The Quant Renaissance Part II: Winter’s Thaw?

Introduction

The quantitative investing landscape has enjoyed a remarkable revival since the ‘Quant Winter' of 2018-2020, a prolonged period of losses for many traditional factor strategies. Underpinned by a rebound in factor returns, this renaissance has renewed investor confidence and widened the performance gap between systematic managers. But must all good things come to an end? After a number of favourable years, some investors are naturally questioning whether this positive momentum can continue – or if we are potentially facing another freeze.

This paper addresses this question by examining how the evolution of systematic equity investing has created opportunities for genuine differentiation. Building on our original 2024 analysis we explore why we believe ‘this time could be different’ and how innovative approaches to quant investing may help investors weather future market dislocations.

The root causes

The 2008-2009 Global Financial Crisis (GFC) demonstrated just how rapidly headwinds can develop for systematic equity strategies. The 2018-2020 period proved that Quant Winters can emerge even outside of traditional recessions. These experiences naturally prompt curiosity about how sustainable current favourable conditions are and the structural resilience of today’s quantitative investing approaches.

Our research identifies two primary drivers of past underperformance: adverse macro environments and the inherent sensitivity of traditional factors to them. These factors often exhibit strong cyclical relationships with macroeconomic variables, making them vulnerable to regime shifts, policy changes, and broader economic dislocations. Crucially, this sensitivity introduces risk that cannot be easily diversified away through traditional portfolio construction techniques.

This relationship extends beyond simple correlation with economic indicators. Many factors are inherently linked to the economic cycle, performing well during specific phases of economic expansion or contraction but struggling during other phases or transitions. Additionally, factors often embed time-varying bets on macroeconomic variables such as interest rates, inflation expectations, or currency movements, creating exposures that may not be immediately apparent.

Isolating specific sensitivities, such as the relationship between Value and inflation or Momentum and volatility, can help identify when strategies are most vulnerable.

However, while individual macro relationships provide valuable insights into factor behaviour, economic conditions reflect the complex interaction of multiple variables, including interest rates, inflation, growth expectations, policy settings, and market sentiment. To capture these interactions comprehensively, we leverage Man Numeric’s MacroScope model,¹ a framework that identifies how variables combine to create distinct market environments and their implications for systematic investment strategies.

Four distinct regimes

By applying the MacroScope framework, we identified four distinct macroeconomic regimes that capture the essential characteristics of different investment environments:

• Crisis/recession periods (Regime One): Times of acute economic and financial stress, characterised by defensiveness, elevated volatility, correlation spikes between asset classes, and significant liquidity constraints across markets. During these periods, traditional investment relationships often break down as investors prioritise capital preservation over return generation. This creates challenges for some factors and simple factor strategies as the assumptions underlying their construction may no longer hold.
• Recovery/early expansion periods (Regime Two): Post-crisis phases when accommodative policy measures remain active and markets begin to normalise. These periods often present favourable conditions for systematic investment strategies as the return of ‘normal’ market functioning allows factor-based approaches to generate alpha.
• Mid-cycle expansion (Regime Three): Periods of stable economic growth, moderate volatility and balanced policy settings, featuring normal factor performance patterns and economic conditions that often represent the optimal environment for quantitative investment strategies.
• Late cycle/overheating (Regime Four): Periods of extended economic expansion with potential overheating concerns, policy tightening, and compressed risk premiums. These phases typically precede Quant Winters, featuring ‘everything rally’ dynamics where correlations between different investment approaches increase. The 2018-2020 Quant Winter perfectly illustrates the transition from this regime to crisis: beginning with late-cycle crowding in high growth stocks (a less popular quant exposure) and compressed risk premiums, it triggered a sharp reversal as crowded positions unwound through correlation spikes and volatility clustering.

Figure 1 illustrates how the economy tends to cycle through these regimes over the past three decades, with mid-cycle expansion (Regime Three) being the most common state, dominating much of the 1990s and periods in the 2010s. Late cycle/overheating conditions (Regime Four) appear during extended expansion phases, notably in the mid-2000s housing bubble, brief periods around 2011-2012, and most recently from 2023 onward. Crisis/recession periods (Regime One) are the rarest but capture iconic stress events including the dot-com crash (2000-2001), the GFC (2008-2009), and the COVID pandemic disruption (2020).

Figure 1: Classification of macroeconomic regimes (Aug 1994 – July 2025)

Problems loading this infographic? - Please click here

Source: Man Numeric, as of July 2025. Regime1: Crisis/recession periods, Regime 2: Recovery/early expansion periods, Regime 3: Mid-cycle expansion, Regime 4: Late cycle/overheating

Macro-dependent performance

This regime framework demonstrates that generic factor performance can be highly dependent on macro conditions across the entire factor universe. To illustrate this systematic relationship, we can examine how individual factors behave across different regimes.

Using momentum as an example, during crisis/recession periods (Regime One), momentum factors show consistent underperformance across all geographic regions, reflecting the fundamental challenge that cross-sectional trend-following equity strategies face during market stress periods when established trends reverse rapidly and correlations between securities spike.

Cross-sectional momentum factors also show dramatic regime dependence. Volatility in these factors increases substantially during recession periods compared to normal market conditions. This increased volatility reflects both the instability of underlying securities and the strategy’s amplified sensitivity to market dislocations. This compounds the challenge momentum strategies face during stress periods as higher volatility reduces risk-adjusted returns even when absolute returns are positive.

Figure 2a: Momentum² factor performance by macroeconomic regime

Long-short quintile spreads, sector-neutral

Problems loading this infographic? - Please click here

Figure 2b: Momentum factor volatility by macroeconomic regime

Problems loading this infographic? - Please click here

Source: Man Numeric, S&P Capital IQ, Jan 2005 – July 2025.

Multi-factor: not so diversifying after all?

Importantly, these patterns extend beyond momentum to encompass all major factor categories, revealing that traditional quantitative approaches share common vulnerabilities to macro regime shifts. This systematic sensitivity means that the benefits of multi-factor diversification may be overstated during periods when macro regime shifts create correlated headwinds.

Our analysis reveals that historically macro factors have explained significant portions of return variation in generic factors. This relationship shows pronounced cyclical patterns that fundamentally challenge traditional assumptions about factor diversification.

Figure 3: Percentage of generic factor return variance explained by macro factors and percentage of eVestment quant manager excess return variance explained by macro factors

Problems loading this infographic? - Please click here

Source: Man Numeric, eVestment, Jan 2007 – July 2025. Evestment US large cap quant manager universe. Man Group cannot guarantee accuracy of third party data.

During normal market conditions, macro factors typically explain 15-25% of generic factor return variation, as shown by the navy blue line in Figure 3, leaving substantial room for idiosyncratic factor performance and diversification benefits. However, during stress periods, this percentage can spike dramatically — reaching peaks of over 40% during the GFC around 2009 and again during the 2018-2019 period. These substantial increases in R-squared values during Quant Winters demonstrate that macro factors often become the dominant driver of factor returns (putting bottom-up company fundamentals in the backseat). This occurs precisely when diversification is needed most, fundamentally altering the risk-return characteristics of factor-based strategies during those periods.

This pattern of increased macro sensitivity during stress periods becomes even more concerning when examining the crowding phenomenon among quantitative investment managers, as illustrated by the aqua blue line representing US quant managers from eVestment data. Notably, macro factors consistently explain a larger proportion of variation in live manager returns compared to generic factor returns. This difference likely reflects several aspects of real-world portfolio construction that amplify macro sensitivity beyond what pure factor returns would suggest.

Constraints and timing

Quant portfolios are constructed with sector constraints, position sizing rules and risk budgeting frameworks that can inadvertently concentrate macro exposures. For example, liquidity constraints and transaction cost considerations often push managers toward larger, more liquid stocks that tend to have higher macro betas and more correlated responses to systematic shocks. Many managers also employ similar risk management overlays such as volatility targeting that trigger synchronised portfolio adjustments during stress periods, effectively creating common macro exposures that don't exist in the underlying factors themselves.

The timing of factor exposures also matters – managers who rebalance on similar schedules or respond to similar signals may inadvertently synchronise their macro sensitivities even when their underlying factor philosophies differ. Additionally, optimisation techniques commonly used in portfolio construction may amplify certain macro exposures when seeking to maximise risk-adjusted returns or minimise tracking error against common benchmarks.

These portfolio construction dynamics manifest clearly in the aqua blue line's behaviour during crisis periods. When macro sensitivity increases – as evidenced by the pronounced spikes during both the 2008-2009 GFC and the 2018-2020 Quant Winter, with peaks exceeding 50% – much of the diversification benefit that investors expect from allocating across multiple quant managers disappears. The synchronised increase in macro factor explanatory power reveals that despite apparent differences in methodologies and factor emphases, traditional quant approaches have historically become increasingly similar in their fundamental macro exposures during stress periods. As these shared macro sensitivities dominate returns precisely when diversification is most valuable, the result is reduced portfolio-level diversification benefits and increased systemic risk during Quant Winters.

Preparing for future market cycles

Some quant managers have addressed the core structural vulnerabilities behind past Quant Winters, simultaneously creating new sources of alpha generation and risk management. This transformation encompasses three dimensions that collectively represent a new approach to systematic investing, each building upon the others to create a more robust and resilient framework:

1. Enhanced diversification through alternative data and novel data science techniques

The expansion of alpha capabilities represents the most visible aspect of this transformation. Recognising that traditional factors' reliance on conventional financial data creates systematic vulnerabilities to macro regime shifts, the industry has embraced alternative data sources that operate independently of traditional financial metrics and macro sensitivities.

As detailed in our original paper, the number of alternative data models has nearly tripled in recent years, spanning diverse datasets including geolocation and foot traffic data, patent filing information, credit card transaction data, satellite imagery, and social media sentiment analysis. These sources provide real-time insights into business fundamentals before they appear in financial statements. This creates temporal advantages and independence from traditional accounting frameworks that reduce susceptibility to macro sensitivities.

Unlike traditional financial statements, which might respond synchronously to macro shocks, alternative data captures idiosyncratic, company-specific information in near real-time. For example, satellite imagery of retail parking lots or credit card data can reveal individual company performance that diverges from broader retail sector trends. Additionally, the high frequency, breadth, and diversity of these sources mean managers are less likely to converge on similar macro exposures, further reducing the synchronised sensitivity observed in traditional quant approaches.

Machine learning is the second pillar of this diversification strategy. These techniques capture complex, non-linear relationships in high-dimensional datasets where traditional statistical approaches struggle, providing pattern recognition capabilities that extend far beyond conventional analytical capacity.

Correlation analysis offers clear evidence of this diversification’s effectiveness. As previously documented, correlations between Numeric’s current alpha models and those from the 2018-2020 Quant Winter average only 40-50% across regions, demonstrating meaningful differentiation. For context, correlations between different systematic approaches can exceed 70-80% due to shared reliance on traditional financial data. This lower correlation reflects a substantively more diversified model framework.

Figure 4: Correlation between backcast alpha and as-was-alpha³

Problems loading this infographic? - Please click here

Source: Man Numeric, as of 31 December 2025. Hypothetical past model performance is not indicative of future results.

2. Dynamic factor selection and model combination

Building on the foundation of diversified alpha sources, the evolution from static to dynamic factor combination represents a fundamental shift in systematic investment philosophy. Recognising that static approaches create systematic vulnerabilities during regime transitions, the industry has developed adaptive methodologies that respond to changing market conditions. This addresses the core weakness of predetermined factor weights: the assumption that optimal combinations remain stable over time.

Figure 5: Generic quant combo versus an optimised factor-selection model

Problems loading this infographic? - Please click here

Source: Man Numeric, S&P Capital IQ, June 2005 – July 2025.

Traditional quantitative management approaches combine factors using static weights based on backward-looking optimisation, creating vulnerabilities when market conditions change. To demonstrate these limitations, we constructed a 'Generic Quant Combo' with static weights of 40% Value, 30% Momentum, and 10% each in Capital Efficiency, Earnings Quality, and Size. Representative of traditional processes, this static combination cannot adapt to evolving factor relationships or macro environments. Instead, it maintains predetermined exposures that become disadvantageous during regime transitions, creating unnecessary volatility while failing to capitalise on well-positioned factors.

In contrast, our dynamic approach utilises machine learning and macro-informed weighting to create a forward-looking, adaptive factor combination. This methodology continuously evaluates changing market conditions and factor relationships to optimise combinations in real time, moving beyond backward-looking metrics to embrace predictive analytics and regime-aware positioning. Consequently, our Factor Selection model demonstrates notably smaller drawdowns during previous Quant Winters and greater resilience. This reflects its ability to reduce exposure to underperforming factors while increasing allocation to those better positioned for the specific macro environment.

3. Macro regime resilience

Perhaps the most critical evolution in systematic investing is the development of macro regime resilience – directly addressing the core vulnerability behind previous Quant Winters. Having established diversified alpha sources and dynamic factor selection, this final dimension focuses on creating approaches that maintain effectiveness across different macro regimes.

This does not necessarily require explicit regime detection or regime adaptive modelling. Rather, it involves constructing alpha sources and factor combinations that are inherently more robust to macro regime changes through their fundamental design and diversification properties. This capability represents a key aspect of the systematic investing evolution, combining diversified inputs with robust methodologies designed to create more inherently resilient investment approaches.

Traditional approaches often treated macro sensitivity as an unavoidable characteristic of factor-based strategies, accepting that certain macro environments would create systematic headwinds. This reflected the limitations of static, backward-looking methodologies that could not adapt to changing conditions. However, our analysis demonstrates that systematic approaches can be designed to be effective across different macro regimes. Through a sophisticated understanding of regime dynamics, adaptive positioning strategies can respond to changing macro environments.

Figure 6: Annualised information ratio by macro regime

Problems loading this infographic? - Please click here

Source: Man Numeric, S&P Capital IQ, Jan 2005 – July 2025. Hypothetical past model performance is not indicative of future results.

Comparing annualised information ratios across macro regimes provides clear evidence of this evolution and its increased resilience. The analysis tracks three different alpha combinations that represent different stages in the evolution of systematic investing:

1. Generic Quant Combo: representing traditional static factor approaches
2. As-Was-Alpha: representing live alpha during previous Quant Winters
3. Backcast Alpha (calculating backwards from a desirable future point to determine its feasibility): representing current alpha models incorporating all three dimensions of transformation

The generic quant combination exhibits the characteristics that made traditional approaches vulnerable. It demonstrates the lowest average information ratio and largest difference in directional performance between regimes. This high directional dispersion indicates highly differentiated outcomes depending on macro conditions, creating systematic periods of underperformance that reflect the macro sensitivities identified in our earlier analysis. The as-was alpha shows meaningful improvement over the generic combination with reduced regime dependence, though it still exhibits some vulnerability to macro regime shifts.

As Figure 6 shows, current backcast alpha demonstrates strong and consistent performance across all four macro regimes, providing compelling evidence that systematic approaches can maintain effectiveness regardless of macro conditions. This regime resilience addresses the core vulnerability that created previous Quant winters stemming from reduced macro sensitivity through alternative data sources, enhanced diversification across uncorrelated alpha sources and the incorporation of sophisticated understanding of macro dynamics into the alpha modelling process.

Conclusion: a foundation for the future

While it remains to be seen whether "this time is different”, the opportunity for transformation is undeniable. The systematic investing industry now has the tools to address the core vulnerabilities behind previous Quant Winters.

Some managers have evolved their approaches to systematic investing to address vulnerabilities that were an Achilles heel in past market gyrations driven by macro volatility. By embracing a more sophisticated, adaptive, and resilient methodology, they have built a strong foundation for navigating future conditions. We believe this evolved quantitative approach is better suited for the challenges ahead.

The substantial decrease in correlation between current and historical alpha sources demonstrates how some managers have evolved beyond the limitations of traditional factor-based methodologies. This shift results from the development of genuinely uncorrelated return drivers using alternative data and advanced modelling. Our performance analysis confirms that these systematic approaches can maintain effectiveness across different economic environments, fundamentally addressing the regime dependence that created historical vulnerabilities.

This evolution represents more than technological advancement; it reflects a deeper understanding of market dynamics and systematic risk. The integration of alternative data, machine learning, and regime awareness has created what we believe is a new foundation for systematic investing – one that maintains the benefits of quantitative approaches while finally addressing their historical vulnerabilities.

1. See: https://www.man.com/insights/under-the-macroscope
2. Momentum factor defined as 39-week price momentum lagged by four weeks from S&P Capital IQ Alpha Factor Library.
3.As-Was-Alpha: Representing live alpha during previous Quant Winters. Backcast Alpha (calculating backwards from a desirable future point to determine its feasibility)

All data Bloomberg unless otherwise stated.