Skip to content
All library documents

Basel Capital Adequacy Ratios as a Loss Cushion

Article Quant Q&A · Author: Shyam
Read the original

Summary

The document explains capital adequacy as a buffer intended to absorb loan losses, not a guarantee that a bank cannot fail. It characterizes Basel's 8% minimum as a historical benchmark and notes that the ratio depends on subjective risk weights. As a result, regulatory measures can classify exposures differently from market assessments of their risk.

An illustrative bank balance sheet shows how loan defaults and recoveries reduce equity, while existing provisions absorb some losses. Further provisioning against remaining loans also reduces the cushion, even if the bank remains solvent under the regulatory measure. The example emphasizes that banks may curtail lending before reaching insolvency as their capital buffer shrinks. It is an explanatory illustration, not a derivation of the Basel threshold or a complete account of capital rules, risk-weight calculations, or bank failure mechanisms.

Key ideas

  • Capital adequacy is a cushion against losses rather than a promise that a bank cannot fail.
  • The ratio depends on risk-weighted assets, and those weights may not match market views of risk.
  • Loan losses and additional provisions can erode equity and reduce a bank's capital buffer.
  • Banks may restrict lending as the cushion shrinks, before insolvency becomes immediate.
  • The example does not derive the Basel minimum or cover the full regulatory framework.

Tags

Fixed incomeRisk management
Full text
# Capital Adequacy Ratio Basel 3


# Capital Adequacy Ratio Basel 3












Wanted to understand how Basel arrived at the figure of 8% for capital adequacy .

Is it fair to say that if banks maintain CAR, then even if their loan portfolio goes bad, they have enough capital to provide for it? And does that also mean that banks cannot go bust if they maintain their CAR?

Thanks

## Answer by demully (score 1)

https://quant.stackexchange.com/a/50875

I honestly wouldn't read too much into the 8% figure. It's just a historical ballpark figure for minimum adequacy to weather all but the biggest shocks (in which case, it would be sensible and appropriate for the central banks to step in and play their lender-of-last resort role).

Most banks after their post-GFC workouts run much higher than this these days. Running 8% equity:CoreTier1 ratios these days would be seen by regulators and the market as "flashing some ankle", if not a full-on Janet Jackson "nipplegate". The effective minimum is well above 8% these days!

Plus the 8% is based on equally subjective risk-weightings for different kinds of loans. Back in the EUrozone Crisis, a bank owning PIIGS (even French at times) govvie bonds yielding 10% was Basel "safer" than a bank lending to German students at 4%. The market CDS of those Banks was obviously not consistent with the regulatory measures ;-)

What the CAR tries to guage is how much loans can go bad before the bank runs out of equity (rather than they cannot ever go bust thus). It's more of a "margin of safety" thing. How much equity do they have to cover potential losses?

Note that this does NOT mean that lending is immune to losses that the banks can cover without failing. Say I have 100 loans, 10 equity with 1 provisions booked for bad/doubtful debts (BDD) = 10% CAR. 2 loans go bad, and I recover 25% = 1.5 losses. Less the 1 provision for these already booked, that's a 0.5 hit to profits and equity. So I now have 98 loans, 9.5 equity, 0 BDD = 9.7% CAR. But I now have no provisions for BDDs on the rest of my book. Provisioning for the same 1% of assets, that's a 0.98 hit to profits and equity. So I end up 98 loans, 8.52 equity with 0.98 provisions = 8.7% CAR.

I'm not bust, or even "underwater" from a regulatory perspective - but my capital cushion has been wiped out if I expected 1% of loans to sour and 2% did. So I will curtail my lending (even start calling in riskier debts) long before my solvency ever becomes a pertinent question. This is the real (and still unresolved) issue with banks!

hope this helps

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.