Library

PJRmi is an open source software tool that we built to facilitate seamless interoperation between Python and Java. Rather than creating systems with separate components, PJRmi allows developers to write applications in a blend of languages, working with each where it makes the most sense. In this piece, we explore the design principles underlying PJRmi and share aspects of its implementation and illustrative workflows.

For most of 2024 and into 2025, purchasing S&P 500^® exposure via futures proved considerably more expensive than buying it in cash. We argue this apparent “mispricing” reflects efficient pricing of the leverage embedded in index futures. In short, we believe demand for balance sheet that finances long U.S. equities significantly outstripped available supply. We consider how changes in those supply-demand dynamics could normalize that financing market.

Surveys of professional forecasters assign low probabilities to extreme inflation and growth outcomes for the United States in 2024, suggesting near certainty of a “soft landing” in the year ahead. We believe these forecasts are too narrow, and more generally that historical distributions can be a useful baseline when considering the range of plausible outcomes. History reminds us that the distribution of inflation outcomes is wide and suggests that the potential for more robust growth should not be written off.

Over our firm’s history, we’ve developed sophisticated optimizers that we deploy across systematic and discretionary strategies. In this piece, we share some lessons we’ve learned from interacting with optimizers as part of discretionary investment processes. From calibrating trading speed to accounting for tail risks, we explore how an optimizer’s comparative advantages can help us improve our intuition as investors in pursuit of better portfolio outcomes.

Investors seeking to reduce an equity portfolio’s exposure to carbon emissions often implement exclusions or concentrated underweights at the stock or sector level. In this piece, we examine tradeoffs related both to concentration of emissions and to correlations between emissions and equity style factors. We show that those tradeoffs have implications for portfolio construction and active risk. Our analysis suggests that a thoughtful approach to portfolio construction can help mitigate active exposures and free up risk budget to add value beyond carbon reduction objectives.

Although the U.S. Federal Reserve appeared slow to grapple with accelerating inflation in early 2022, market-implied inflation expectations generally remained grounded. In this piece, we argue that the Fed’s eventual shift to an aggressive monetary policy stance was critical to that anchoring, offering an analysis based on market indicators that estimates the effect of monetary policy on inflation expectations. Our analysis supports the view that absent aggressive Fed policy, inflation expectations may well have become unanchored, with far-reaching implications for asset prices and correlations.

Originally published by the Brookings Institution’s Hutchins Center on Fiscal and Monetary Policy, this piece offers a resource (including a modeling tool available on GitHub) for projecting the structure of U.S. Treasury debt based on assumed funding needs and an assumed path of gross issuance amounts.

In the early 2000s, equity returns—whether at the individual stock, sector, style factor, or index level—showed only modest differences in their cross-sectional sensitivity to interest rate changes. Twenty years later, equity portfolios might be expected to react to changes in interest rates with much greater variation depending on their composition. This piece explores the origins of that trend and considers its implications for portfolio management.

Aggregate ESG ratings for U.S. companies have increased in recent years. This piece investigates whether that increase is the product of structural factors, such as evolving market composition or ESG scoring rules, or if it reflects changes in company behaviors. We find that ratings have increased even after controlling for important structural factors, and present evidence that company-specific changes may have helped drive that trend.

Google LLC’s Community Mobility Reports represent a valuable data source for use in economic analysis, but in the absence of seasonal adjustments—made more difficult by the short history of the data series—the raw data can be misleading. This piece outlines a novel approach to seasonally adjusting mobility data and sheds new light on important trends in U.S. economic activity since the onset of the COVID-19 pandemic.

The application of quantitative research to the China A-shares market presents both significant promise and distinct challenges. In this piece, we argue that certain of the market’s structural features—in particular, the historical prevalence of trading halts—necessitate important active decision making during the research process that can have a meaningful impact on portfolio risk and results.

Pyflyby, developed by the D. E. Shaw group, offers developers a number of innovative utilities to mitigate the risk of interruption and improve efficiency when working in IPython and Jupyter. In this piece, originally published as a Quansight Labs blog post, Quansight introduces pyflyby and provides examples of its functions and features.

The residential mortgage-backed segment of the U.S. structured credit market illustrates how investment opportunities can arise when top-down thematic concepts converge with bottom-up asset mispricings. This case study focuses on the investment dynamics of government-sponsored enterprise credit risk transfer bonds, a category of securities that came under enormous pricing pressure at the outset of the COVID-19 crisis.

The dynamics of European non-core assets (“NCAs”) offer a window on the important roles that deal sourcing, data collection, value assessment, and stress testing can play in investment decision-making in the asset-backed space. This case study takes an in-depth look at a large portfolio of Spanish NCAs to highlight the impact those elements can have on the early stages of an investment process.

The correlation between U.S. stock and government bond prices shifted into positive territory in the first half of 2021. In this piece, we argue that changes in inflation expectations and perceptions of U.S. Federal Reserve policy are central to understanding that move in correlation, and we consider how those dynamics could play out over the longer term.

Jason Furman, Aetna Professor of the Practice of Economic Policy at Harvard University and a former top economic adviser to President Obama, joins Brian Sack, our director of global economics, for a wide-ranging conversation on the implications of fiscal policy expansion. Introduced by Darcy Bradbury, our head of public policy, Jason and Brian discuss key issues at the intersection of fiscal and monetary policy, as well as prospects for inflation.

For quantitative research involving time series data, it is important to avoid relying on data available only at a later point in time (T2) when generating results for an earlier point in time (T1). The ability to time travel through versioned data sets is critical for accurate and reproducible results. Quansight Labs and the D. E. Shaw group collaborated to develop Versioned HDF5, which allows users to implement versioning within large data sets. In this piece, adapted from a series of Quansight blog posts, Quansight introduces Versioned HDF5 and discusses the design and performance of data libraries using this system.

The adverse market events brought on by the 2020 COVID-19 pandemic represented a key test for government bonds as safe haven assets. Markets suffered a large negative shock at a time of low interest rates. This piece considers whether government bonds are likely to serve as an effective source of diversification going forward.

In September 2019, the Treasury repo market, through which more than $1 trillion flows every day, experienced a sharp spike in short-term interest rates, straining overnight funding markets. This blog post−co-authored by Brian Sack−proposes several ways that the Fed could improve the resiliency of its monetary policy operating system.

In the face of economic damage and structural changes following the global financial crisis, major central banks tested the lower limits of their nominal policy rates over much of the following decade. In this piece, we explore certain implications of the lower-bound constraint on financial markets and conclude that market pricing failed to fully account for this new paradigm.

In this paper, we contend that the negative stock-bond correlation regime observed over the preceding two decades has been driven largely by major central banks’ success in lowering inflation and anchoring inflation expectations. We confront that hypothesis with a combination of theoretical and empirical evidence and discuss potential implications of this correlation regime.

Even though Quantitative Easing (or “QE”) has assumed a prominent role as a policy instrument for central banks, the implementation of QE has differed significantly both across countries and over time. This Policy Brief—which was co-authored by Brian Sack, the D. E. Shaw group’s director of global economics—drew on the literature measuring the effects of QE as well as the historical experience in implementing QE to lay out a simple strategy for how to use it.

We argued that, while the emergence of large-scale shale gas production in the United States had greatly expanded natural gas (NG) supply and generally moderated long-term volatility, the impact of shale gas on the NG market was less uniform and more complicated than might have been suggested from readings of the popular press.

This Policy Brief—which was co-authored by Brian Sack—proposed a new operating framework to allow the Fed to conduct monetary policy while maintaining a substantially elevated balance sheet and abundant liquidity in the financial system. In particular, it argued that the Fed should set the interest rate at which it would offer over-night reverse repurchase agreements as its policy instrument and maintain the interest rate paid on bank reserves at the same level.

We believed a fresh look at 130/30 was warranted, as this investment approach could, when robustly implemented, improve risk-adjusted performance relative to long-only portfolios. The poor historical results of many 130/30 products, we argued, stemmed largely from implementation problems specific to certain managers or investment approaches rather than flaws in the theoretical foundations of 130/30.

This piece presented an examination of the two main paradigms in reinsurance—diversification and risk aversion—followed by an analysis of why our particular approach to reinsurance investments typically yielded a riskier, more concentrated, but we also think ultimately more attractive, portfolio than the strategies employed at the time by many of our peers.

The market for U.S. residential mortgage-backed securities (RMBS) took a turn for the worse in 2011, particularly in the subprime segment. We sought to bring some clarity to this corner of the capital markets by focusing on two overlooked dynamics: loan servicer behavior and the negative duration of subprime floaters.

We believe that managers of multi-strategy funds offer investors advantages they could not obtain from managers of single-strategy investments, including a better value proposition, better alignment of incentives, more efficient capital allocation, and better risk management. This paper first considered the structural advantages of the multi-strategy approach and then addressed some common criticisms.

After some spectacular successes at the height of the financial crisis, the macro hedge fund sector muddled along for much of 2009 and 2010. On balance, we felt optimistic about opportunities for macro at the time, finding the environment compelling on the basis of its offering broad (if not always deep) liquidity, coupled with a large amount of pricing inefficiency. In this piece, we attempted to account for the difference of opinion between our 2010 view of the risk-adjusted opportunity and the apparent consensus among the broader macro community.

In February of 2010, the U.S. Securities and Exchange Commission voted to implement a rule that reimposed constraints on the execution of short sales. Dubbed the “alternative uptick rule,” the rule has two major components, a “circuit breaker” and a “passive bid test.” We considered the alternative uptick rule in this entry of our Market Insights series.

This piece was an examination of aspects of leverage that merit consideration by investors evaluating levered investment portfolios. Put simply, folks think a lot about the question, “How much leverage?” but not very much at all about the questions, “Why is the leverage being used?”, “Under what terms and conditions is the borrowing being done?”, and “How is the leverage quantity being computed?”

In a financial crisis, it’s often said, “all correlations go to 1.” That may or may not be the case. More certain, however, is that crisis-induced dislocations will move various asset classes and market indicators in surprising directions. Covering much of the painful period of the most recent financial crisis, this piece examined the performance of a number of different asset classes from January of 2008 through June of 2009.

We took a closer look at some real-world examples of common investor mistakes across various classes of market participant, whether amateur or professional and whether banker, hedge fund manager, insurance company executive, fund-of-funds operator, high-net-worth individual, or mutual fund investor. We consider ourselves something of an authority on the subject of mistakes not only because we’ve seen others make them, but also because, over thirty-plus years, we’ve made our share.

During the financial crisis, cash-synthetic basis became so volatile and pervasive across a number of credit instruments that many took to referring to it simply as “the basis.” We discussed our approach to incorporating cash-synthetic basis in our analysis, including an examination of the relationship between cash bonds and related credit default swaps and interest-rate swaps.