A collection of our white papers and peer-reviewed research articles related to macro / multi-asset investor behavior, hedging, risk regimes, liquidity risk, private assets, portfolio construction, and more.
By Megan Czasonis, Ding Li, Grace (TianTian) Qiu, Huili Song, and David Turkington
We present an adaptive method for constructing an inflation robustness equity factor which outperforms attribute- and sector-based strategies during 65 inflation events from 2008 to 2025.
We introduce a novel framework to identify which equities will prove robust and which will falter during episodes of inflationary stress. By analyzing firm-level attributes and comparing them to historical performance during inflation events, we construct stock-level inflation robustness scores that capture nonlinear, conditional relationships between stock characteristics and inflation sensitivity. Tested across 65 inflation events from 2008 to 2025, these scores reliably predict stock performance, with a long-short portfolio delivering 9.1% annualized outperformance during inflationary periods. The robustness factor outperforms traditional attribute- and sector-based strategies and remains statistically significant even after controlling for sector and factor exposures. This adaptive methodology enables investors to build inflation-hedged portfolios with greater precision than conventional equity benchmarks.
By Mark Kritzman, Sébastien Page, and David Turkington
Our 2010 research paper – recently flagged by the Financial Analysts Journal as among the 23 most influential articles in its 80 year history—defended how data-drive portfolio optimization adds value when it is thoughtfully applied.
Investors rarely aspire to mediocrity, yet critics love to dismiss the quest for an “optimal” portfolio as quixotic. In an article published in the Financial Analysts Journal 15 years ago, we challenged the notion that equal-weighting (the infamous 1/N rule) beats optimization. That claim implies all information including expected returns, risks, and correlations, is worthless. By digging deeper we found a reason those studies generated such a counterintuitive conclusion: they used naïve five-year return extrapolations as forecasts. When we applied simple, sensible expectations instead, mean-variance optimization delivered clear out-of-sample value across assets, industries, factors, and stocks. Optimization is far from perfect and it deserves to be extended to address real-world complexities, but it is a far better starting point than just giving up.
By Alexander Cheema-Fox, Edward S.Cuipa III, and Robin Greenwood
Institutional investor flows can help distinguish between flights to safety and flights from carry, revealing that the CHF, JPY, and USD play distinct roles depending on interest rate differentials and whether market stress is driven by macroeconomic uncertainty or funding pressure in currency markets.
We examine the roles of safe-haven and funding currencies during market stress by analyzing how institutional investors reallocate across currencies during crises. Using proprietary FX and local-currency sovereign bond flows from State Street, we show that while the CHF, JPY, and USD all act as safe havens, investor flows and currency behavior differ substantially across crises depending on interest rate differentials and whether the episode is driven by macroeconomic uncertainty or by a carry trade unwind. A panel regression framework, leveraging FX forward and sovereign bond flows, reveals that some crises, like the 2008 Global Financial Crisis and COVID-19, reflect true flights to safety, whereas others, such as the 2024 carry crash and Liberation Day, are characterized as flights from carry involving CHF and JPY inflows and muted or negative USD flows. These findings underscore the importance of distinguishing between safe-haven and carry dynamics in currency markets and suggest that institutional investor flows offer a powerful tool for classifying crisis episodes.
By Megan Czasonis, Mark Kritzman, and David Turkington
As the race to design sophisticated data analytics continues, we show why relevance-based prediction offers an ideal way to measure the importance of an input variable to a prediction.
T-statistics act as a hallmark for rigor by pinpointing the effect of a single variable and distinguishing signal from noise. However, they have significant limitations: (1) t-stats do not capture ‘shared’ information, (2) t-stats are not prediction-specific, and (3) t-stats only consider linear relationships. In a recent paper, we introduce an alternative method, called Relevance-Based Importance (RBI), which measures the importance of every variable to the reliability of every individual prediction. RBI recognizes that it is almost never the case that a variable is always important, or that it is never important. Rather, it's more likely that variables are sometimes important, depending on the circumstance. We show that RBI brings the virtues of t-statistics but also adapts to each unique situation, making it robust to complexities where t-stats fall short.
Fresh off his 2025 James R. Vertin Award win, State Street Associates’ Founding Partner Mark Kritzman joins David Turkington, head of State Street Associates, for a discussion on the key themes that have underpinned Mark’s body of research, the importance of connecting research with intellectual curiosity, and the key qualities of research that can stand the test of time.
By Megan Czasonis, Yin Li, Huili Song, and David Turkington
Our innovative "interrogation" method detects unreliable machine learning predictions in advance, overcoming limitations of the traditional cross-validation method.
We introduce a new method called "interrogation" to warn when a machine learning model has underfit or overfit a data sample, offering a more efficient alternative to traditional cross-validation. Unlike cross-validation, which can be cumbersome and computationally expensive, interrogation evaluates models trained on all available data by breaking down their prediction logic into linear, nonlinear, pairwise, and high-order interaction components. This method successfully identified near-optimal stopping times for training neural networks without using validation samples, boosting confidence that models are well-calibrated and can perform reliably on new data. Interrogation is model-agnostic, providing transparency and reliability even for black-box models.
By Alexander Cheema-Fox, Megan Czasonis, Piyush Kontu and George Serafeim
We explore the world’s first set of financial accounting data on firms’ sustainable activities.
Though sustainable investing has grown in popularity over the past decade, measuring sustainability remains a key challenge. Investors often rely on environmental criteria—such as analyst ratings and carbon emissions—that are insufficient or rely on qualitative analysis. However, for the first time, with the advent of the EU’s Taxonomy for Sustainable Activities, investors have access to financial accounting data that follows standardized and transparent criteria for quantifying the percentage of a firm’s revenues and expenditures that align with sustainable activities. In a recent paper, we explore this novel dataset for a cross-section of large European firms, documenting patterns and analysing how firms’ aligned activities relate to fundamentals and environment ratings. We find that the EU Taxonomy data provide information that is distinct from existing sources and offers insights that can help investors and regulators, alike.
By Megan Czasonis, Mark Kritzman, and David Turkington
We show that relevance-based prediction offers an elegant solution to the problem of incomplete data, preserving valuable information and enhancing prediction reliability in a way that is not possible using traditional models.
When setting out to form data-driven predictions, it’s common to encounter incomplete information, such as a time series with shorter history lengths or observations with missing data. Traditional methods for addressing this challenge either discard valuable data or manufacture replacements based on limiting assumptions, leading to unreliable results. We propose a novel technique called Relevance-Based Prediction (RBP), which elegantly navigates the pitfalls of missing data by retaining more information and accounting for the relative importance of observations for which only partial data is available. We show that RBP offers an elegant solution to the problem of incomplete data, preserving valuable information and enhancing prediction reliability in a way that is not possible using traditional models.