While economic information undeniably wields a significant and widespread influence on financial markets, the systematic incorporation of macroeconomic data into trading strategies has thus far been limited. This reflects skepticism towards economic theory and serious data problems, such as revisions, distortions, calendar effects, and, generally, the lack of point-in-time formats. However, the emergence of industry-wide quantamental indicators and the rise of statistical learning methods in financial markets make macroeconomic information more practical and powerful. Successful demonstrations of statistical learning and macro-quantamental indicators have been achieved, with various machine learning techniques poised to further improve the utilization of economic information.
FX trading signals with regression-based learning
Regression-based statistical learning helps build trading signals from multiple candidate constituents. The method optimizes models and hyperparameters sequentially and produces point-in-time signals for backtesting and live trading. This post applies regression-based learning to macro trading factors for developed market FX trading, using a novel cross-validation method for expanding panel data. Sequentially optimized models consider nine theoretically valid macro trend indicators to predict FX forward returns. The learning process has delivered significant predictors of returns and consistent positive PnL generation for over 20 years. The most important macro-FX signals, in the long run, have been relative labor market trends, manufacturing business sentiment changes, relative inflation expectations, and terms of trade dynamics.
Macro trends and equity allocation: a brief introduction
Macroeconomic trends affect stocks differently, depending on their lines of business and their home markets. Hence, point-in-time macro trend indicators can support two types of investment decisions: allocation across sectors within the same country and allocation across countries within the same sector. Panel analysis for 11 sectors and 12 countries over the last 25 years reveals examples for both. Across sectors, export growth, services business sentiment, and consumer confidence have predicted the outperformance of energy stocks, services stocks, and real estate stocks, respectively. Across countries, relative export growth, manufacturing sentiment changes, and financial conditions have predicted the outperformance of local stocks versus foreign ones for the overall market and within sectors.
Understanding dollar shortages and related market dynamics
A dollar shortage is a state of FX and rates markets where covered interest rate parity between the U.S. and another currency area would result in excess dollar demand. Covered interest rate parity is the equality for short-term interest rate differentials and FX forward implied carry. Since the great financial crisis, arbitrage between onshore and offshore dollar credit markets through FX swaps has been impaired. In contrast, the dollar’s dominance in international transactions has remained intact. The consequence of market segmentation and dollar dominance has been sporadic dollar shortfalls in times of market turmoil or tightening financial conditions: a rush for liquidity turns into a net “dash for dollars,” and dollar rates in the offshore market rise above those in the onshore markets. Since higher dollar rates in the offshore market drive both offshore borrowers and lenders to buy dollars in the FX spot market directly, the dollar appreciates, at least temporarily.
Regression-based macro trading signals
Regression is one method for combining macro indicators into a single trading signal. Specifically, statistical learning based on regression can optimize model parameters and hyperparameters sequentially and produce signals based on whatever model has predicted returns best up to a point in time. This method learns from growing datasets and produces valid point-in-time signals for backtesting. However, whether regression delivers good signals depends on managing the bias-variance trade-off of machine learning. This post provides guidance on pre-selecting the right regression models and hyperparameter grids based on theory and empirical evidence. It considers the advantages and disadvantages of various regression methods, including non-negative least squares, elastic net, weighted least squares, least absolute deviations, and nearest neighbors.
Generic derivative returns and carry (for strategy testing)
Backtesting of macro trading strategies requires good approximate profit-and-loss data for standard derivatives positions, particularly in equity, foreign exchange, and rates markets. Practical calculation methods of generic proxy returns not only deliver valid strategy targets but are also the basis of volatility adjustments of trading factors and for calculating nominal and real “carry” of macro derivatives. A methodological summary for equity index futures, FX forwards, and interest rate swaps shows that generic return and carry formulas need not be complicated. However, decisions on how to simplify and set conventions require good judgment and adjustment to institutional needs.
Equity market timing: the value of consumption data
The dividend discount model suggests that stock prices are negatively related to expected real interest rates and positively to earnings growth. The economic position of households or consumers influences both. Consumer strength spurs demand and exerts price pressure, thus pushing up real policy rate expectations. Meanwhile, tight labor markets and high wage growth shift national income from capital to labor.
This post calculates a point-in-time score of consumer strength for 16 countries over almost three decades based on excess private consumption growth, import trends, wage growth, unemployment rates, and employment gains. This consumer strength score and most of its constituents displayed highly significant negative predictive power with regard to equity index returns. Value generation in a simple equity timing model has been material, albeit concentrated on business cycles’ early and late stages.
Advanced FX carry strategies with valuation adjustment
FX forward-implied carry is a popular ingredient in currency trading strategies because it is related to risk premia and implicit policy subsidies. Its signal value can often be increased by considering inflation differentials, hedging costs, data outliers, and market restrictions. However, even then, FX carry is an imprecise and noisy signal, and previous research has shown the benefits of enhancements based on economic performance (view post here). This post analyses the adjustment of real carry measures by currency over- or undervaluation. As a reference point, it uses point-in-time metrics of purchasing power parity-based valuation estimates that are partly or fully adjusted for historical gaps. The adjustment is conceptually compelling and has historically increased the performance of carry signals across a variety of strategies.
Tracking systematic default risk
Systematic default risk is the probability of a critical share of the corporate sector defaulting simultaneously. It can be analyzed through a corporate default model that accounts for both firm-level and communal macro shocks. Point-in-time estimation of such a risk metric requires accounting data and market returns. Systematic default risk arises from the capital structure’s vulnerability and firms’ recent performance, as reflected in equity prices. The metric is both an indicator and predictor of macroeconomic conditions, particularly financial distress. Also, systematic default risk has helped forecast medium-term equity and lower-grade bond returns. This predictive power seems to arise mostly from the price of risk. When systematic default risk is high, investors require greater compensation for taking on exposure to corporate finances.
Optimizing macro trading signals – A practical introduction
Based on theory and empirical evidence, point-in-time indicators of macroeconomic trends and states are strong candidates for trading signals. A key challenge is to select and condense them into a single signal. The simplest (and often successful) approach is conceptual risk parity, i.e., an equally weighted average of normalized scores. However, there is scope for optimization. Statistical learning offers methods for sequentially choosing the best model class and other hyperparameters for signal generation, thus supporting realistic backtests and automated operation of strategies.
This post and an attached Jupyter Notebook show implementations of sequential signal optimization with the scikit-learn package and some specialized extensions. In particular, the post applies statistical learning to sequential optimization of three important tasks: feature selection, return prediction, and market regime classification.
