The years 2018-2020 were bleak for quantitative managers. A shift in Federal Reserve monetary policy upended traditional quant factors, while a market frenzy for growth stocks led to a value drawdown.
Covid worsened quant investors’ plight, with fiscal stimulus and easy monetary policy further concentrating market winners and culminating in a momentum meltdown.
The quantitative investing landscape has enjoyed a remarkable revival since this ‘Quant Winter', but questions linger about whether it could experience another spell in the deep freeze.
To answer these questions, it is important to understand why Quant Winters occur in the first place.
Looking at past bouts of underperformance, our research identified two primary drivers: first, adverse macro environments and the sensitivity of traditional quant factors to them. This means factors can be vulnerable to regime shifts, policy changes and broader economic dislocations.
During stress periods, macro factors often become the dominant driver of factor returns, putting bottom-up company fundamentals in the shade.
Second, crowding – and not just in raw factor exposures. Our analysis of live manager returns shows that the portfolio construction process itself amplifies the problem.
Sector constraints, liquidity screens, and similar rebalancing schedules push managers into correlated positions, increasing macro sensitivity beyond what raw factor returns alone would suggest.
The result is that during Quant Winters, much of the diversification benefit investors expect from allocating across multiple quant managers disappears.
This tendency has led traditional quant approaches to become very similar in their exposures just when diversification is needed most.
To address these core structural vulnerabilities, in recent years some quant managers have sought to create new sources of alpha generation and risk management. These encompass three dimensions.
Alternative data
Recognising that a reliance on conventional financial data creates systematic vulnerabilities, the industry has embraced alternative data sources that operate independently of traditional financial metrics.
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.
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.
Machine learning is a key pillar here as its pattern-recognition capabilities far exceed conventional statistical approaches. This can reduce correlations to significantly below the 70-80% common between traditional systematic approaches, which have a shared reliance on traditional financial data.
A more dynamic approach
Traditionally, quant management combines factors using static weights, with set percentages for value, momentum and size, for example. However, this approach creates vulnerabilities during changes in macro regimes, maintaining predetermined exposures that become suboptimal and causing unnecessary volatility while failing to capitalise on well-positioned factors.
Today, some managers have developed factor selection models that can continuously evaluate changing market conditions and factor relationships to optimise combinations in real time, moving beyond backward-looking metrics to embrace predictive analytics.
As one component within a comprehensive investment process, these models can notably reduce drawdowns by adjusting exposure away from underperforming factors and toward those better positioned for the specific macro environment.
Macro regime resilience
Perhaps the most critical evolution in systematic investing is the development of macro regime resilience. This final dimension focuses on creating approaches that maintain effectiveness across different macro regimes.
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, systematic approaches can now be designed to be effective across different macro regimes, including periods of crisis/recession, recovery, expansion and late-cycle/overheating.
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.
Through a sophisticated understanding of regime dynamics, adaptive positioning strategies can respond to changing macro environments.
This 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.
A foundation for the future
While it remains to be seen whether ‘this time is different’, the opportunity for transformation is compelling. The systematic investing industry now has the tools to address its core vulnerabilities.
By embracing a more sophisticated, adaptive, and resilient methodology, some managers have built a strong foundation for navigating future conditions. We believe this evolved quantitative approach is better suited for the challenges ahead.
Our performance analysis confirms that modern approaches can maintain effectiveness across different economic environments, fundamentally addressing the regime dependence that created historical vulnerabilities.
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 flaws.
Ori Ben-Akiva is the director of portfolio management, and Valerie Xiang a portfolio manager, both at Man Numeric.