As a portfolio management tool, Markowitz optimization has disappointed most investment managers. Regardless of the theory’s elegance, it has not worked well in practice. The likelihood of realizing an efficient frontier with any optimized allocation in the real world is often quite low. The failures of this standard model approach have led to a search for alternative asset allocation solutions, such as risk budgeting. New allocation solutions may ultimately create true breakthroughs for portfolio management. Most innovations are fads that catch the attention of managers for only a short period; nevertheless, the industry needs to understand new techniques to separate what is useful from what may be a passing fancy.
The risk parity approach is viewed by a growing number of managers as a viable portfolio structuring alternative to older optimization methods. Until now, however, there has not been a comprehensive work on the subject. Risk parity started as a simple heuristic for forming “all-weather” portfolios across major asset classes but has been increasingly used and touted by money management firms as an approach to diversification and better portfolio construction. The risk parity approach has been marketed to attract investors who have been disappointed with some of the classic asset allocation approaches, such as the naive 60/40 equity/bond capital-weighted portfolio benchmark.
Risk-based portfolio alternatives performed particularly well during the financial crisis, when low-volatility, “safe” assets were preferred. With the strong equity performance over the past few years, however, the glamour of a risk parity strategy that allocates away from high-volatility equities in favor of bonds has faded. Asset allocation solutions fitted to the recent past, however, are likely to be the failures of tomorrow. They certainly will not be successful in all environments. Hence, there is a strong need to understand the fundamentals of this approach and its potential pitfalls.
Risk budgeting is the general term for allocating risk across a set of assets, whereas risk parity is a special case of allocating risk equally across an asset set. Many investors do not make this distinction. Although a risk-budgeting approach is fundamentally at odds with classic optimization tools based on both expected return and risk, the intuition of allocating only by risk has a strong appeal. Whether risk budgeting will become a primary asset allocation process is unclear; nevertheless, its prevalence as a core tool in investment managers’ toolkits is likely to grow, and it is necessary for managers to understand the pros and cons of this approach in order to more effectively advise their investors.
Risk parity is still in its infancy, so a book that covers the foundational material and provides breadth on the practical use of risk budgets is welcome. Introduction to Risk Parity and Budgeting captures the key principles of this process and will further establish risk parity as a viable alternative portfolio management tool. This highly technical work presents all the math, statistics, and practical applications necessary for understanding the risk parity framework. The book is divided into two parts: a detailed theoretical section that marches through the development of optimization and risk parity techniques and a section that applies risk parity to various asset classes and problems. There is also a technical tutorial and an accompanying website for those who want to delve deeper into the problems of structure and application.
A review of the general return-to-risk optimization model through theory and simulations lays the foundation for why the efficient frontier framework is flawed. Optimization failures have occurred for a long time, and effective alternatives have been hard to find. Mean–variance optimization is problematic because it provides results that are unstable and highly sensitive to model inputs. Small changes in return, volatility, or correlation lead to significant differences in allocation weights. Mismeasurements of variable inputs are rampant. Given the failure of unconstrained optimization, researchers have followed a number of potential solution paths. These alternatives include constraining the problem in order to minimize mistakes in the framework and using statistical and heuristic methods to create or impose stability. A recurring result is that a minimum-variance portfolio or one that assumes that returns are equal creates a more stable and efficient portfolio. The proper response is clear: Drop return estimation from the analysis and focus on risk, which provides the foundation for risk budgeting.
Author Thierry Roncalli’s review of optimization provides context for comparing risk-budgeting with other approaches. Risk parity can be effectively measured against the minimum-variance and equal-weighted models. The theoretical section on risk parity effectively builds from simple cases with two assets to more complex analysis. The math gets increasingly difficult, but there are enough examples, tables, and simple analysis to allow readers to appreciate the basics of risk allocation schemes. Risk budgeting may not be a perfect alternative, but it provides a level of stability that has eluded traditional optimization alternatives.
Weighting portfolio assets by risk is similar to weighting by factors other than capitalization, which is a growing index fad. Risk budgeting as an alternative weighting scheme is the subject of one chapter. Roncalli presents the concept of inefficient capitalization-weighting schemes and relates them to an implicit trend-following mechanism. A risk-weighting system is uncorrelated with more traditional index systems and eliminates the cap-weighted biases. The empirical evidence is inconclusive, however, on whether risk-based weighting dominates when costs, turnover, and management effort are included.
Roncalli applies risk-budgeting tools to a number of asset-class problems. A discussion of the use of risk budgeting for bond portfolio management shows the advantage of moving away from a traditional debt- or GDP-weighted mix. Roncalli demonstrates how an approach that targets credit risk spread volatility, for example, would have prevented the EU sovereign debt crisis and would have outperformed other weighting schemes. A debt-weighted benchmark in a financial crisis is an easily beaten straw man. An alternative asset example that features commodities and hedge fund strategies again makes the case for risk weighting as a viable alternative to traditional capitalization weights. Once again, however, empirical evidence on the superiority of this risk parity approach is mixed. The book ends with the multi-asset-class portfolio allocation problem, which has been a marketing success since the global financial crisis began, but a balanced analysis shows there are still potential drawbacks based on the environment faced.
Introduction to Risk Parity and Budgeting provides a comprehensive presentation of risk budgeting, but it does have the feel of a very long academic paper, with numbered examples and remarks. Any academic would be comfortable with this presentation style, but it makes it difficult for practitioners to see how the technique can be applied in practice. The presentation could be enhanced with better labeling of graphics and tables that force the reader to skip around for reference information. Additionally, the standard shorthand of referencing past research leaves readers wanting more details on previous studies. These are classic issues with any academic work that intends to cross over to a wider practitioner audience. Nevertheless, Roncalli’s intent may have been to write a seminal academic finance book and not one for portfolio managers. Unfortunately, he misses an opportunity to connect with this wider audience on a critical portfolio management topic.
There are some minor content problems in the book. It does not fully review the issue of risk parity being a heuristic with limited theoretical foundation, nor does it fully address the critical question of leverage with risk parity. The approach of equal risk weighting results in an overall risk level that can be much lower than those of traditional approaches. Consequently, leverage is a necessary tool for setting risk targets. Also important is the issue of risk parity overweighting less volatile assets, which paradoxically have shown a tendency for higher risk-adjusted returns. A high-volatility avoidance strategy is viable but should not be a required assumption for risk parity to work.
This important book provides a complete framework for describing risk parity and risk budgeting while contrasting these approaches with traditional optimization methods. Roncalli does an excellent job of comparing simulations of strategies and approaches. A manager who reads this book—instead of reading a set of Wall Street research papers, consultant opinions, and some practitioner work—will be able to effectively discuss and evaluate all issues surrounding this important directional change in portfolio management. Through being evenhanded and generally unbiased in his conclusions, the author shows that risk parity is not a perfect solution but is an approach with significant merit that should be considered an important portfolio-structuring alternative and not a fad.
—M.S.R.