
Information doesn’t move markets all at once. It trickles in and gets absorbed over days or weeks. To model this, short-term exponential moving averages (with a 3-day half-life) are applied to smooth recent updates.
Additionally, signals are aggregated in four ways:
- Across Indicators within each macro category (e.g., averaging growth-related changes),
- Across Categories using unweighted averages of group signals to form a composite,
- Across Countries by weighting changes according to GDP share in USD terms,
- Across Local and Global Levels by combining local signals with global trends.
These layers of aggregation create signals that capture broad economic shifts while minimizing noise. The result is a set of robust indicators with strong autocorrelation and rapid responsiveness.
Backtesting Results: Risk-Adjusted Performance and Accuracy
A 25-year backtest (2000–2024) across all 22 markets showed clear predictive value. Weekly information state changes were significantly correlated with subsequent weekly bond returns. The composite signal, as well as the individual macro categories, all demonstrated forward-looking relationships, with growth and inflation leading the way.
- Sharpe Ratio: The composite signal produced a Sharpe of 1.4 for a market-neutral strategy.
- Sortino Ratio: The downside risk-adjusted return hit 2.2.
- Seasonality: Value generation was stable across time, with only brief underperformance during 2009–2010.
- Drawdowns: Contained, especially when compared to standard macro signals.
Daily prediction accuracy exceeded 50% for most macro themes, and adding a global overlay further enhanced precision. Local-global combined signals achieved a Sharpe of 1.5 and a Sortino of 2.5—beating even strong local-only signals.
Key Takeaways
- Macro updates can drive short-term signals: Traders often underestimate just how frequently public macroeconomic data shifts. These changes offer exploitable patterns for high-frequency or short-horizon strategies.
- Normalization and aggregation are crucial: Without transforming the data to a common scale and format, real-time macro trading is impractical. The NICA framework solves this challenge.
- Broad applicability: Signals based on inflation, growth, labor, sentiment, and financial conditions all contributed to profitable outcomes. Combining them offers stronger, more consistent results.
- Outperformance in volatile regimes: Strategies built on macro information updates excelled during crisis periods, including the 2008 recession and the COVID-19 shock.
Macro information state changes represent a bridge between traditional economic insight and systematic trading discipline. With proper normalization and aggregation, they provide a timely, data-driven lens into market expectations. As more asset managers embrace this approach, macro-based signals are poised to play a larger role in shaping high-frequency and short-duration portfolios.