New Improvements to Riskalyze’s Methodology for Bonds

By: Michael Page
Director of Risk and Methodology, Riskalyze

Riskalyze has always had a strong commitment to getting risk right with deep methodology. We’ve talked about that track record extensively at our Fearless Investing Summit in 2019, and in the Methodology section on our website.

Our Risk and Analytics team is continually doubling down on that effort, constantly working to develop new ways to improve our analysis across different asset classes and investment solutions. Today, we’re excited to share a new round of improvements we’ve made to analyzing the risk of bonds.

We had three major themes for these enhancements: customizing our methodology for different types of bonds, increasing the precision of our analytics for all bonds, and factoring in credit risk.

First, let’s talk about customizing for different bond types. Over the last few years, we’ve significantly expanded our coverage from 30,000 to more than 500,000 individual bonds on the platform. While our initial methodology was optimized for coupon-bearing bonds, this expansion meant we needed to improve our approach for their zero-coupon and perpetual cousins.

Let’s use perpetual bonds and duration as an example. If we were calculating the Macaulay Duration for many bonds, you’d just find the weighted average time of payment until maturity. However, perpetual bonds have no maturity date and their payment weights approach zero as time approaches infinity. Similar situations arise for several other key bond metrics across different types of bonds, and we’ve built new methodology to customize how we handle those for zero-coupon and perpetual securities.

Second, we wanted to increase the precision and accuracy of our analytics for all bonds. Every one of the bond metrics we use in our analysis — like Accumulated Interest, Dirty Price, Maturity, Yield to Maturity, Macaulay Duration, Modified Duration, Convexity, PV01, DV01, and more — are now calculated to eight decimal places for accuracy.

And finally, we wanted to find a way to incorporate credit risk into bond analysis. This was a big challenge for our team because, on the one hand, credit risk is clearly an important factor to assess the risk of holding a bond. But on the other hand, simply relying on credit ratings as a proxy for credit risk has been a dubious proposition in the past.

That’s because credit ratings are largely subjective, and as we all know from the 2007-08 financial crisis, that subjectivity led to a massive disconnect between credit ratings and the actual credit risk of many securities. Even the big ratings agencies say in their disclosures that their credit ratings “should not be viewed as assurances of credit quality,” and another said, “credit ratings are opinions on relative credit quality and are not a predictive measure of specific default probability.”

As a company whose risk methodology has been built on the foundation of objectivity, we really struggled with leveraging subjective credit ratings in the equation, and we ultimately found a much more objective path — calculating the implied probability of default based on the market price of each bond — to ensure that a high quality of measure of credit risk is reflected in the overall analysis of that bond and your client’s entire portfolio.

This new methodology is live in your Riskalyze account already, and it’s a free upgrade for customers on all of our service plans.

We’re excited to deliver these incredible advances in methodology for calculating and analyzing the risk in individual bonds. With expanded methodology for different bond types, greater precision in calculation, and an objective approach to evaluating credit risk, Riskalyze has doubled down on putting the power in your hands to objectively evaluate all kinds of portfolios for clients and prospects alike.