🧠 It’s Not Just About Having a Risk Model — It’s About Using It Right
Congratulations! You’ve got a shiny economic capital model that spits out numbers, quantifies tail risk, and impresses your board. But here’s the catch:
“Is anyone actually using this model to make decisions?”
Enter: the Bank for International Settlements (BIS) — the global whisperer of banking best practices. The BIS has a list of 10 commandments (well, recommendations) that supervisors and banks should consider if they want to make their internal models actually useful.
So let’s walk through them, and along the way, ask:
“How do these great ideas work in practice — and what could possibly go wrong?”
📜 BIS Recommendations: The “How to Not Screw Up Internal Models” Starter Pack
1. Capital Adequacy Must Meet Decision-Making
If your economic capital model is just for show, it’s as useful as a gym membership you never use.
☑️ The bank should demonstrate how the model informs real decisions — like which risks to accept.
But who actually makes those decisions?
2. Senior Management Must Be Awake and Involved
The boardroom isn’t just for approving mergers and long lunches. Economic capital models need buy-in from the top.
Because if the CEO thinks VaR is an acronym for a new crypto coin, we’re in trouble.
So if senior management’s on board… how do we ensure they’re working with good info?
3. Transparency is King
Risk numbers shouldn’t be black-box magic. If your risk model speaks Latin and only the quant team understands it, you’ve already lost.
☑️ Keep outputs traceable and flexible enough for stress testing.
But before we start testing things, how do we know we’re even measuring the right risks?
4. Start with the Right Risk Identification
If you misidentify the enemy, no amount of modeling will save you. A robust capital model depends on pinpointing the correct drivers, exposures, and positions.
Great — but once we’ve found the risks, how do we measure them?
5. Risk Measures Must Be Understood, Not Just Impressive
No risk measure is perfect. VaR is too shallow, Expected Shortfall is hard to explain, and spectral measures sound like ghosts.
☑️ Use multiple measures and know their limitations.
Okay, but once we have multiple risks measured — how do we combine them without breaking logic?
6. Risk Aggregation Isn’t Math Soup
Simply adding risks together like Legos doesn’t cut it. You need:
- Consistent parameters
- Logic that reflects business reality
- Aggregation methods that match your bank’s DNA
Sounds complicated. What if our aggregation goes wrong?
7. Validate Like Your Capital Depends on It (Because It Does)
Validation isn’t a checkbox. You must prove your model works — not just once, but repeatedly, under stress, and with real data.
But how do you validate assumptions about how risks interact — especially credit risk?
8. Mind the Credit Risk Dependencies
Modeling credit risk is like modeling group behavior at a wedding.
“Will they all behave? Or cascade into chaos together?”
Copulas, ASRF, and portfolio models help — but every approach has assumptions that can unravel under stress.
Speaking of stress… what about the people we’re trading with?
9. Counterparty Credit Risk Is a Beast
When your counterparty fails, you don’t just lose money — you lose faith in the system.
Measuring CCR requires simulation, aggregation, stress testing, and a deep breath.
And what about the slow-burning risk sitting in the banking book?
10. Interest Rate Risk in the Banking Book (IRRBB) Is No Snoozefest
IRRBB deals with rate-sensitive assets like mortgages and deposits. Embedded options + rate changes = serious modeling headaches.
And here comes the dilemma:
- Earnings-based models = easier, but don’t play nice with other risks
- Economic value models = better for integration, but often ignored by business units
So these BIS recommendations are great… but how do they play out inside the bank?
🧩 Economic Capital in Real Life: What’s Working, What’s Not?
Now that we’ve got the BIS blueprint, the next step is understanding how economic capital plays out in practice — and what constraints and opportunities it creates.
Let’s break it down.
📘 Credit Portfolio Management:
“Should I Say Yes to This Loan — Or Diversify Instead?”
Constraint: You can’t just evaluate loans individually anymore. You have to consider incremental portfolio risk. One good loan can still mess up your portfolio if it creates a sector overload.
Opportunity: With good portfolio models, you can optimize hedges, reduce concentration, and create a beautifully diversified lending book.
Once you’ve set up the portfolio… how do you price the individual deals?
💸 Risk-Based Pricing:
“Is This Deal Worth It — Or Am I Paying to Lose?”
Constraint: RAROC-based pricing means deals below a threshold get rejected. No more sweetheart loans unless they generate a risk-adjusted return and make the shareholders smile.
Opportunity: Precision pricing = higher profitability. Plus, you can still override the model — with management approval — if the relationship is worth it (like that customer who brings in ten other accounts).
And speaking of customers, can we tell who’s actually profitable?
👥 Customer Profitability Analysis:
“Are We Making Money Off This Client — Or Just Keeping Them Warm?”
Constraint: Aggregating all the risks per customer is harder than herding cats. Plus, segmenting customers by return/risk requires pristine data.
Opportunity: Once you figure it out, you’ll spot freeloaders, and reallocate capital toward MVP clients — maximizing the risk-reward equation.
But will employees actually follow this logic if their bonus isn’t tied to it?
🏆 Management Incentives:
“Do Managers Actually Care About Economic Capital?”
Constraint: In many banks, compensation has nothing to do with economic capital. Sad trombone.
Opportunity: When you tie incentives to capital efficiency, managers pay attention. Suddenly, they care about portfolio risk, pricing logic, and the capital impact of that “just one more deal.”
Cool. But who’s watching the watchers? What ensures this whole system stays legit?
🧭 Best Practices & Governance:
“Is Anyone Making Sure This Economic Capital Framework Isn’t Just… Theater?”
Governance is the glue that holds everything together. Without it, models are misused, capital is misallocated, and risk becomes a guessing game.
Best practices include:
- Senior management commitment – no lip service, real engagement
- Clear roles across business units – centralized or decentralized, just be clear
- Defined policies & owners – someone owns the model, updates it, explains it
- Timely measurement & disclosure – monthly or quarterly, but consistent
🚨 Key Governance Concerns:
- Is senior management really involved, or just signing off stuff they don’t understand?
- Are stress tests integrated — or just box-ticking exercises?
- Are risks being measured correctly — absolute vs. relative, diversified or not?
- Is economic capital treated as the only measure — or one of many tools?
- Do we have a clear definition of available capital?
- Is the model transparent enough to be used — and trusted — for real decisions?
🎯 Final Word: Economic Capital Is Not Just a Model — It’s a Culture Shift
When done right, economic capital becomes more than a calculation — it becomes a language of risk, spoken from the front line to the boardroom.
It aligns:
- Risk with strategy
- Capital with opportunity
- Decisions with long-term value
But to get there, banks need to follow best practices, confront limitations, and most importantly:
Turn the model from a number factory into a strategic compass.
Leave a Reply