Selected Work

Detailed descriptions of current accelerators and prior implementation work. The distinction matters: reusable capsules are described as ready-to-implement; delivery experience is described as work undertaken through prior consulting and SAS roles.

Ready-to-implement SAS accelerator

IRRBB Stress Capsule

A SAS Stored Process framework for internal IRRBB stress testing. It starts with a bank's existing bucketed cashflows and yield-curve inputs, applies stress scenarios and behavioural adjustments, then produces EVE, NII, reverse-stress and management-report outputs in a controlled run.

What the capsule does

Creates four stress sets—RBI, Baseline, Medium and Severe—each across seven scenarios.
Calculates EVE across the full 19-bucket IRRBB maturity structure using stressed discount curves.
Calculates NII over a one-year horizon from the first six buckets using the GAP method.
Applies scenario-specific TDRR and CPR adjustments before EVE/NII are calculated.
Maintains a run identifier, version, timestamps and append-only result history.
Produces a management dashboard and a nine-sheet Excel report pack.

Reverse stress: rather than asking only what happens under a fixed rate shock, a binary search identifies the minimum parallel-up or parallel-down shock that moves the EVE result beyond the 15% Tier-1 outlier limit. This gives management a limit-headroom measure.

Open the illustrative capsule walkthrough →

Behavioural-model workbench

Behavioural Models & Bank-Workbook Validation

The framework is deliberately modular: each product model is built, reconciled and validated separately before being integrated into the IRRBB cashflow engine. This avoids a single large implementation becoming impossible to validate.

Retail NMD

Daily balance history is converted into bucket-level percentage drops using fixed business-day look-backs.
The 95th-percentile worst drop is calculated per bucket and de-cumulated to derive the non-stable portion.
Stable balances are adjusted for deposit-rate pass-through (PTR/Beta); the remaining core is capped according to the regulatory treatment and distributed over long buckets.
Current-account treatment differs from savings: no paid interest means PTR is zero; the model therefore focuses on stability rather than repricing.

Term Deposit Redemption Rate (TDRR)

Monthly early redemption is measured against the prior-month outstanding balance for each of 19 maturity buckets.
One-sided mean + 2 standard-deviation outlier treatment replaces exceptional spikes rather than deleting observations.
Short, medium and long tenor groups are forecast separately with log-linear macroeconomic regressions.
Group forecasts are apportioned back to individual buckets, compounded to a lifetime redemption rate and translated into adjusted IRRBB cashflows.

Validation pattern: each calculation is reconciled to the bank's workbook—formula cells, not just column labels. Once the standard model tallies, SARIMAX challengers and walk-forward backtests can assess autocorrelation and out-of-sample performance.

Open the NMD walkthrough → · Read the NMD note → · Read the TDRR note →

Quantitative risk / validation

AIRO & Embedded Interest-Rate Optionality

AIRO work focuses on the part of IRRBB that ordinary contractual cashflows miss: instruments where the customer has an option to change the cashflow pattern if rates move. Examples include caps, floors, collars, callable/puttable instruments and early withdrawal of wholesale term deposits.

Validate the curve used for forward rates and discounting, option position, strike, reset frequency and exercise convention.
Decompose caps and floors into caplets/floorlets where appropriate, aggregate values by product and stress scenario, and retain drill-down outputs.
Calculate option-adjusted EVE by adding the scenario option value to standard EVE.
Apply Black-76 or a Hull-White interest-rate model according to the product, market-data availability and level of optionality required.

Read the AIRO valuation method → · Full IRRBB methodology →

Career delivery experience

Risk-System Delivery, Market Risk & ALM

Prior roles at SAS, Ryzklytix, EY and Wipro covered the production delivery side: turning risk methodology into configured and automated SAS solutions, data flows, risk calculations, reports and controls.

IRRBB, ALM and market-risk solution configuration, requirements and stakeholder walkthroughs.
Pricing, valuation, historical simulation and Monte Carlo VaR configurations using SAS risk platforms.
Cashflow schedule generation, maturity/duration analysis, liquidity and interest-rate-risk calculations.
Source data extraction, transformation, SAS OLAP/web reporting and regulatory reporting outputs.
SAS 9.4 and Viya migration, configuration and functional validation support.

Public descriptions use generic, client-safe language. No client data, production source code or confidential implementation parameters are published.