Following Axioma's risk model and performance analysis acquisition in August of 2005, Axioma fully researched and updated risk models will be delivered to the market soon. The following questions and answers provide insights on Axioma's new models and the unique factors that differentiate them from all others on the market today. [Click here for Frequently Asked Questions.]
What makes Axioma Robust Risk Models different from other risk models and why are they better?
Axioma Robust Risk Models are unique in the marketplace because they are the only models that directly take uncertainty into account. For years, the key innovation that made Axioma Portfolio a success involves its ability to exploit uncertainty in alphas for portfolio construction. This gives our clients the power to capitalize on the notion of uncertainty in their return estimates. As risk models contain uncertainty as well, we are leveraging that same innovation in our new risk models. There is nothing comparable to Axioma's risk models on the market today.
In addition — and this is very significant — Axioma Robust Risk Models were specifically designed to function as an input to Axioma Portfolio, our portfolio construction and rebalancing solution. While clients can use any risk models they wish in conjunction with Axioma Portfolio, using Axioma Robust Risk Models greatly simplifies the implementation process for clients and improves the overall results of the investment strategy.
What specific features differentiate Axioma's risk models?
Axioma's risk-model philosophy is based on the premise that risk is inherently unobservable and thus cannot be measured precisely. As Aristotle noted nearly 2,300 years ago, "It is the mark of an educated person to look for precision only as far as the nature of the subject allows." There will always be uncertainty in any risk-model estimation. Because of the difficulty of the problem, it has been fundamentally ignored — until now.
At the heart of the matter is a fundamental question: Can that knowledge of uncertainty be used to our advantage? By developing methodologies that explicitly take into account uncertainty in model estimation, Axioma's products trade more upon information rather than the noise in estimation. This results in better exceptional and intuitive portfolios.
We believe that risk should be reported and estimated transparently, consistently and in a timely manner.
Transparency is essential. Axioma has built a glass box, not a black box. That is the only way for all the constituents-asset owner, portfolio manager, research, CIO, etc.-to understand the risk dynamics of the portfolio.
Consistency in risk reporting and estimation also is crucial. The asset owner must understand risk reports. For a portfolio manager, talking to the asset owner in one language and using a different language to communicate internally leads to misunderstandings and problems.
And timeliness is critical. Conventional risk models estimate and report monthly. Estimating risk monthly produces severe changes in estimated risk from the day prior to estimation to model estimation day. This results in unintuitive and hard-to-explain jumps in predicted volatility each month. If portfolio performance is monitored daily, then risk should be, too. This results in both smoother and more timely estimates of portfolio risk.
In addition to transparency, consistency and timeliness, our risk-model philosophy embraces one more goal, which is that risk should be used effectively to make better decisions in money management. More to the point, we believe that investment professionals can make better decisions and obtain a meaningful advantage by using an integrated approach to risk modeling and optimization.
What led to the innovations built into Axioma Risk?
When Axioma began to explore the risk model space, our initial research was aimed at determining how well different risk models performed. That research led to an unexpected discovery. We found that risk models were typically tested against either benchmarked or random portfolios, but almost never against optimized portfolios. In addition, little had been published on the subject, so it was clear that Axioma had ventured into uncharted waters.
After extensive further research, we discovered that conventional risk models underestimate risk for optimized portfolios. This occurs because of flaws in the ways that risk models interact with portfolio optimization engines. Axioma realized that if we could get the optimizer to take advantage of the knowledge in the risk model, then we could build a superior tool that would enable investment professionals to achieve improved performance.
Why did Axioma decide to develop its own risk models?
The impetus came from our clients. Axioma successfully innovated the portfolio construction space with Axioma Portfolio and our clients asked us to do the same thing for risk models. And it made sense because risk models are a natural extension of Axioma's product line and an integral part of the portfolio construction process. We now provide our clients with a fully integrated portfolio-management platform that combines risk models, portfolio construction and performance attribution in one package. Axioma's clients no longer need to rely on the risk models of other providers, most of which are based on technology that's 20 years old.
Axioma acquired the risk model and performance attribution components of Goldman Sachs Asset Management's Portfolio Analysis and Construction Environment-better known as PACE-in September 2005. Why did Axioma choose the PACE system as the basis for Axioma Robust Risk Models?
After a rigorous analysis of the marketplace, we found that the PACE solution was by far the best. PACE is a time-tested system that was built by portfolio managers for portfolio managers. PACE has been used extensively by GSAM and includes country, regional and global models, and more than a decade (1994) of history.
PACE's data warehousing components were a major factor in our decision. Data accuracy is key to any risk model. The PACE acquisition included all the data capturing and scrubbing processes, the risk-model construction and update processes, the database structures and scripts, etc.
In addition, PACE had been used for years in conjunction with Axioma's portfolio construction software, so it was an excellent fit.
