How Basel I Evolved with the 1995–1996 Amendments
💥 A Market Crash and a Basel Wake-Up Call
By the mid-90s, the financial world had been rocked by several market shocks — most famously the 1987 stock market crash. Banks were beginning to explore Value at Risk (VaR) and quantitative risk models. But Basel I, still stuck in 1988, was aging like a floppy disk in the era of Windows 95.
One of its biggest flaws?
Banks were penalized for doing the responsible thing — hedging.
Imagine you’re wearing both sunscreen and a raincoat. If the Basel I system were in charge, you’d still get sunburnt and soaked… twice, because it treated both exposures as separate and dangerous — ignoring the fact that they cancel each other out.
So what changed in 1995 that finally gave banks credit for protecting themselves?
🔁 1995 Amendment: The Rise of Netting
Enter the Netting Amendment. It allowed banks to offset their exposures when using legal agreements $($like the ISDA Master Agreement$)$.
Instead of counting each derivative position independently $($even if they were meant to hedge each other$)$, Basel now allowed: $CEA=max(∑Vi,0)+∑[0.4×Dj+0.6×Dj×NRR]$
Where:
- $V_i$: market value of each derivative
- D_j: add-on amount
- NRR: Net Replacement Ratio (exposure with netting / without netting)
🧠 Analogy:
Think of NRR as your gym buddy splitting the bill for a pizza after the workout — you still gain weight, but not as much.
📌 Example:
If your total positive exposure is \$25M and netting brings that to $18M,
then NRR = 18 / 25 = 0.72
Now you only pay add-ons on a reduced exposure — a big win for risk-aware banks.
But this still didn’t address a glaring issue: what about assets in the trading book?
📈 1996 Amendment: Market Risk, Meet Capital
Up to this point, Basel I largely ignored the elephant in the room — market risk. Trading book assets like bonds, derivatives, or commodities could lose value in seconds, and yet… there was no capital buffer for them.
The 1996 Market Risk Amendment changed that by:
- Requiring mark-to-market accounting for trading book assets
- Introducing two ways to calculate market risk capital:
- Standardized Measurement Method
- Internal Model-Based Approach (IM)
Standardized Method:
Like assembling IKEA furniture without the manual — rigid and no room for creativity. It assigned capital charges per item without considering diversification.
Internal Model Approach:
Like using Google Maps with traffic updates — smarter, faster, and optimized.
Banks could now use VaR models they developed internally, converting market risk into capital using:
$Market\ RWA=12.5×max(VaR_{t−1},m×VaR_{avg})$
But how do you make sure banks aren’t cheating with overly optimistic models?
🧪 Enter Backtesting: Basel’s Lie Detector
The amendment didn’t just trust banks blindly. It introduced backtesting.
Banks had to compare their daily 99
If losses exceeded VaR estimates too often, it meant:
- Your model is broken 🧯
- Or you’re just incredibly unlucky 🎲
Depending on the number of exceptions, the multiplicative factor (m) was adjusted:
| Exceptions | m Value |
|---|---|
| < 5 | 3 |
| 5 to 9 | 3.4 to 3.85 |
| > 10 | 4 (and higher scrutiny) |
👉 So the worse your predictions, the more capital you had to hold. Accountability meets math.
But what if even after all this, market risks overwhelm credit risks?
🧱 Tier 3 Capital: The Market Risk Buffer
To meet market risk capital needs without tapping into Tier 1 or 2, Basel introduced Tier 3 Capital:
- Subordinated unsecured debt
- Original maturity ≥ 2 years
🛠️ Think of it as the duct tape of bank capital — handy in a pinch but phased out by Basel III.
So what’s the final recipe for total capital under this upgrade?
🧮 Total Capital Requirement (Post-1996)
The full capital requirement now became:
$Total \ Capital=0.08×(Credit\ RWA+Market \ RWA)$
Where:
- Credit RWA = on-balance + off-balance (with netting, CEAs, etc.)
- Market RWA = calculated using VaR model (as per the formula above)
📌 Example:
If yesterday’s VaR = \$10M, 60-day average = $8M, m = 3:
$Market \ RWA=12.5×max(10,3×8)=12.5×24=300⇒Capital=0.08×300=\$24M$
But of course, this still wasn’t the end of Basel’s evolution…
🧠 Final Reflection: Basel Grows a Brain (and a Backbone)
The 1995 and 1996 Amendments marked Basel’s puberty — it:
- Started recognizing offsetting risks (netting)
- Accounted for trading book volatility
- Made banks prove their risk models were trustworthy
From here, Basel would mature into Basel II (with more risk sensitivity) and eventually Basel III (adding liquidity, leverage, and crisis-era stress).
But in many ways, the 1995–96 updates were the moment Basel learned how to measure reality — not just count paper.