Risk Analytics The credit risk decision discipline.

From data to decision, with discipline at every step.

Risk decisions get made every day. Analytics should make them sharper, not slower.

Risk Analytics turns what the Lending Process learns about the borrower into numbers the institution actually uses. A rating. A probability of default. An expected loss. A risk-adjusted return. A ceiling the Board has set.

Four phases. Three calculations and the Risk Appetite that frames them. Auditable, calibratable, built to live inside daily lending. Not parked in a separate model room.

The shape of the discipline

Three calculations under one ceiling.

Every number starts with a rating. Runs through expected loss. Lands on RAROC. The Risk Appetite Statement sets the ceiling. Sequential in calculation. Linked in governance.

Foundation

Rating & Scoring

The borrower assessment layer. Combines heuristic judgment (questionnaires, sector knowledge) with statistical scoring (logistic regression on historical data) to produce a grade and, where data permits, a probability of default.

Calculation

Expected Loss & RAROC

EL combines PD with loss given default and exposure at default. RAROC translates that loss expectation into a return-on-capital decision metric used for pricing, hurdle rates, and capital allocation.

Framing

Risk Appetite Statement

The Board-approved ceiling that sets thresholds, tolerance zones, and escalation triggers. The RAS is what makes the calculations governance, not just analytics. Every figure ends with a comparison to the appetite.

How the stack works

Four phases, top-to-bottom.

Phase 4 frames it from above. Phases 1 to 3 do the math. Pricing, limits, allocation, and provisioning fall out of the bottom. Realized RAROC, NPL, and concentration feed back into the appetite. The next cycle starts sharper than the last.

Risk Analytics computational stack with Risk Appetite Statement framing PHASE 04 / FRAMING Risk Appetite Statement Sets thresholds, tolerance zones, escalation triggers CONSTRAINS THE STACK BELOW PHASE 01 / FOUNDATION Rating & Scoring Output Grade A-E PD % Heuristic + statistical Quant + qual + external data Calibrated, validated, documented FEEDS PHASE 02 / CALCULATION Expected Loss Formula EL = PD x LGD x EAD Calibrate to portfolio defaults Stage 1 / 2 / 3 transitions (IFRS 9) FEEDS PHASE 03 / DECISION RAROC Formula (Income - EL - Cost) / Capital Pricing, hurdle rates, allocation Compared to RAS thresholds DECISIONS PRODUCED Risk-based pricing Sector and obligor limits Capital allocation Provisioning and capital FEEDS BACK
Phase 01

Rating & Scoring

Grade A-E + PD %

Hybrid heuristic and statistical models. Quantitative, qualitative, and external data combine into a calibrated, validated grade.

Open Phase 01
Phase 02

Expected Loss

EL = PD x LGD x EAD

The loss expectation per facility and per portfolio. Calibrated to actual default and recovery experience. Feeds provisioning and pricing.

Open Phase 02
Phase 03

RAROC

(Income - EL - Cost) / Capital

Risk-adjusted return on capital. The single metric that makes pricing, hurdle rates, and capital allocation comparable across deals and segments.

Open Phase 03
Phase 04

Risk Appetite Statement

Board-approved limits

The framing constraint. Four pillars, three lines of defence, three tolerance zones. Where ambition meets capacity. Where analytics becomes governance.

Open Phase 04
Across Q-Lana

Related

Risk Analytics earns its keep when paired. Three places it shows up next.

The difference

Analytics that get used, not parked.

Most rating models live in a separate file. RAROC sits in a quarterly board pack. The Risk Appetite Statement gathers dust between reviews. Q-Lana puts the four phases on the same page the credit team opens every morning. The number on the screen is the number that was governed.

For CROs and Heads of Credit Risk

Walk us through the calculation you would re-build first.

No pitch deck. A working session on where your rating model, EL engine, RAROC, or RAS is strongest, where it leaks, and what would change with the right operating discipline.

Q-Lana Weekly

One newsletter. Five lenses. In your inbox every week.

Curated news with practitioner commentary. One Deep Dive in rotation. One applied tool. Read in fifteen minutes. Used the same week.

Q-Lana. The Operating System for SME and Corporate Finance.
Information becomes intelligence. Intelligence becomes advantage.