Pure macro(economic) strategies are trading rules that are informed by macroeconomic indicators alone. They are rarer and require greater analytical resources than standard price-based strategies. However, they are also more suitable for pure alpha generation. This post investigates a pure macro strategy for FX forward trading across developed and emerging countries based on an “external strength score” considering economic growth, external balances, and terms-of-trade.
Rather than optimizing, we build trading signals based on the principles of “risk parity” and “double diversification.” Risk parity means that allocation is adjusted for the volatility of signals and returns. Double diversification means risk is spread over different currency areas and conceptual macro factors. Risk parity across currency signals diminishes vulnerability to idiosyncratic country risk. Risk parity across macroeconomic concepts mitigates the effects of the seasonality of macro influences. Based on these principles, the simplest pure macro FX strategy would have produced a long-term Sharpe ratio of around 0.8 before transaction costs with no correlation to equity, fixed income, and FX benchmarks.
The below post is based on proprietary research of Macrosynergy.
A Jupyter notebook for audit and replication of the research results can be downloaded here. The notebook operation requires access to J.P. Morgan DataQuery to download data from JPMaQS, a premium service of quantamental indicators. J.P. Morgan offers free trials for institutional clients. Also, there is an academic research support program that sponsors data sets for relevant projects.
This post ties in with this site’s summary of macro trends and systematic trading strategies.
What are pure macro strategies?
Pure macro trading strategies are trading rules that are informed only by macroeconomic data, as opposed to price data or mixed information sets. In the systematic space, a pure macro strategy refers to an algorithmic trading rule that is based only on macro-quantamental data without using market-based indicators, such as trend and carry. Quantamental indicators are not regular economic data but real-time information states of the market and the public with respect to an economic concept. They are based on concurrently available vintages of regular economic data series. Macro quantamental indicators are suitable for testing relations with subsequent returns and backtesting related trading strategies, like price-based indicators. For this post, we use indicators of the J.P. Morgan Macrosynergy Quantamental System (“JPMaQS”).
The drawback of pure macro strategies is that they limit the range of plausible value-generating principles, intentionally disregarding valid price-based indicators. The benefit of pure macro strategies is that they are suitable for generating true alpha. Statistical alpha can be divided into true alpha, which reflects the quality of the investment process, and fake alpha, which is a premium for non-directional systematic risk. Fake alpha arises from exposure to conventional factors that are not correlated with the market portfolio (view post here). Many types of price-based trading signals, such as carry and trend, have become “mainstream.” Market exposure has increased over past decades thanks to specialized funds and relatively low costs of signal production. This gives rise to crowded positioning and risk premia.
By contrast, pure macro indicators require more investment in the underlying data (vintages) and domain knowledge in the field of macroeconomic data and equilibrium theory. As a result, pure macro strategies are less common, less standardized, and more fragmented across information types than popular price-based strategies. Moreover, pure macro strategies often thrive on public inattention to fundamental change and increase the efficiency of financial markets. Related investment gains are not a zero-sum game.
A pure macro FX strategy
In this post, we present a pure macro FX forward trading strategy for seven developed markets and 19 emerging market currency areas. This group contains all “small” countries (excluding the U.S., the euro area, and Japan) that have since 2000 featured tradable, liquid FX forward or non-deliverable forward markets, as well as largely convertible currencies and flexible exchange rates.
- The developed market currencies are AUD (Australian dollar), CAD (Canadian dollar), CHF (Swiss franc), GBP (British pound), NOK (Norwegian krone), NZD (New Zealand dollar), SEK (Swedish krona),
- The emerging market currencies are BRL (Brazilian real), CLP (Chilean peso), COP (Colombian peso), CZK (Czech Republic koruna), HUF (Hungarian forint), IDR (Indonesian rupiah), ILS (Israeli shekel), INR (Indian rupee), KRW (Korean won), MXN (Mexican peso), MYR (Malaysian ringgit), PEN (Peruvian sol), PHP (Philippine peso), PLN (Polish zloty), RON (Romanian leu), THB (Thai baht), TRY (Turkish lira), TWD (Taiwanese dollar), and ZAR (South African rand).
Periods where any of these currencies have been untradable, not convertible, or pegged, have been excluded from the main body of the analysis. These exceptions can be viewed in the above-linked Jupyter Notebook.
The strategy is informed by quantamental indicators related to the competitiveness of economies: growth, external balances, and terms-of-trade. Within these types, we select a set of available quantamental indicators that are all plausibly related to the broad concept, combining them with equal weights. Neither the selection of the indicators nor their weighting is in any way optimized, as this post merely serves as proof of concept.
- Economic growth trends: This group comprises estimated GDP trends and business sentiment trends. There are two types of GDP trends, measured as % over a year ago in 3-month moving averages, available in quantamental format from JPMaQS: intuitive GDP trends, which replicate the standard estimation procedures of market economists (documentation here), and technical GDP trends, which replicate point-in-time standard nowcasting of econometricians (documentation here). We use both types outright and relative to the base currency of each country (USD for most, but EUR for some European countries, and a basket of USD and EUR for GBP and TRY). Furthermore, we use changes in manufacturing business sentiment (documentation here), 3 months over the previous 3 months, and 6 months over the previous 6 months to cover more recent changes in economic activity.
- External balances trends: This group consists of changes in external trade balances and external current account balances as a percent of GDP at different horizons (documentation here). In particular, the group uses changes of seasonally adjusted external balance ratios of the following types: the latest 3 months over the previous 3 months, the latest 6 months over the previous 6 months, and the latest 3 months over the previous 5 years.
- Terms-of-trade trends: Terms-of-trade are country-specific ratios of average export and import prices. This group uses JPMaQS’s two types of quantamental terms-of-trade data (documentation here). The first type, commodity-based terms of trade, uses commodity trade and price data alone. The second type, mixed terms of trade, uses (a) low-frequency broad external trade data to capture medium-term dynamics and (b) daily commodity-based terms-of-trade to estimate recent and intermittent dynamics. Here we use the following terms of trade changes: last month (21 days) over a year ago, last month over the previous 1-year average, and last month over the previous 5-year average.
The quantamental indicators are combined within the group into a conceptual trend score and, then, across groups into a single “external strength” score. For this combination, we first normalize the individual indicators around their zero values, which are all “natural” neutral levels. Normalization is performed sequentially across time to avoid hindsight in signal calculation and based on the full panel of available countries and periods. The resulting scores are trimmed are winsorized (“capped”) at three standard deviations, a standard procedure to de-emphasize outliers. After normalization, the individual indicators are combined with equal weights, a method that could be called “conceptual risk parity” since it gives the variation of each indicator approximately equal weight in the calculation of the composite. If individual scores are not available for a linear composite, the composite score is calculated based on the available ones.
Subsequently, the combined group indicators are re-normalized to prepare for calculating a global external strength score. Then the group scores are combined with equal weights, preserving the principle of “conceptual risk parity.”
The external strength score is a plausible predictor of volatility-targeted FX forward returns. These are returns of FX forwards of a long in the local currency versus the base currency in percent of risk capital on a position that is scaled to a 10% (annualized) volatility target. The generic return calculation assumes a rollback to full 1-month maturity at the end of the month. The case for the volatility-targeted position is compelling for a macro strategy to make cross-country risk-taking comparable and to account for differences in the response of exchange rates to macro trends across countries. For example, small open economies require less of an exchange rate adjustment for economic outperformance than large, closed economies.
The explanatory power of the pure macro composite
We investigate the explanatory power of the external strength score for subsequent FX forward returns from 2000 to 2023 (July) based on the Macrosynergy panel test (view post here). This test is more meaningful than standard correlation p-values because it adjusts targets and features of the predictive regression in a cross-country panel for common (global) influences. The stronger these global effects, the greater the weight of deviations from the period-mean in the regression, thus avoiding “pseudo-replication” without disregarding any country-specific experiences.
Panel regression shows a positive relation between end-of-month information states of external strength and subsequent weekly and monthly vol-targeted FX returns. The test suggests that the probability of this relation being systematic rather than accidental is between 98.3% and 96.3%. Note that standard pooled significance tests based on parametric and non-parametric correlation would put that probability at over 99.9%.
The positive predictive relation is a subtle one at the monthly frequency but pervasive. Thus, positive correlation prevailed in 63% of all years since 2000 and 77% of all countries. Similarly, the accuracy (ratio of correctly predicted return directions) and balanced accuracy scores for the prediction of monthly returns have averaged 51% and 52.1%, respectively, with above-50% ratios in more than two-thirds of all years and countries. Also, for all group macro trends (economic growth, external balances, and terms of trade), accuracy scores have been above 50%, and all parametric and non-parametric correlation coefficients have been positive.
We estimate the economic value of a composite external strength score based on a naïve PnL computed according to a standard procedure used in Macrosynergy research posts. A naive PnL is calculated for simple monthly rebalancing in accordance with the external strength score at the end of each month as the basis for the positions of the next month and under consideration of a 1-day slippage for trading. The trading signals are capped at 2 standard deviations in either direction for each currency as a reasonable risk limit and applied to volatility-targeted positions. This means that one unit of signal translates into one unit of risk (approximated by estimated return volatility) for each currency. The naïve PnL does not consider transaction costs or compounding. For the chart below, the PnL has been scaled to an annualized volatility of 10%
Value generation of the pure macro signal has been consistent and economically significant. The 23-year Sharpe ratio has been 0.72, and the Sortino ratio 1.04. Importantly, the correlation of the PnL with key market benchmarks, such as the S&P500 or the U.S. treasury, has been near zero. Also, there has been almost no correlation with EURUSD forward returns or returns of a basket of all small-country currencies. All this suggests that the PnL generated by the pure macro signal is additive to standard long-only benchmarks.
The benefits of double diversification and risk parity
Double diversification means spreading the risk over different currency areas and plausible conceptual macro factors. The argument for diversification across currency areas with similar institutional features is that it evens out the influence of idiosyncratic disturbances across currencies. The argument for diversification across macro factors is that it mitigates the seasonality risk: the influence of macroeconomic trends often comes in multi-year episodes, and the economic state of a country does not change in high frequency.
High diversification can be achieved by allocating across currency areas and macro concepts on a risk parity basis.
- Risk parity across currency signals means that per unit of signal, we allocate the same risk across sections and periods, irrespective of how successful the signal has been in the past for the individual currency area. This means we consider all cases in a panel, i.e., combinations of cross-sections and periods, as drawings from the same population. In the current setting, this is accomplished by applying signals to vol-targeted FX forward positions.
- Risk parity across macroeconomic concepts means that all macro trends contribute equally to the variation of the composite factor and that all macro trend constituents contribute equally to the variation of the macro trend, irrespective of their individual explanatory power. This means that we consider all plausible macro trend factors as equally important predictors. In the currency setting, this is accomplished through the normalization of trends and constituents.
The term “plausibility” is very important and means that the predictive relation between the macro trend and FX forward return is supported by economic theory. Typically this means that the sign of the macro trend is expected to drive the FX forward return in a certain direction and that it is unlikely that the market uses the related macro information instantaneously and completely, in accordance with the theory of rational inattention (view post here). Risk parity across macro concepts does not mean that we lump together a set of macro indicators but rather that we apply an intelligent logical structure: we give equal weights to plausible concepts and, within these concepts, give equal weight to constituents with approximately equal claim to representing the concept.
The experience of macro trend-based FX trading since 2000 provides ample evidence that supports the benefits of double diversification.
Diversification across currencies has proven relevant because returns and macro trend signals display idiosyncratic dynamics across currency areas. The first point is illustrated by the below correlation matrix of volatility targeted weekly FX forward returns across the 26 developed and emerging market currencies considered in this research. While positive correlation dominates, coefficients are mostly below 50% and often near zero or negative.
Second, external strength scores post both negative and positive correlation across countries, and high positive correlation is rare. This means that even when the correlation is high across returns, signal diversity can still mitigate risk concentration.
Another way to illustrate the benefit of signal parity is to compare the above naïve PnL for 26 global currencies with the naïve PnL for the same strategy but for the 7 developed market currencies alone. Trading developed markets with macro trends may be more convenient and has also been profitable. However, the below chart shows that a developed market FX strategy would have produced less than half the risk-adjusted return of the global portfolio, with a Sharpe ratio of just 0.32., and greater seasonality. Also, inspection of the influence of the individual macro trends reveals that external balances broadly failed to predict developed market FX returns. That probably reflects that, unlike emerging markets, developed markets have, in recent decades, not experienced balance of payments stress. However, this experience reasonably does not preclude balance of payment issues in developed countries in the future.
Diversification across macroeconomic concepts has proven relevant because the predictive power of these concepts often comes in seasons. This is a phenomenon that can be seen in many asset classes, exemplified by the importance of inflation equity returns (view post here) or the predictive power of import growth for fixed income returns (view post here)
In the pure FX strategy case, all three types of macro trends have contributed to PnL generation but in different “seasons.” Economic growth trends played an important role in predictive FX trends in the 2000s but have generated only modest value since 2010. Conversely, external trade trends produced no value in the 2000s but greatly added to PnL generation in the 2010s and 2020s. The mirror image of these probably is not accidental: in times of strong international capital flows, high-growth economies tend to attract FX inflows even if their external balances deteriorate, while in times of financial shocks and de-globalization, external deficits are a greater concern. The important point is that jointly these two trends produced consistent value. Finally, terms-of-trades have helped PnL generation across decades, but naturally only in episodes where international commodity prices changed significantly.
Within certain limits, risk parity can even beat hindsight. The diversified risk parity signal would have outperformed not only all three major macro trend signals but also each and every signal based on any of the 18 underlying constituents. The chart below compares the performance naïve PnLs of the parity-based diversified external strength score and all trend constituents, i.e., all the individual quantamental series (normalized) behind the three main macro trends, as explained above. Even the best score chosen with hindsight (merchandise trade balance trend) would only have produced a Sharpe of 0.51 versus 0.68 for the composite.
The below correlation matrix also shows monthly correlations across all constituents of the macro trends used in this post. The (JPMaQS) ticker names are explained in the documentation notebooks linked above. These indicators are block-wise positively correlated, i.e., within the trend category they represent. However, there is not much correlation across blocks and even a negative tilt of correlation between growth trend indicators and external balance trend indicators. This illustrates the potential for diversification across concepts. Additional macro concepts could be applied to the present pure macro-FX strategy, such as labor market tightness, producer price growth, credit growth, or international investment positions. Beyond plausibility and evidence for direct predictive power, an important criterion for extending a model is low correlation and different seasonality relative to the existing set.
