Skip to content
FinLexicon Wiki
Directory

Liquidity Coverage Nsfr Prudential

Non-Maturity Deposits (NMD) Behavioral Modeling: Deposit Betas, Core Caps & SRP31 Duration Ceilings

Contractually withdrawable at par overnight yet behaviorally sticky across rate cycles, Non-Maturity Deposits (NMDs) are governed under Basel SRP31 by strict core proportion ceilings (50%–90%) and average maturity caps of 4 to 5 years.

By Arthur Pendleton, CPA, JDReview not recorded · Updated
Non-Maturity Deposits (NMD) Behavioral Modeling: Deposit Betas, Core Caps & SRP31 Duration Ceilings — FinLexicon Wiki 2026 quantitative & supervisory data card
FinLexicon Wiki (Basel LCR, NSFR, HQLA Haircuts & Intraday Liquidity Buffers): Convex deposit betas, replicating tractor swaps, and SRP31 5.0Y/4.5Y/4.0Y maturity ceilings.Source: FinLexicon Wiki Institutional Research DeskLicense: CC-BY-4.0

Checking accounts, NOW accounts, corporate cash-management balances, and instant-access savings accounts—collectively termed Non-Maturity Deposits (NMDs)—carry a contractual repricing maturity of zero days (overnight). Yet because millions of retail payroll and operational accounts exhibit low rate sensitivity ("deposit beta"), treasury ALCO desks hedge NMDs with multi-year receiver swaps or fixed-rate bonds. To prevent banks from overstating deposit duration and hiding IRRBB EVE duration mismatches, Basel Framework paragraphs SRP31.73–SRP31.82 impose hard statutory caps on both core NMD proportions and behavioral average maturities.

Decomposing NMDs Into Overnight Non-Core and Replicating-Portfolio Core Slices#

Every NMD category must first be segmented into homogeneous depositor pools—Retail Transactional, Retail Non-Transactional (savings/MMDA), and Wholesale Non-Financial Corporate (financial-institution sight deposits can NEVER be treated as core NMDs and must be slotted 100% overnight!). Within each eligible pool, as connected to our Basel LCR, NSFR & HQLA cornerstone guide, ALCO models separate balances into:

  • Non-Core NMDs (Rate-Sensitive / Volatile): Slotted into the overnight (<= 1 day) repricing time bucket. This includes both rate-chasing balances expected to reprice immediately (Deposit Beta) and balances expected to migrate into high-yielding time deposits or money-market funds when policy rates surge.
  • Core NMDs (Stable & Rate-Insensitive): Stable balances unlikely to reprice even under significant interest-rate shocks, which treasury hedges via a replicating portfolio of staggered 1-to-10-year fixed-rate tranches (provided the weighted-average maturity stays within SRP31 caps!).
Basel SRP31 Non-Maturity Deposit segmentation, beta pass-through, and core maturity cap workflow
Basel SRP31 Non-Maturity Deposit segmentation, beta pass-through, and core maturity cap workflowVerified Exhibit

Basel SRP31 Table 2 Statutory Ceilings: 90%/70%/50% Core & 5.0Y/4.5Y/4.0Y Maturity Caps#

Under the Basel SRP31 standardized framework (Table 2, SRP31.81)—and mirrored as supervisory backstops by the European Central Bank, EBA (5-year retail/non-financial average cap), and UK PRA—banks cannot exceed strict two-dimensional caps:

  1. Retail Transactional NMDs: Maximum 90% Core proportion; Maximum 5.0 years average maturity of core (=> Max 4.50 years portfolio-wide weighted average maturity!).
  2. Retail Non-Transactional NMDs (Savings): Maximum 70% Core proportion; Maximum 4.5 years average maturity of core (=> Max 3.15 years portfolio-wide WAM!).
  3. Wholesale Non-Financial Corporate NMDs: Maximum 50% Core proportion; Maximum 4.0 years average maturity of core (=> Max 2.00 years portfolio-wide WAM!).
  4. Wholesale Financial Institution NMDs: 0% Core (100% Overnight, 0.0 years WAM!).

Non-Linear Deposit Betas and Convex Regime Shifts After Zero-Rate Decades#

Why did econometric NMD models calibrated on 2009–2021 data fail across regional and universal banks in 2022–2023? Because deposit pass-through betas (beta = Delta_r_deposit / Delta_r_policy) are convex and state-dependent!

  • For the first +100 bps hike off zero, bank deposit betas often sit near 0.15–0.25. Once central bank policy rates cross 3.50%–5.00% and mobile banking apps advertise 5.0% Treasury money-market funds in two taps, cumulative betas jump toward 0.55–0.75 while non-interest-bearing checking accounts cannibalize into high-yield CDs! Institutions like Bank of America, BNP Paribas, and Ally Financial stress-test regime-switching S-curve betas rather than linear regressions.
Quantitative Distribution · SRP31 Max Core Share
SRP31 Max Core ShareSRP31 Max Core Avg Maturity
01Retail Transactional (Check…90.0%02Retail Non-Transactional (S…70.0%03Wholesale Non-Financial Cor…50.0%04Wholesale Financial Institu…0.0% (Forbidden…
NMD Depositor SegmentSRP31 Max Core ShareSRP31 Max Core Avg MaturityMax Total Pool WAMCorresponding LCR / NSFR Weight
Retail Transactional (Checking/Payroll)90.0%5.0 Years4.50 Years3%–5% LCR Outflow / 95% ASF
Retail Non-Transactional (Savings/MMDA)70.0%4.5 Years3.15 Years10%+ LCR Outflow / 90% ASF
Wholesale Non-Financial Corporate50.0%4.0 Years2.00 Years25% Op / 40% Non-Op / 50% ASF
Wholesale Financial Institutions / NBFI0.0% (Forbidden!)0.0 Years (Overnight)0.00 Years25% Op / 100% Non-Op / 0% ASF
EU EBA RTS 2022/10 Aggregate RuleModel-validated5.0 Years Hard Ceiling<= 5.00 YearsExcludes financial & high-value deposits

Reconciling LCR 30-Day Run-Off, NSFR 95%/90% ASF, and IRRBB NMD Duration#

Supervisors cross-audit a bank’s behavioral assumptions across liquidity and interest-rate silos: claiming a high-yield online savings account is "70% core with 4.5-year duration" for IRRBB while simultaneously classifying it as a higher-outflow digital/less-stable deposit in 30-day LCR run-off tables triggers immediate SREP/horizontal exam challenges!

Comparison of modeled vs SRP31-capped effective NMD duration and resulting Delta-EVE impact
Comparison of modeled vs SRP31-capped effective NMD duration and resulting Delta-EVE impactVerified Exhibit

Structural Tractor Replicating Portfolios in Funds Transfer Pricing (FTP)#

To stabilize net interest margins, treasury desks credit retail branches via a rolling 60-month "tractor" swap strip (e.g., 1/60th rolling into a new 5-year receiver swap each month) calibrated inside Contingent Funding & FTP governance across the Basel LCR, NSFR & Prudential Liquidity topic hub.

Formula or model from the articleNot independently verified
Core_Share_Eff = min( Cap_Core_SRP31 , (1 - Decay_Stress) * (1 - Beta_Shock) )
  • WAMCoreEff = min( CapMatSRP31 , ModeledCoreWAM )
  • WAMTotalPool = CoreShareEff * WAMCoreEff + (1 - CoreShareEff) * (1 / 365)
  • Worked Retail Savings NMD Calibration ($40.0B Digital + Branch Savings Pool under +200 bps Up-Shock):
  • 10-Year Historical Decay in Stress = 15% | Up-Shock Deposit Beta (including migration to Term CDs) = 0.40
  • Modeled Core_Share = (1 - 0.15) * (1 - 0.40) = 0.85 * 0.60 = 51.0% (Below 70.0% SRP31 Ceiling -> 51.0% binds)
  • Modeled Core Ladder = Equal 1Y-to-8Y Tractor (Avg = 4.50Y, exactly at 4.50Y SRP31 Retail Non-Transactional Cap)
  • Effective Pool WAM = 0.51 * 4.50Y = 2.295 Years ($20.4B Core Hedged Across 1–8Y; $19.6B Slotted Overnight)

Frequently asked questions

What exactly is a "60-month tractor" replicating portfolio for Non-Maturity Deposits?

A 60-month (5-year) tractor invests 1/60th of the Core NMD balance each month into a new 5-year fixed-rate asset or receive-fixed swap. At steady state, maturities are evenly spaced from 1 month to 60 months, giving a smooth 2.5-year average duration that dampens NII volatility.

Why does deposit cannibalization (migration from 0% checking into 4.5% CDs at the same bank) hurt IRRBB even if total cash stays at the bank?

Even though liquidity (LCR/NSFR) appears unchanged when a customer shifts $100,000 from 0% checking into a 6-month 4.8% CD at the same bank, economically a 4.5-year zero-cost liability vanished overnight and repriced to market rates—destroying both NII and the EVE duration hedge!

Are banks allowed to apply the 90%/5.0Y SRP31 caps mechanically without behavioral modeling?

Only institutions using the simplified Standardized IRRBB Framework apply Table 2 baselines subject to historical caps; internal-model IMS banks must prove empirically via multi-cycle econometric decay and beta models that their core share and maturity are justified below those ceilings.

How does EBA RTS 2022/10 treat high-net-worth retail deposits above EUR 100,000 in NMD modeling?

Mirroring LCR40 retail rules, EBA IRRBB guidelines require banks to segment high-value/private-banking balances separately from mass-market payroll checking because affluent clients exhibit wholesale-like deposit betas (0.60–0.85) during tightening cycles.

Publication record

Published
14 November 2026
Updated
15 November 2026
Author
Arthur Pendleton, CPA, JD
Review
Review not recorded
Automated fact check
Passed in the stored pipeline record
Source URLs
3 recorded

AI assistance recorded: anthropic/claude-opus-4.6. This field does not establish human review or source verification.

Academic & Institutional Attribution

Cite This Research & Embed Live Data Card

Open Attribution License (CC BY 4.0)

Standard academic & institutional research paper citation format.

Arthur Pendleton, CPA, JD. (2026, November 14). Non-Maturity Deposits (NMD) Behavioral Modeling: Deposit Betas, Core Caps & SRP31 Duration Ceilings. FinLexicon Wiki. https://finlexicon.org/wiki/non-maturity-deposits-nmd-core-pass-through-behavioral-caps
Explore Liquidity Coverage Nsfr Prudential Hub (20) →

FinLexicon Wiki Data, Benchmarks & Contributor Records

Non-Maturity Deposits (NMD) Betas & SRP31 Duration Caps | FinLexicon Wiki