RIA Insights: Why It’s Time for Legacy Risk Models to R.I.P.

The good news? Technology has paved the way for new risk measurement tools that advisors and RIAs can use to better measure, communicate, and manage risk.

In this blog, we explain why legacy risk models fall short and the impact this has on investors. You’ll also learn how combining “fat tail risk” modeling with today’s advanced technology can produce a more accurate risk score that helps advisors provide the insight clients need to protect their portfolios during extreme market events.

Table of Contents:

• The not-so-hidden flaws embedded in legacy financial risk modeling
• ‘The Gaussian Distribution is pretty much useless for risk estimation’
• The major flaw in the Value-at-Risk (VaR) method—it doesn’t address worst case scenarios
• Why correlation matrix limitations should also be considered
• Underestimating tail risk can compromise portfolio stability and client trust—it’s time for a new, improved risk score
• Enter QuantumRisk™: A proprietary risk engine designed to help advisors assess portfolio exposure during market shocks.

The not-so-hidden flaws embedded in legacy financial risk modeling

Most foundational finance models are based on standard deviation Gaussian Models like the bell curve that use normal distribution as the main statistical pillar. Why is this approach flawed? The reason why standard deviation fails in investing is because risk practitioners who rely on standard deviation use backward-looking models that underestimate real-world volatility.

Experts in the fields of finance and mathematics have decried the viability of the Gaussian Model in calculating risk and financial planning as far back as the 1960s. Amplify Director of Investment Research Ron Piccinini, Ph.D., explained this in his 2023 white paper on risk scoring.

‘The Gaussian Distribution is pretty much useless for risk estimation.’

According to Dr. Piccinini, “Sadly, the Gaussian Distribution simply does not work well (or at all) to describe market or economic data and is pretty much useless for risk estimation. This has been known in the financial literature since the 1960’s. For example, in 1963, Mandelbrot (one of the most gifted mathematicians in history) shows how poorly the Gaussian Distribution fits financial data. Eugene Fama, the father of Efficient Market Hypothesis, also concludes in his doctoral dissertation, submitted in 1964, that the probabilistic distribution of stock prices is heavy tailed.”

The major flaw in the Value-at-Risk (VaR) method—it doesn’t address worst case scenarios

In light of Mandelbrot and Fama’s insights, early risk practitioners developed financial risk modeling to compare the risk between investments, which involved comparing the respective standard deviations of the investments. The resulting method, Value-at-Risk, refers to a quantile of risk (typically 95% or 99% or higher). VaR is one of the primary methods used for commercial risk scoring today.

The problem here is that quantile methods like VaR don’t consider what happens in the worst 5% of scenarios (the tail). They only tell you where the risk begins, not what happens when you get there.

And unlike the natural/Gaussian world, where large deviations from the mean have a limited effect on the phenomena studied, that isn’t the case in the financial, heavy-tailed world, where rare events like the pandemic are extremely impactful.

Why correlation matrix limitations should also be considered

When it comes to using correlations for portfolios with multiple investments, this classic way to calculate risk is also inadequate, considering the technology we have available today, like high-performance computing, LLMs, and AI.

According to Dr. Piccinini, “The better approach is to do away completely with the correlation assumption and to directly recreate the theoretical returns of the portfolio using today’s weights. By simply adding together the weighted return vectors, no co-movement information is lost. A heavy tail ETL (expected tail loss) computed on the resulting time series contains all the interactions during periods of market volatility.”

Underestimating tail risk can compromise portfolio stability and client trust—it’s time for a new, improved risk score

Advisors have been saddled with backward-looking financial risk modeling for decades. But without accounting for tail risk—the potential for extreme losses in extreme market conditions—these legacy models put investors and their portfolios at risk. It’s a misguided way to conduct a portfolio risk assessment.

Advisors need one metric that builds trust and protects against the frequency and the magnitude of risk events (fat-tail risk). That’s why Dr. Ron Piccinini—one of the leading experts in “fat tail risk modeling”—has been so passionate about finding a better risk scoring solution to aid advisors in investment risk management.

Enter QuantumRisk™: A proprietary risk engine designed to protect portfolios from market shocks

Informed by Dr. Piccinini’s pioneering work in “fat tail risk” modeling, QuantumRisk™ calculates risk in real time, leaving standard deviation assumptions and outdated bell curves in the past. Now available within the Amplify Platform, QuantumRisk offers the high-level precision and complex calculations that advisors need to help clients make smarter decisions about their portfolios.

QuantumRisk™ was also made possible due to the advances in high-performance computing and graphics processing we have available today. Combined with Dr. Piccinini’s insight on “fat tail risk” modeling, QuantumRisk models real-world probabilities and market stress scenarios. It gives the advisor a clearer picture of portfolio risk by showing both the likelihood and severity of potential outcomes—going beyond routine market fluctuations.

What else should advisors and RIAs know about QuantumRisk?

Unlike traditional risk measurement tools, QuantumRisk was designed to crack the code regarding how to measure real-world portfolio risk. It’s more than an advanced tool. QuantumRisk makes it light years easier for advisors to measure, calculate, and manage risk. What else does this proprietary risk engine do?

QuantumRisk™:

• Simulates millions of real-world outcomes in less than a second.
• Uses an advanced 0–1,000 scale, where higher values equate to higher exposure to volatility and risk.
• Gives advisors a realistic, actionable view of portfolio exposure—beyond “average” volatility.
• Promotes smarter advisor-client conversations with a visual, client-ready experience.

Most important, it enables advisors to provide clients with the knowledge they need to make better informed financial decisions. In turn, clients can gain greater confidence that they’re on the right track.

Amplify Technology, LLC (“Amplify”) is not a registered investment adviser. Its services are for informational purposes only and do not constitute investment advice or recommendation. Please consult a registered investment adviser before using Amplify and its services.

QuantumRisk™ is a proprietary tool developed by Amplify Technology, LLC. It is for informational and educational purposes only, designed to support financial professionals in evaluating portfolio risk. It is not investment advice and does not make recommendations to buy or sell any security. Outcomes will vary depending on advisor use and client circumstances.