# Operational and Model Risk

> Operational risk comes from failed processes, people and systems; model risk from wrong or misused models. Learn real examples and the key controls.

Source: https://learn.tradelabsai.com/portfolio/operational-and-model-risk/  
Track: Portfolio and Performance · Level: Advanced · Updated: 2026-10-03  
Publisher: TradeLabs AI (https://tradelabsai.com). Education, not financial advice.  
Cite as: TradeLabs Learn, "Operational and Model Risk", https://learn.tradelabsai.com/portfolio/operational-and-model-risk/

Some of the largest trading losses in history had nothing to do with markets moving the wrong way in ordinary fashion. They came from a software deployment gone wrong, a rogue trader hiding losses, a fat finger order or a model that everyone trusted but nobody truly understood. Operational risk is the risk of loss from failed internal processes, people, systems or external events. Model risk is the risk of loss from models that are wrong, misapplied or misunderstood. Both are harder to measure than market risk, and both can be catastrophic.

## Operational risk categories

| Category | Examples |
|---|---|
| People | Errors, fraud, rogue trading, key person dependence |
| Processes | Missing controls, poor reconciliation, settlement failures |
| Systems | Software bugs, outages, cyber attacks, data errors |
| External events | Power failures, natural disasters, pandemics, vendor failures |

Banking regulation, through the Basel framework, requires banks to hold capital for operational risk.

## Famous operational failures

| Event | What went wrong | Lesson |
|---|---|---|
| Barings Bank, 1995 | Nick Leeson hid losses of about £827 million in a secret account; the bank collapsed | [The Fall of Barings Bank](https://learn.tradelabsai.com/history/the-fall-of-barings-bank/) |
| Société Générale, 2008 | Jérôme Kerviel built unauthorised positions; unwinding them cost about €4.9 billion | [Lessons From Market Failures](https://learn.tradelabsai.com/history/lessons-from-market-failures/) |
| Knight Capital, 2012 | A faulty deployment sent erroneous orders, losing about $440 million in 45 minutes | [Risk Controls and Kill Switches](https://learn.tradelabsai.com/algo-trading/risk-controls-and-kill-switches/) |
| Fat finger errors | Many cases of mistyped order sizes or prices causing large moves | [Alerts, Error Handling and Reconnection](https://learn.tradelabsai.com/algo-trading/error-handling/) |

The common thread is weak controls: separation of duties, reconciliation, limits and monitoring.

## Model risk

| Source | Example |
|---|---|
| Wrong assumptions | Assuming normal returns when tails are fat. See [Fat Tails](https://learn.tradelabsai.com/math/fat-tails/) |
| Estimation error | Unstable correlations or volatilities. See [Covariance and Correlation](https://learn.tradelabsai.com/math/covariance-and-correlation/) |
| Overfitting | A backtest that captures noise. See [Overfitting and Curve Fitting](https://learn.tradelabsai.com/research/overfitting-and-curve-fitting/) |
| Implementation errors | Bugs in pricing or risk code |
| Misuse | Applying a model outside the conditions it was built for |
| Stale models | Relationships that have changed. See [Structural Breaks and Regime Changes](https://learn.tradelabsai.com/math/regime-changes/) |

**Example: The London Whale**
In 2012, JPMorgan's Chief Investment Office built a huge position in credit derivatives. Losses eventually exceeded $6 billion. Investigations found that a new value at risk model introduced that year had a spreadsheet error, including dividing by a sum instead of an average, which roughly halved the reported VaR and made the position look less risky than it was. The model was approved with limited testing. The case is a standard example of model risk combining with weak oversight. See [Value at Risk (VaR)](https://learn.tradelabsai.com/portfolio/value-at-risk/).

## Controls for operational risk

1. **Separation of duties:** traders do not control settlement, accounting or risk reporting.
2. **Daily reconciliation** of positions and cash with independent sources. See [Trade Accounting and Reconciliation](https://learn.tradelabsai.com/industry/trade-reconciliation/).
3. **Pre trade limits and kill switches.** See [Risk Controls and Kill Switches](https://learn.tradelabsai.com/algo-trading/risk-controls-and-kill-switches/).
4. **Change management** for software, with testing and rollback.
5. **Audit trails** of all actions. See [Logging, Audit Trails and Incident Response](https://learn.tradelabsai.com/algo-trading/audit-trails/).
6. **Business continuity and disaster recovery plans.** See [Failover, Backups and Disaster Recovery](https://learn.tradelabsai.com/algo-trading/disaster-recovery/).
7. **Mandatory leave** for staff, which can expose hidden activity.

## Controls for model risk

US banking regulators' guidance known as SR 11 7 (2011) set out widely followed practices:

- **Independent validation** of models before use and periodically.
- **Documentation** of assumptions, limitations and intended use.
- **Ongoing monitoring** comparing model outputs with outcomes, such as VaR backtesting.
- **Model inventory** with owners and approval status.
- **Conservative use** where uncertainty is high, with overrides documented.

## For individual traders

Individuals face the same risks at small scale: an order typed with an extra zero, a bot with a bug, a spreadsheet error in position sizing, or trusting a backtest too much. Double checking order tickets, using platform limits, testing code and reviewing assumptions are cheap protections. See [Pre-Trade Checklist](https://learn.tradelabsai.com/start-here/pre-trade-checklist/).

## Frequently asked questions

### What is operational risk in trading?

The risk of losses from failed processes, human error or misconduct, system failures or external events, rather than from market moves.

### What is model risk?

The risk of losses from models that contain errors, rest on wrong assumptions or are used outside their intended purpose.

### How do firms reduce operational risk?

Through separation of duties, reconciliation, limits, kill switches, change management, audit trails and continuity planning.

Next, learn about risks to the whole financial system in [Systemic Risk](https://learn.tradelabsai.com/portfolio/systemic-risk/).

## Continue learning

- Next lesson: [Systemic Risk](https://learn.tradelabsai.com/portfolio/systemic-risk/)
- Previous lesson: [Liquidity Risk](https://learn.tradelabsai.com/portfolio/liquidity-risk/)
- Related: [Liquidity Risk](https://learn.tradelabsai.com/portfolio/liquidity-risk/): Liquidity risk is the danger of being unable to trade quickly at a fair price, or running short of cash. Learn its two types, how to measure it and controls.
- Related: [Risk Controls and Kill Switches](https://learn.tradelabsai.com/algo-trading/risk-controls-and-kill-switches/): Pre trade risk checks and kill switches stop a trading algorithm before a bug becomes a disaster. Learn the essential limits, how to layer them and how to test them.
- Related: [Logging, Audit Trails and Incident Response](https://learn.tradelabsai.com/algo-trading/audit-trails/): An audit trail records every signal, order, change and fill so trading can be reconstructed later. Learn what to log, the regulatory rules and how it helps traders.
- Related: [Failover, Backups and Disaster Recovery](https://learn.tradelabsai.com/algo-trading/disaster-recovery/): Power cuts, server crashes and broker outages happen. Learn how traders and trading systems plan for disasters, with backups, failover and tested recovery steps.
- Related: [Overfitting and Curve Fitting](https://learn.tradelabsai.com/research/overfitting-and-curve-fitting/): Overfitting means a strategy fits noise instead of a real pattern. Learn the warning signs, why it happens, how to measure it and practical ways to avoid it.
- Related: [The Fall of Barings Bank](https://learn.tradelabsai.com/history/the-fall-of-barings-bank/): In 1995 Nick Leeson's hidden losses of £827 million destroyed Barings, Britain's oldest merchant bank. Learn how it happened and the control failures behind it.