Model Risk
The risk of loss arising from errors or incorrect assumptions in financial models used to value assets or manage risk.
What is Model Risk?
Model risk is the potential for adverse outcomes resulting from using flawed, misapplied, or inappropriately calibrated quantitative models to price securities, measure risk, or make investment decisions. Financial models are simplifications of reality that depend on assumptions — about return distributions, correlations, volatility, liquidity, and causation — that may be wrong or may break down under stress. Model risk has three main sources: conceptual errors (wrong theoretical framework), implementation errors (bugs or data errors in the model), and inappropriate use (applying a model outside its validated domain or market conditions). The 2008 financial crisis illustrated extreme model risk: credit rating models and CDO pricing models systematically mispriced mortgage risk using assumptions of normal distributions and stable correlations that failed catastrophically. Regulators including the Federal Reserve's SR 11-7 guidance require banks to formally validate, stress test, and document quantitative models.
Example
A trading desk uses a Value at Risk model calibrated on 2003–2006 data — a period of unusually low volatility and high asset correlations with risk. The model estimates a daily 99% VaR of $50 million. During the 2008 crisis, actual daily losses exceeded $200 million on multiple occasions — 4× the model's estimate — because the model had not accounted for fat-tail distributions or correlation spikes under stress.
Source: Federal Reserve SR 11-7 — Guidance on Model Risk Management