Economic Regimes Unveiled: Strategic Insights for Asset Performance

Our study begins amidst the backdrop of the heightened inflationary pressures experienced during 2022 and 2023. This period sets the stage for widespread speculation about a looming recession in 2023. Contrary to these anticipations, however, 2023 has displayed remarkable economic resilience. This unexpected turn of events underscores the complexity of financial markets and highlights the critical need for a deep understanding of economic regimes.

‍

Our analysis reveals a prevailing trend of moderate growth for 2023, with a 60% probability of continuation over the next three months and a 50% chance extending throughout the year. While the risks of recession or stagflation are present, they remain below a 20% probability, indicating a lower likelihood of economic downturn or inflationary stagnation in the near term.

‍

As a result, in this context of moderate growth, certain assets are poised to outperform. Specifically, in private markets, sectors like distressed debt and buyouts show relative out-performance due to restructuring opportunities. Hedge funds, especially those focusing on equity hedge and event-driven strategies, thrive on corporate events and steady market trends. Similarly, factor-based returns show a clear preference for quality and value strategies, signalling a market favouring well-established, undervalued assets.

‍

This research focuses on systematically categorising economic regimes, and not only provides a nuanced appreciation of market dynamics but also offers valuable insights into strategic asset allocation and investment decision-making.

‍

The financial sector is characterised by its constantly evolving nature. As a complex adapative system, market regimes develop and then shift, which in turn can dramatically affect the investment landscape. For investors, the ability to identify and interpret these regimes is crucial, allowing them to adapt their investment strategies and anticipate economic shifts, thereby making strategic portfolio decisions with greater clarity and therefore greater effectiveness.

‍

Although authoritative sources like the National Bureau of Economic Research (NBER)¹ provide valuable macroeconomic insights, a more nuanced and timely regime identification is essential for staying ahead in the dynamic world of investment. Bridging this gap requires an analytical approach that not only understands broad economic trends but also explores the intricacies that affect specific investment categories.

‍

Our exploration adopts a machine learning perspective utilising the Gaussian Mixture Models (GMM)² to analyse and categorise economic indicators, thereby delineating distinct market regimes. Through this lens, we examine the nuances of economic state characteristics, historical periods, and their impact on asset performance. We place a particular emphasis on alternative asset classes, such as hedge funds, factor strategies, and private markets. This focus aligns seamlessly with our initiatives at Klarphos, where we are deeply invested in these diverse strategies and asset classes, which we consider as essential components of an investment portfolio, especially for institutional investors with a long-term horizon.

Figure 1: NBER Business Cycle - Unemployment Rate and Recessions since 1950

^{Source: NBER; period 1950 - 2023}

‍

Determining the regimes

‍

Establishing these regimes can be approached from several angles. Traditionally regimes have been defined through a combination of market wisdom, historical perspective, and key economic indicators, such as CPI, GDP, and unemployment rates. For instance, the GDP growth rate serves as a principal indicator of recession. While quarterly GDP data provides valuable historical insight, it may not fully encapsulate the swift and nuanced economic shifts that occur within each quarter. Additionally, given GDP's inherent nature as a lagging indicator, its data points often arrive with a delay, during which the economic landscape may have already transitioned into a new phase. This temporal gap presents a significant hurdle in historical economic analysis based solely on official GDP figures.

‍

To address these complexities, our analysis integrates a careful selection of economic indicators, ranging from leading to lagging, and coincident metrics. We conducted the study within the US context, and included a broad array of economic measures that capture the overall health and activity of the economy, price stability, business confidence levels, and labour market conditions, as well as market-centric indicators such as equity returns, market stability indices, and the trajectory of interest rates³.

‍

The GMM's strength lies in its reliance on empirical data, allowing for the generation of outputs that depict an array of market scenarios. Nevertheless, it is worth noting that while the GMM outputs are data-rich, they may not always convey an intuitive grasp of the economic climate. To enhance our understanding, we dissect these outputs by examining the behaviour of the feature variables across identified states. For instance, one state closely aligns with US recessions, prompting us to label it the "recessionary" regime. To affirm this classification, we have validated our findings for all five regimes through authoritative and academic sources.

‍

Aligned with the analytical capabilities of the GMM, we have developed a classification for the five distinct market regimes identified in our study - extending beyond the traditional four-state model and with the desire to offer a more nuanced understanding of market dynamics.

‍

Figure 2: Economics States of the US since 1962 (highest probabilities only)

^{Source: Klarphos; period 1950 - 2023}

‍

Regime 1: Moderate Growth

Characteristics: this market regime is characterised by a tempered economic pace, where growth is consistent but restrained, avoiding the highs of boom cycles. Market advances during this period are steady, with stability in prices and a measured performance in equity returns, reflecting a phase of moderate progress. Industrial expansion is present but not pronounced and employment shows subtle improvements. Overall market volatility is lower indicating a climate of cautious optimism among investors.

‍

Exemplary periods: during the late 1990s, the economic landscape underwent significant structural changes. Industries were adapting to new technologies and grappling with the early stages of globalisation, which tempered economic expansion. This period, reflecting moderate growth, was marked by a careful balance between maintaining traditional industrial outputs and integrating emerging tech sectors. It was a time of strategic realignment, where the full impact of the digital age was still on the horizon, leading to a cautiously optimistic market that moved forward with an awareness of the impending technological transformation.

‍

Assets returns: in a moderate growth phase, assets with a strong foundation for potential value appreciation tend to outperform. For private markets, this includes sectors like real estate, distressed debt, and buyouts, where the potential for restructuring can lead to significant gains. Hedge funds particularly shine in equity hedge and event-driven strategies, which thrive on corporate events and steady market trends. Factor-based returns show a clear preference for quality and value strategies, signalling a market favouring well-established, undervalued assets. These trends suggest a strategic advantage in investing in areas with solid fundamentals poised for growth or recovery.

‍

‍

Figure 3: Asset Returns during Moderate Growth

^{Source: Bloomberg, Preqin, HFR; period 2000 - 2023.09}

‍

Regime 2: Expansion

Characteristics: the expansion phase is often characterised by a robust and widespread increase in economic activity. This is a period where growth is not only consistent but accelerates, reflecting a confident economic outlook. The labour market strengthens and consumer spending rises in response to increased employment and income. Financial markets tend to perform well with significant gains in stock valuations signalling investor optimism.

‍

Exemplary periods: the 1960s are often celebrated as a golden age for the US economy, a decade marked by robust industrial output, rising consumer demand, and significant improvements in civil infrastructure. Progress in civil rights led to a more diverse workforce, further propelling economic growth. The early 1990s, following the recessionary period of the late 1980s, were a time of expansion. The economy was lifted by the advent of the internet and a growing tech industry, which led to increased productivity and initiated a new era of innovation and economic activity.

‍

Assets returns: during an economic expansion phase most private market assets perform strongly, with venture capital, private equity, and real estate displaying significant returns. This is indicative of the high investor confidence that typically accompanies periods of accelerating economic activity. Investors are likely capitalising on the broad-based growth, with increased consumer spending and a strong labour market fuelling returns in these sectors.

‍

Hedge funds also perform well, especially those with event-driven and equity-focused strategies, benefitting from the heightened corporate activity and rising stock valuations that are characteristic of expansion periods.

‍

Conversely, the factor returns show a mixed picture, with several strategies like dividend yield underperforming, as investors may be bypassing these more conservative strategies in favour of direct equity investments or alternative assets that offer higher growth potential.

‍

Figure 4: asset returns during expansion

^{Source: Bloomberg, Preqin, HFR; period 2000 - 2023.09}

‍

Regime 3: Stagflation

Characteristics: a period defined by economic stagnation paired with high inflation. Key indicators during such periods are markedly negative, with economic output dwindling, unemployment rates climbing, and consumer prices soaring. This combination of a sluggish economy with inflation leads to a decline in purchasing power and investor confidence, often causing a surge in market volatility.

‍

Exemplary periods: the stagflation of the 1970s was ignited by a series of economic disturbances: the collapse of the Bretton Woods system, a peak in Texas oil production, and an OPEC oil embargo, which caused a sharp increase in oil prices. These factors led to a persistent rise in inflation in the US economy, coinciding with high unemployment and slow growth. Moving to 2022, heightened inflationary pressures emerged, driven by fiscal stimulus in response to the COVID-19 pandemic and global political conflicts, such as the war in Ukraine and subsequent sanctions on Russia.

‍

Assets returns: during stagflation, the investment landscape becomes challenging, as reflected in the overall subdued or negative returns across various asset classes. Private equity and buyout sectors experience the worst downturns due to higher cost of capital and higher discount rate, whereas natural resources often see favourable outcomes due to their ability to contribute to the inflationary pressures caused by higher inflation levels. Hedge fund strategies face setbacks as well, with the exception of defensive and market neutral strategies. Most factors have negative returns with profitability and value factors outperforming as investors gravitate towards companies with strong, reliable earnings as a safe haven in economically turbulent times.

‍

Figure 5: asset returns during stagflation

^{Source: Bloomberg, Preqin, HFR; period 2000 - 2023.09}

‍

Regime 4: Recession

Characteristics: recession is characterised by an economy in sharp decline, triggered by, for example, financial shocks, a downturn in asset values, or policy errors. Similar to stagflation, we have traits of rising unemployment and a slowdown of economic activity. However, the key difference lies in the general fall in prices during a recession, as opposed to the persistent inflation that is seen in stagflation. A recession is marked by a notable slump in GDP and consumer spending, alongside an increase in market instability. While both economic phases impact investor confidence, a recession is particularly defined by a widespread retreat in economic indicators, and monetary policy responses tend to differ; recessions usually call for stimulative measures, while stagflation may prompt more restrictive actions to control inflation.

‍

Exemplary periods: the early 1980s experienced a recession in the wake of the stagflation of the late 1970s, which had been precipitated by the oil/Iran crisis. Elevated interest rates, a tool to curb the rampant inflation of that era, played a role in dampening economic growth. The 2008 recession, triggered by the subprime mortgage crisis and subsequent financial turmoil, marked another period of economic downturn. Similarly, the 2020 recession ensued from the widespread disruption caused by the COVID-19 pandemic, leading to a rapid and significant slump in global economic activity.

‍

Assets returns: in recessionary periods, asset returns generally falter as seen in the retreat across real estate, natural resources, and most private equity categories. Market conditions challenge investors, pushing them towards more defensive positions. Hedge funds with market neutral strategies or fixed income focus demonstrate relative stability, likely benefitting from their ability to exploit market dislocations, monetary stimulus, and the safe-haven status of bonds.

‍

‍

Figure 6: asset returns during recession

^{Source: Bloomberg, Preqin, HFR; period 2000 - 2023.09}

‍

Regime 5: Recovery

Characteristics: in the aftermath of economic downturns, the recovery phase is characterized by a gradual uptick in leading economic indicators and a calming of market volatility, signalling a stabilising environment. Employment figures slowly improve as industries reorient and consumer prices begin to level off. Stock markets respond positively, supported by a collective sigh of relief from investors.

‍

Exemplary periods: a notable historical period occurred between 1983 and 1986. During this time, the US economy was emerging from the early 1980s recession, attributed to tight monetary policies that curbed the high inflation of the late 1970s. The subsequent easing of these policies, along with a reduction in tax rates, spurred growth and marked the beginning of a recovery phase. A similar pattern is observed from late 2009 through to 2013. This period succeeded the Great Recession, with the economy beginning to rebound mid-2009, which was characterised by gradual improvements in the job market, bolstered by significant monetary stimulus and federal intervention.

‍

Assets returns: the recovery phase typically sees a resurgence in private markets, with real estate and distressed private equity showing robust performance, reflecting investor confidence in assets poised for rebound. Hedge funds, particularly those adopting event-driven and relative value strategies, capitalise on this environment by harnessing market adjustments and heightened corporate activity. Meanwhile, in the public markets, strategies emphasising value and momentum demonstrate strength, as investors shift towards sectors with the potential for growth in a revitalising economy.

‍

Figure 7: asset returns during recovery

^{Source: Bloomberg, Preqin, HFR; period 2000 - 2023.09}

‍

Where we are today

‍

A zoomed-in view of the economic landscape from 2010 to 2023, as interpreted by the probabilities derived from the GMM, shows the stacked probabilities of various economic states and offers a comprehensive view of the economy's evolution over time. Each colour band reflects a distinct economic state. The height of each band is proportional to its probability, summing up to a total height of one, or 100%, which signifies absolute certainty. Periods where the bands are filled with multiple colours, imply mixed economic signals. However, typically one state's features are more predominant, offering clarity on the overall economic condition.

‍

The recession of 2020, as defined by the NBER, was notably short, spanning from February to April. However, the GMM model's identification of a more extended recessionary period suggests a deeper impact on the economy, reflecting a broader array of economic dynamics and the lingering effects of the COVID-19 pandemic across various sectors. Similarly, while 2022 witnessed inflation rates above the long-term average, the trend indicates a shift towards an easing inflation scenario.

‍

2023 stands out on the chart as a time of moderate growth. This classification may seem counter-intuitive given the market's lingering concerns over "higher-for-longer" inflation and a potential recession, as many have anticipated, due to aggressive inflation-curbing interest rate hikes.

‍

Despite these concerns, including a brief period of banking sector instability and rising short and long-term interest rates, the US economy has shown remarkable resilience so far this year, with the third quarter highlighting its strength - real GDP growing at an annual rate of 4.9%, the most rapid expansion we have seen in almost two years. This growth surpassed consensus forecasts, both globally and within the US. Core inflation has declined from its 2022 peak, while the labour market has performed remarkably well, with unemployment rates falling to about half a percentage point below their pre-pandemic levels, indicating a strengthening job market. Based on all this data, our model even identified an expansionary month in July.

‍

Figure 8: Market Regimes' Probabilities of US from 2010 to 2023 (coloured band height denotes likelihoods)

^{Source: Bloomberg, Preqin, HFR; period 2000 - 2023.09}

‍

Where are we transitioning to in 2024?

‍

Current data does not point to a booming economy but importantly, it does not signal a high probability of contraction either. There is an emerging picture of moderate growth, characterised by steady, controlled expansion buoyed by cautious optimism from both consumers and businesses. Our historical analysis has generated two types of transition probability matrices to understand economic regimes.

‍

The first matrix traces transitions on a month-to-month basis, providing a detailed view of immediate economic fluctuations. The second matrix, intentionally omitting the main diagonal, presents a clearer picture of regime changes by focusing solely on the probabilities of transitioning from one state to another, thus excluding periods of status quo.

‍

Figure 9: Transition Matrix - Month to Month                       Figure 10: Transition Matrix - Regime Shift

^{Source: Klarphos; between 1960 - 2023.09}

‍

Figure 9 shows a pronounced tendency for the economy to persist in its current condition. Although all regimes generally exhibit a likelihood of continuation, moderate growth stands out as more variable. This state's characteristic mix of indicators may lead to smoother transitions to different economic conditions. Meanwhile, 'stagflation' emerges as a less persistent state, suggesting its typically short-lived nature.

‍

In contrast, 'recession' appears as the most entrenched regime, underscoring the challenges policymakers face in making change, particularly when structural issues are at play. This reflects the complexities of applying monetary policy as a remedy; while inflation can be curbed, deep-seated economic weaknesses are not as readily alleviated by such interventions.

‍

Regime Shift Transition

‍

Figure 10 focuses on change probabilities. In the current state of 'moderate growth', which could be described as navigating cautiously, the matrix reveals a combined 30% chance of moving into stagflation or recession. Conversely, the combined likelihood of progressing into recovery or expansion is notably higher, reinforcing the current observations of the odds of a 'soft-landing'.

‍

Additionally, if inflation rates climb again the matrix suggests a relatively minor chance (below 20%) of descending into a recession, with a robust tendency (over 70%) to cycle back to moderate growth.

‍

Another way to answer the question, “Which economic states are we likely to encounter in 2024?” is to calculate the historical distribution probabilities of the economy experiencing various states over a specified future period, such as the ensuing three months or twelve months.

‍

For example, within our dataset covering 740 months from January 1962 to September 2023, we have identified 291 months that are characterized by economic expansion. We examined the subsequent economic conditions that followed each of these 291 expansionary months over the following three and twelve months and calculated the distribution probabilities of different economic states.

‍

This approach is visualized in tables. The vertical axis is labelled 'Initial States', and represents the economic condition at the start of the observed period, while the horizontal axis lists the 'Subsequent States' that the economy may transition into after the start period. The intersecting cells across these axes display the probabilities of transitioning from the initial state to the subsequent state over a specified period, i.e., three months and twelve months.

‍

Probabilistic Forecasts of Economic States

‍

Figure 11: Transition Matrices - Next 3 and 12 months

^{Source: Klarphos; between 1960 - 2023.09}

‍

• The table indicates that the likelihood of entering a recession or experiencing stagflation in the next three months is low, combined at only 14.3%.

• In contrast, the probability of the economy maintaining a state of moderate growth is high, at close to 60%. This signals that moderate growth is the most anticipated economic state in the near term, supporting the 'soft-landing' scenario.

• The chance of moving into expansion is low but not negligible at 26.7%. This migration may not materialise but increases optimism.

• When we extend the forecast to twelve months, the transition matrix table shows that the chance of moderate growth occurring remains the highest at 50%, suggesting that this state will likely predominate for half of the upcoming year.

• The combined probability of encountering either a recession or stagflation increases to 19% over the 12-month period, vs. 15% in a 3-month outlook, equating to an anticipated 2.3 months of recessionary conditions within the year. This suggests a slightly higher risk of economic downturn or inflationary stagnation compared to the short-term outlook, but still represents a less likely scenario than continued moderate growth.

‍

In summarising our study on market regime analysis, we have detailed the historical transitions between economic states and their evolution. This analysis sheds light on how assets perform across various economic cycles, and offers an empirical guide for asset allocation and portfolio management. The regime mapping uncovers the trajectories that asset classes have followed amid economic shifts. The transition probabilities also provides a basis for a more informed approach to asset selection. Moreover, looking into 2024, historical data suggests the likelihood of recession or stagflation within the next 3/12 months to remain below 20%, and the chance of entering expansion is even higher at 26 to 29%. However, the base case remains a moderate growth condition, with a 60% probability of continuity over the next 3 months and a 50% chance over the next 12 months.

‍

It is crucial to acknowledge the limitations inherent in our analysis. The probabilities are based on past trends, dependent on the data's breadth and the periods we have examined, as well as the economic indicators we chose to include.

‍

Moving forward, Klarphos is poised to build upon this initial analysis and expand the scope of our work, delving into the intricacies of regime mapping and its relevance for multi-asset portfolio construction and how this historical lens can guide asset allocation strategies.

‍

‍

‍

Endnotes

1) The National Bureau of Economic Research (NBER) Business Cycle Dating Committee meticulously charts the chronology of US business cycles, defining and dating the peaks and troughs that demarcate economic expansions and recessions . An NBER-defined recession reflects a significant, widespread decline in economic activity persisting for more than a few months. This determination relies on diverse indicators, notably real personal income less transfers and nonfarm payroll employment, to capture a broad economic impact, not confined to a single sector. In practical application, the NBER's dating of business cycles is a retrospective process, awaiting sufficient data to confirm a transition between economic phases.

2) Gaussian Mixture Models (GMMs) stand as a sophisticated approach within unsupervised machine learning, designed to discern intricate structures in multivariate data without predefined classifications. These models hypothesize that data emanates from a composite of several Gaussian distributions, each defined by distinct means (μ) and covariances (Σ), thereby delineating the probabilistic frontiers of subpopulations within the multidimensional feature space. In contrast to the deterministic clustering of K-means, GMMs implement a probabilistic, or 'soft', clustering paradigm. This modality assigns a membership likelihood for each data instance across the array of Gaussian distributions, thereby accommodating the inherent uncertainty and feature overlap characteristic of complex datasets.

3) Our econometric framework incorporates indicators such as the US GDP, PMI, unemployment rate, S&P 500 returns, realised and implied volatilities, the 3-month Federal Reserve rate, and yield spreads (2-year/10-year and 10-year/30-year). These indicators were interpolated for temporal alignment, standardised to a uniform scale, and thoughtfully weighted, reflecting their economic import.

‍

‍

‍

‍

Important Information

This document is informative purposes only. It does not constitute research, investment advice nor solicitation to invest in any investment product or service that Klarphos offers or may offer in the future in any jurisdiction. The information contained herein is based on projections, estimates and/or other financial data and has been prepared internally by Klarphos. Opinions expressed therein are current opinions as of the date of this document only and are subject to change at any time without notice.

No representations are made as to the accuracy of the observations, assumptions and projections. No subscriptions to any Klarphos products are possible based solely on this document. Any investment decisions should be made in accordance with the legal documentation of a fund such as its offering memorandum.

Klarphos is not entitled to provide any tax or legal advice.

Past performance is not indicative of future returns. There can be no assurance that the strategy objectives will be realized or that the strategy will not experience losses. Target returns are hypothetical and are neither guarantees nor predictions of future performance. There can be no assurance that the target returns will be achieved.

Klarphos is an Asset Manager specialized in customized portfolio solutions and advisory services for institutional clients based in Luxembourg. Klarphos concentrates its asset management on Alternative Investments and also provides advisory services for strategic asset allocation and ALM optimization. The asset manager employs an international team of specialists and is regulated by the Luxembourg financial regulator CSSF as an Alternative Investment Fund Manager (AIFM).