Valuation Methodology Is Not Optional

Something gets overlooked in alternative asset lending discussions.

The moment you use a model to value collateral, you trigger regulatory requirements for model risk management. OCC Bulletin 2011-12 applies. For credit unions, NCUA guidance parallels these requirements.

Skip this step and you're building a regulatory liability instead of a sustainable capability.

What Counts as a Model

What constitutes a "model" under regulatory guidance?

OCC 2011-12 defines a model as a quantitative method that applies statistical, economic, financial, or mathematical theories, techniques, and assumptions to process input data into quantitative estimates.

For alternative asset valuation, that includes:

• Algorithms that estimate startup equity value based on comparable transactions
• Models that calculate cryptocurrency haircuts based on volatility analysis
• Methods that estimate private fund NAV between reporting periods
• Any systematic approach that transforms observable data into collateral value

If you're doing anything more sophisticated than looking up a price on Bloomberg, you're probably using a model.

Why This Matters

Model risk matters because models can be wrong.

Consider a model that values startup equity based on recent financing rounds. The model might correctly identify a $10 million valuation from a Series B closed six months ago. But if the company's runway shortened, key employees departed, or market conditions changed, that $10 million valuation might be dramatically wrong.

The model produced a number. The number might not reflect reality.

Now imagine you lent against that valuation with a 40% advance rate. You have $4 million in loans secured by equity that might be worth $3 million. Or $2 million. Or zero.

Model risk becomes credit risk becomes capital risk.

The institutions that collapsed in the crypto winter didn't fail because their technology was weak. They failed because their models produced values that diverged from reality and their risk management didn't catch it.

OCC 2011-12 Requirements

OCC 2011-12 establishes specific requirements. These aren't suggestions.

Documented methodology. Every model must have documented methodology describing its theoretical basis, assumptions, limitations, and expected operating conditions. You need to explain why the model should produce accurate estimates and under what circumstances it might not.

For alternative assets, this means documenting why your cryptocurrency haircut model uses particular volatility windows, why your startup valuation model weights particular comparables, why your private fund valuation assumes particular discount rates.

Development discipline. Models should be developed following sound practices. This includes rigorous testing against historical data, sensitivity analysis to understand how inputs affect outputs, and clear understanding of the conditions under which the model performs well or poorly.

Independent validation. Models must be validated by parties independent from the development process. This doesn't mean different employees; it means independent review that can challenge assumptions and verify performance.

For many community institutions, independent validation may require external expertise. That's acceptable. What's not acceptable is validating your own work.

Ongoing monitoring. Models must be monitored on an ongoing basis to assess performance. When market conditions change, models may behave differently. Monitoring catches this before it becomes a problem.

Board oversight. Senior management and board of directors must understand model risk and provide appropriate oversight. This requires understanding of what models are used, what risks they present, and how those risks are managed.

Practical Implementation

Regulatory requirements translate into practical implementation for alternative asset lending.

Methodology documentation. Create written documentation for each valuation approach. For cryptocurrency, document the data sources (exchanges, aggregators), the haircut calculation methodology, the volatility analysis approach, and the conditions under which the methodology might produce unreliable results.

For startup equity, document the comparable transaction selection criteria, the adjustment methodology for differences between companies, the treatment of preference stacks and liquidation terms, and the limitations of the approach.

This documentation serves multiple purposes: regulatory compliance, institutional knowledge preservation, and the discipline that comes from having to explain your thinking clearly.

Validation program. Establish a validation program that reviews models periodically. For alternative asset valuation models, annual validation is typically appropriate, with more frequent review if market conditions change significantly.

Validation should assess:

• Whether model methodology remains theoretically sound
• Whether model performance matches expectations (backtesting)
• Whether input data sources remain reliable
• Whether model documentation remains current
• Whether model limitations are understood and communicated

Performance monitoring. Track actual outcomes against model predictions. When you value startup equity at $5 million and later see an exit at $3 million, that's data. Systematically collecting and analyzing this data improves model calibration over time.

For cryptocurrency, compare predicted liquidation values to actual liquidation values when loans are closed. For private funds, compare model estimates to subsequent NAV statements. Build institutional knowledge from actual experience.

Governance structure. Establish clear responsibilities for model risk management. Who owns the models? Who validates them? Who monitors them? Who approves changes? Who reports to the board?

The structure matters less than having one. Many institutions assign model oversight to an existing risk committee with appropriate expertise supplementation.

The Compliance Advantage

Here's the counterintuitive point: model risk management creates competitive advantage, not just regulatory compliance.

Institutions with robust model governance understand their models deeply. They know when models work well and when they don't. They can adjust lending terms based on model reliability. They can explain their methodology to examiners, members, and partners.

Institutions without model governance are flying blind. They use models they don't fully understand. They can't explain their methodology under scrutiny. They can't distinguish between model limitations and model failures.

When markets get volatile, the first group adjusts intelligently. The second group either overreacts (missing opportunity) or underreacts (taking losses).

The discipline imposed by OCC 2011-12 is the discipline that keeps you from blowing up.

NCUA Parallels

For credit unions, NCUA guidance parallels OCC requirements without identical language.

NCUA Letter to Credit Unions 16-CU-08 addresses third-party due diligence and oversight, which applies when using external valuation providers. NCUA's supervisory expectations for complex products include appropriate risk management and board oversight.

The principles are consistent: understand your models, validate them independently, monitor their performance, and maintain appropriate governance.

Credit union examiners increasingly ask about model risk management when they see non-traditional lending activities. Being prepared matters.

Common Mistakes

Common mistakes I see institutions make:

Treating vendor models as exempt. If you use a third-party valuation service, you're still responsible for understanding and governing that model. OCC 2011-12 explicitly addresses third-party models. Using a vendor doesn't eliminate your oversight responsibility.

Validating development, not performance. Some institutions validate whether a model was built correctly but don't validate whether it actually works. Performance validation, comparing model outputs to actual outcomes, is essential.

Documenting once and forgetting. Methodology documentation isn't a one-time exercise. Models evolve. Markets change. Documentation must stay current.

Treating model risk as compliance overhead. Model risk management is risk management. It protects your institution from systematic errors. Treating it as bureaucracy misses the point: this is protection.

The Path Forward

For institutions entering alternative asset lending, model risk management should be addressed at the outset, not added later.

Before launch: Document methodology. Establish validation procedures. Define governance structure. Brief the board on model risk.

At launch: Monitor performance closely. Validate assumptions against early results. Refine methodology based on experience.

Ongoing: Maintain regular validation cycles. Update documentation as methodology evolves. Keep the board informed of model performance and any concerns.

This approach treats model risk management as integral to the product, not as compliance overhead layered on top.

What This Means for Your Program

Valuation methodology is not optional. The moment you use models to value alternative asset collateral, you've accepted responsibility for managing model risk.

That responsibility includes documentation, validation, monitoring, and governance. It requires investment and discipline. It does not disappear because you find it inconvenient.

The discipline required by model risk management is the discipline that produces sustainable alternative asset lending. Institutions that skip this step tend to learn the hard way.

The crypto-lending graveyard is full of companies that valued assets without methodology. Don't join them.

OCC Bulletin 2011-12 establishes model risk management requirements for national banks. NCUA supervisory expectations establish parallel requirements for credit unions. Both apply when models are used to value alternative asset collateral.