The 2008 global financial crisis exposed serious structural problems within the credit rating industry, particularly the conflicts of interest inherent in the “issuer-pay” model. In this system, rating agencies are paid by the very institutions they rate, incentivizing favorable ratings for complex financial products that should have raised red flags. As a result, risky financial instruments received AAA ratings, which misled investors and contributed to the economic collapse.
In response to this, several reforms were introduced. The Financial Stability Board (FSB) put forward principles aimed at reducing dependence on credit ratings and promoting better internal credit risk analysis by investors. Meanwhile, the Dodd-Frank Act in the U.S. imposed stricter regulations on credit rating agencies to ensure greater transparency and accountability.
Despite these efforts, critics argue that not enough has changed. The “Big Three” rating agencies—Moody’s, Standard & Poor’s, and Fitch—still dominate the market, and financial institutions continue to rely heavily on their ratings. This ongoing reliance on conflicted agencies raises questions about whether the lessons from the crisis have been fully learned.
One element that’s often oversimplified in media portrayals, such as in The Big Short, is how these high ratings were calculated. The film skims over the sophisticated statistical modeling, particularly David Li’s Gaussian Copula model. Introduced in 1999, it was revolutionary for its ability to assess risk by incorporating geographical diversification. Initially, it worked well by spreading risk across different regions, which made mortgage-backed securities appear safer. However, this reliance on geographic diversity overlooked the growing systemic risks tied to individual mortgage defaults, leading banks to lend excessively.
When the Federal Reserve raised interest rates, defaults soared nationwide, and the very models designed to mitigate risk unraveled. The complexity of these financial products and the blind trust in ratings based on these models contributed to the financial meltdown. While rating agencies have since adjusted their practices, including modifying geographical diversification models, the underlying challenge of accurately assessing risk remains.
David Li’s work, which significantly impacted financial modeling, can be explored in greater detail in his paper. For deeper insights into the methodologies used, you can read it here.
h/t RIP_Soulja_Slim