Economic cycles have been a crucial driver for financial markets, making regime-based dynamic asset allocation (DAA) a common practice in supporting optimal portfolio construction. Changes in key economic indicators, monetary policy trends and the evolution of financial leverage have proven a reliable guide in identifying business cycle phases. The new economic paradigm, triggered by the 2008 Great Financial Crisis (GFC), overcame the traditional bidimensional approach, which leveraged on growth and inflation for the definition of economic phases. In a post-GFC world of ever-growing debt, the role of monetary policy was so strong and persistent that it became one of the primary drivers of asset class returns and correlation.
Taking advantage of its multi-year experience in cross asset modelling, Amundi Institute leverages artificial intelligence for its investment approach, in order to tackle complex and fast-evolving financial markets. The Institute developed an innovative tool called the Advanced Investment Phazer (AIP), which embraces the complexity and multi-dimensions of regime determination. It encompasses an extensive range of global variables, aimed at building the optimal portfolio allocation with the potential for outperformance and drawdown minimisation via the diversification of trade bets. Regimes are identified by a clustering algorithm applied to a comprehensive set of macro-financial variables split over four dimensions: growth, inflation, monetary policy and financial leverage.
There are five phases – Correction, Contraction, Recovery, Late cycle and Asset reflation – each featuring reference values for the aforementioned macro-financial measures. The distance between an array of forecasts for our factors and each regime’s thresholds will determine the likelihood of each phase in the future (i.e., the smaller the difference, the higher the probability). The resulting optimal portfolio allocation will hinge on the probability distribution of the regimes over the forecasting horizon and the returns that asset classes have delivered historically in each phase.
For instance, according to the AIP, the economic Recovery regime in place from Q4 2020 to Q4 2021 was replaced by a Late Cycle regime in H1 2022. Throughout this Late cycle phase, asset allocation has been tilted towards high-quality risky assets, partly offsetting the pro-beta stance that dominated in 2021. In H2 2022, a Late cycle regime will be as likely as a Correction regime, stemming from the fallout of the expected economic recession in the Eurozone and the United Kingdom. As such, the asset allocation stance needs to become more defensive, with higher exposure to government bonds and gold. In 2023, as the global economy is expected to slow below trend levels, a Correction regime will stand side by side with a Late cycle regime, suggesting focus should remain on defensive assets (see figure 1).