Imagine a party where everyone enjoys the music, but no one wants to pay the DJ. That was the state of liquidity pricing before the global financial crisis. Banks were dancing to the rhythm of cheap funding, ignoring the true cost of liquidity. Then came 2008 — the music stopped, and everyone scrambled to pay the bill. Welcome to the world of Liquidity Transfer Pricing $($LTP$)$.
🧾 Zero-Cost Approach: The “Free Buffet” Fallacy
The zero-cost approach treated liquidity like free snacks at a party — unlimited and always available. Under this model:
- Assets were not charged for using liquidity.
- Liabilities were not credited for providing liquidity.
- The swap curve was used as the reference, but no liquidity premium was added.
What went wrong? When the crisis hit, the swap spread ballooned — from 0.5 bps to over 120 bps — revealing that liquidity wasn’t free. Banks had funded long-term, illiquid assets with overnight borrowing, creating a ticking time bomb. Like relying on vending machines in a blackout.
This leads us to ask: If treating liquidity as free backfired, can averaging costs solve the problem?
⚖️ Pooled Average Cost of Funds: One-Size-Fits-All
This approach averaged funding costs across maturities. Whether it was a one-year loan or a five-year deposit, everyone got the same liquidity cost or credit.
Example:
If a bank reduces the spread from 9 bps to 6 bps:
- Loans look cheaper \$→\$ More assets.
- Deposits get less credit \$→\$ Fewer liabilities.
Outcome: The loan-to-deposit ratio becomes dangerously skewed.
But the question remains: Is this fair to long-term liabilities or risky assets?
🔄 Separate Average Costs of Funds: Slightly Better, Still Blindfolded
Here, the bank uses different average spreads for:
- Assets \$→\$ e.g., 9 bps
- Liabilities \$→\$ e.g., 3 bps
This model gives more flexibility. For example, banks could incentivize loan generation by lowering the asset charge without hurting deposit incentives.
Panel A and B Illustration:
- A 5-year loan \$→\$ Charged $90$
- A 5-year deposit \$→\$ Credited $30$
Still, this model doesn’t consider maturity-based pricing. All loans or deposits, regardless of term, are priced the same. Like charging a one-hour Uber ride the same as a five-hour road trip.
So… what’s the smarter way?
🧠 Marginal Cost of Funds: Reality Check Mode
The matched-maturity marginal cost approach is the gold standard.
It prices liquidity using:
- The current market cost $($e.g., swap curve + liquidity premium$)$
- Maturity-specific spreads
- Embedded derivative stripping for fixed-to-floating conversions
Example:
Suppose Bank MMM has:
- 1-year term spread: 4 bps \$→\$ Cost: \$40 on \$100,000
- 5-year term spread: 35 bps \$→\$ Cost: \$350 on \$100,000
Bank AVG using average cost would charge \$70 for both.
Key takeaway: The longer the term, the more mispricing if average cost is used.
🧮 Calculating Costs – Amortizing vs Bullet Loans
Amortizing Loan:
Loan of \$500,000 paid in 5 yearly \$100,000 tranches.
Using marginal cost rates:
- Year 1: 4 bps
- Year 5: 35 bps
Use weighted average or duration-style method: $\text{Charge} = \frac{1(4) + 2(8) + 3(15) + 4(24) + 5(35)}{1+2+3+4+5} = \frac{336}{15} = 22.4 \text{ bps}$
🔁 Weighted-Average Life $($WAL$)$ for Portfolios
$WAL = \sum_{i=1}^{n} \left( \frac{p_i}{P} \times t_i \right)$
Where:
- $p_i$ = principal in year $i$
- $P$ = total loan amount
- $t_i$ = year of repayment
Example:
If a \$1 billion mortgage pool behaves like it matures in 8 years, and current spread is 58 bps:
$\text{Charge} = 0.0058 \times 1,000,000,000 = \$5.8 \text{ million}$
💰 Credit for Liquidity – Sticky vs Hot Deposits
- Sticky deposits: e.g., term deposits $→$ Less likely to be withdrawn $→$ Higher credit
- Hot deposits: e.g., demand/savings accounts $→$ Volatile $→$ Lower credit
Example:
Using same spreads as before:
- Bank MMM 5-year term deposit $→$ Credit = $350
- Bank AVG 5-year term deposit $→$ Credit = $70
🧨 Contingent Liquidity Risk Pricing: Cushion or Catastrophe?
Regulators now require:
- Liquidity Coverage Ratio $($LCR$)$ for 30-day shocks
- Net Stable Funding Ratio $($NSFR$)$ for long-term coverage
Liquidity Cushion Needs:
- Reduced retail deposits
- Derivative collateral calls
- Line of credit drawdowns
These must be funded using liquid assets — not just overnight cash.
🎯 Example: Pricing a Line of Credit
Line = $20 million
Outstanding = $5 million
Probability of drawdown = 70
$\text{Dollar Cost} = 7.88 \text{ bps} \times 20,000,000 = \\$15,750$
🏁 Final Thoughts: Pricing Liquidity Like a Pro
Whether it’s funding a 30-year mortgage or pricing a revolving line of credit, the key is precision. Zero-cost and average cost approaches belong to the past. The matched-maturity marginal cost method, aligned with current market rates and funding realities, is the future.
Banks that fail to price liquidity properly aren’t just missing pennies — they’re playing financial Jenga with dynamite blocks.
Let’s keep the party going — but now, everyone pays the DJ. 🎧💸