Capital adequacy and Basel ratios: how much loss can a bank absorb

# Capital adequacy and basel ratios: how much loss can a bank absorb

A bank holds $100 in mortgages and $100 in corporate loans. Under Basel III, the mortgages force it to set aside far less capital than the corporate loans, even though both are worth the same on paper. That single rule shapes what banks lend, to whom, and at what price. Let us see exactly how.

What "capital" actually means here

When regulators talk about capital, they do not mean cash in the vault. They mean the loss-absorbing cushion between a bank's assets and its obligations to depositors and bondholders.

The purest form is CET1 (Common Equity Tier 1): common shares plus retained earnings, minus some deductions. It is the first money to disappear when loans go bad. If a bank loses $5 on defaulted loans, that $5 comes out of CET1 before any depositor is touched.

More CET1 means a bank can eat more losses before failing. That is the whole game.

Risk-weighted assets: not all $100 is equal

A bank does not hold capital against its total assets. It holds capital against risk-weighted assets (RWA): each asset scaled by how risky regulators judge it to be.

The formula for a single asset:

RWA = Exposure × Risk Weight

Risk weights under the Basel standardised approach (the simpler of two methods) roughly follow credit risk:

• Cash and government bonds of strong sovereigns: 0%
• Residential mortgages: often 35% to 50%, depending on loan-to-value
• Most corporate loans: commonly 100%, lower for highly rated borrowers

So a $100 mortgage at a 50% risk weight becomes $50 of RWA. A $100 corporate loan at 100% becomes $100 of RWA. The corporate loan consumes twice the risk budget even though the balance sheet shows the same $100.

Computing the CET1 ratio

The headline metric regulators watch:

CET1 ratio = CET1 capital / Risk-weighted assets

Let us build a tiny bank.

Assets:

• Mortgage book: $100, risk weight 50% → RWA = $50
• Corporate loan book: $100, risk weight 100% → RWA = $100
• Total RWA = $150

Capital:

• CET1 = $12

CET1 ratio = 12 / 150 = 8.0%

That 8.0% means the bank holds $8 of pure equity for every $100 of risk-weighted exposure.

Now change the mix

Suppose the bank shifts $100 from corporate lending into mortgages. Same total assets ($200), same $12 of capital.

• Mortgage book: $200 → RWA = $100
• Corporate loan book: $0 → RWA = $0
• Total RWA = $100

CET1 ratio = 12 / 100 = 12.0%

Nothing about the bank's size changed. By tilting toward lower-risk-weight assets, the same capital now covers a much healthier ratio. This is why capital rules quietly steer banks toward mortgages and away from unrated corporate credit.

How much is enough? The Basel minimums

Basel III sets a stack of requirements, not one number. As a rough guide:

• Minimum CET1: 4.5% of RWA
• Capital conservation buffer: an extra 2.5%, bringing the practical CET1 floor to 7.0%
• Countercyclical buffer: 0% to 2.5%, switched on by national regulators in credit booms
• G-SIB surcharge: an additional buffer for global systemically important banks (the largest, most interconnected institutions)

Our first bank at 8.0% clears the 7.0% conservation-buffer level, but only just. A wave of corporate defaults could push it below, triggering restrictions on dividends and bonuses.

The Basel framework itself is published by the Bank for International Settlements. You can browse the consolidated standards at the BIS Basel Framework.

Standardised versus internal models

Large banks may use the IRB (Internal Ratings-Based) approach, estimating their own default probabilities to set risk weights, subject to regulator approval. Smaller banks use the standardised weights above.

The concern: two banks holding identical portfolios could report different RWA, because one uses aggressive internal models. To curb this, Basel III introduced an output floor. A bank's model-based RWA cannot fall below 72.5% of what the standardised approach would produce. Model gains are capped.

These final Basel III reforms (often called Basel III Endgame in the US and Basel 3.1 in the UK and EU) are phasing in through the mid-2020s and into 2026, with transitional arrangements varying by jurisdiction.

🎬 [VIDEO: "Basel III in 10 minutes" - youtube.com - a concise walkthrough of the capital, leverage, and liquidity pillars]

The leverage ratio: a backstop that ignores risk weights

Risk weights can be gamed. So Basel added a leverage ratio that ignores them entirely:

Leverage ratio = Tier 1 capital / Total exposure (unweighted)

The minimum is generally 3%, with a surcharge for the biggest banks. This catches a bank that has stuffed itself with "low risk weight" assets that turn out to be dangerous. The 2008 crisis featured banks with respectable risk-weighted ratios but wafer-thin unweighted equity. The leverage ratio is the crude, honest check on that trick.

Why this shapes real lending decisions

Capital is not free. Shareholders demand a return on it. So every loan carries an implicit capital cost.

Consider our two $100 loans again. The corporate loan requires capital against $100 of RWA; the mortgage against only $50. To hit the same return on equity, the bank must charge the corporate borrower a wider margin, or decline the loan.

This is why:

• Mortgages are priced tightly and offered widely.
• Unrated small-business loans carry higher spreads.
• Banks sometimes prefer holding government bonds (0% risk weight) over lending in weak economic periods, a pattern critics call "capital hoarding."

The risk weight is not an accounting footnote. It is a price signal baked into the rulebook.

Stress testing: the forward-looking layer

Ratios measure today. Regulators also ask: what about a severe recession?

Stress tests project a bank's capital under a hypothetical shock (say, unemployment spiking and house prices falling). The US Federal Reserve runs an annual exercise; the European Banking Authority and the Bank of England run their own. A bank might show a comfortable 12% CET1 ratio today but fall toward the minimum in a modelled crisis.

The key insight: a static ratio can look strong while hiding concentration risk. A bank heavy in one region's mortgages could see that "safe" 50% risk-weight book generate large losses at once. Stress tests expose what point-in-time RWA cannot.

You can read how the Federal Reserve frames its exercise on its stress tests page.

A worked comparison to lock it in

Two banks, each with $12 of CET1:

| | Bank A (corporate-heavy) | Bank B (mortgage-heavy) |

|---|---|---|

| Corporate loans | $150 (RW 100%) | $50 (RW 100%) |

| Mortgages | $50 (RW 50%) | $150 (RW 50%) |

| RWA | $150 + $25 = $175 | $50 + $75 = $125 |

| CET1 ratio | 12 / 175 = 6.9% | 12 / 125 = 9.6% |

Same capital. Same $200 balance sheet. Bank A sits below the 7.0% buffer threshold and would face dividend restrictions. Bank B has room to grow. The only difference is the risk profile of what they chose to lend.

Key Takeaways

• CET1 ratio = CET1 capital / risk-weighted assets. It measures how much loss a bank can absorb before depositors are at risk. The practical floor under Basel III is around 7.0% (4.5% minimum plus the 2.5% conservation buffer).
• Risk weights turn a flat balance sheet into a risk budget. A $100 mortgage at 50% consumes half the capital of a $100 corporate loan at 100%, which is why lending mixes tilt the way they do.
• The leverage ratio is the unweighted backstop. At roughly 3% minimum, it catches banks that look safe on a risk-weighted basis but hold thin real equity.
• The output floor (72.5%) caps how much banks can shave capital using internal models, a core piece of the Basel III reforms phasing in through 2026.
• Ratios are static; stress tests are dynamic. A strong ratio today can erode fast in a modelled downturn, especially where exposures are concentrated.