Most GCC banks have spent heavily to get IFRS 9 right — and most of that spend went to the risk side of the house. The expected credit loss engine, the probability-of-default curves, the staging logic, the macroeconomic overlays: these sit in a dedicated risk platform, and rightly so. What far fewer banks have invested in with the same discipline is the finance side of the same standard — the layer where the credit loss number becomes a booked provision, a consolidated group result, a regulator-ready disclosure, a planned cost-of-risk line in next year’s budget, and a capital position that either supports the growth plan or quietly blocks it.
That finance-side performance layer is what Oracle EPM is for. And the gap between a bank’s ECL engine and its performance layer is where the recurring pain lives: the provision risk calculates and the provision finance books do not tie out cleanly; the group close across three or four regulators takes weeks; the profitability view by product and branch is a quarter old by the time it lands; and the capital plan, if it exists at all, was built in a spreadsheet that nobody wants to be responsible for during the SAMA examination.
This article is about where Oracle EPM belongs in a bank’s finance architecture in the Gulf — not as the system that computes credit losses, which it is not, but as the system that turns everything downstream of that calculation into consolidated, planned, disclosed, and defensible numbers. For a CFO or group financial controller in a Saudi, Emirati, or wider GCC bank, that distinction is the whole point.
Why banking is a different planning problem
A manufacturer plans revenue and cost and arrives at profit. A bank’s economics run through three variables that a general planning model rarely handles well, and all three are sharper in the GCC than almost anywhere.
The first is the cost of risk. Under IFRS 9, a bank recognises expected credit losses at all times, not only when a loss becomes evident, and it updates that expectation every reporting date against current conditions and a forward-looking view. The consequence for finance is that provisioning is no longer a year-end adjustment; it is a P&L line that moves with the economic cycle and with the migration of loans between stages. When a performing exposure in Stage 1, carrying a twelve-month loss, deteriorates enough to move to Stage 2, the provision switches to a lifetime basis — and on a multi-year facility that shift can multiply the provision several times over in a single reporting period. A bank that does not plan the cost of risk with staging sensitivity built in is a bank whose forecast will be overtaken by a single sector event.
The second is capital as a constraint, not an output. A bank cannot simply grow the balance sheet that looks most profitable. Every asset carries a risk weight, and the aggregate drives risk-weighted assets, which drive the capital the regulator requires it to hold. SAMA implemented the Basel III final reforms for Saudi banks from the start of 2023, covering credit, market, operational, and credit-valuation-adjustment risk and the leverage ratio, and it can require buffers above the standard level where it judges them appropriate. A growth plan that looks strong on the income statement can breach a capital buffer once its risk-weighted assets are loaded in — and if capital is not modelled forward alongside the plan, the bank discovers this after committing to the plan rather than before.
The third is the regulator as a primary audience. A bank’s numbers are not only management’s and the auditor’s; they are SAMA’s, the CBUAE’s, the QCB’s, and their equivalents across the Gulf, each with its own return formats, its own Pillar 3 disclosure framework, its own quarterly templates, and its own national discretions layered on top of the Basel text. Producing those disclosures from the same governed source that produces the management accounts is not a convenience. When finance and regulatory reporting diverge, it is the divergence itself that draws scrutiny.
None of these three is a technology feature. Each is a business reality that determines whether the performance system fits the institution — and each is a reason a general-purpose planning implementation, dropped into a bank without banking design, will disappoint.
Where the ECL engine ends and EPM begins
This is the distinction that separates a credible bank finance architecture from a confused one, so it is worth stating plainly.
The expected credit loss calculation belongs in a risk or ECL engine — Oracle’s own financial-services analytical applications and IFRS 9 solution, or a comparable dedicated platform. That is where probability of default, loss given default, and exposure at default combine across the three stages; where significant-increase-in-credit-risk triggers are evaluated, including the thirty-days-past-due backstop, rating-notch downgrades, watchlist placement, and sector stress; and where the forward-looking macroeconomic scenarios are applied. Oracle EPM does not compute this, and any proposal suggesting it should is a proposal to build a risk engine inside a planning tool, which is the wrong tool for the wrong job.
Oracle EPM begins where that calculation produces a result. The ECL output — allowance, provision, and the period-on-period movement that flows to the P&L — becomes an input to the finance system, and from that point the work is finance’s:
- Consolidation and disclosure in Financial Consolidation and Close (FCCS): rolling the provision and the underlying financial assets up across the legal-entity structure, eliminating intercompany, translating currencies, and producing the IFRS 9 disclosures and the statutory and regulatory financial statements from a single governed source.
- Planning the cost of risk and capital in the Planning platform (EPBCS): forecasting the provision line under scenarios, modelling stage-migration sensitivity, and planning risk-weighted assets and the resulting capital ratios forward so the growth plan is tested against the capital constraint before it is approved.
- Profitability and transfer pricing in Profitability and Cost Management (PCMCS): allocating cost and funding to products, branches, segments, and customers, and applying funds- or profit-transfer pricing so the bank can see where it actually earns its return, net of the cost of risk and the cost of capital.
- Reconciling finance and risk: holding the bridge between the ECL the risk engine produced and the provision finance booked, so that the two functions report the same number and can explain the difference where one legitimately exists.
Put simply: the risk engine answers how much credit loss do we expect. Oracle EPM answers what does that do to the group result, the disclosures, next year’s plan, our capital, and our profitability by product — and it answers those in one platform rather than five spreadsheets.
Comparison: the risk/ECL layer versus the finance/EPM layer
| Function | Risk / ECL engine | Finance / Oracle EPM layer |
|---|---|---|
| PD, LGD, EAD calculation | Yes — core purpose | No — consumes the result |
| Stage assignment and SICR triggers | Yes | No — receives staged output |
| Macroeconomic scenario overlays on ECL | Yes | Planning scenarios on the P&L impact |
| Provision as a booked, consolidated number | Feeds it | Yes — FCCS consolidation |
| IFRS 9 and regulatory disclosures | Provides data | Yes — produced from governed source |
| Cost-of-risk forecasting and budgeting | No | Yes — EPBCS |
| RWA and capital ratio planning | Supplies risk inputs | Yes — EPBCS capital planning |
| Product / branch / customer profitability | No | Yes — PCMCS |
| Funds / profit transfer pricing | No | Yes — PCMCS |
| Finance-to-risk reconciliation | One side | Holds the bridge |
What makes the GCC build different
A bank finance model that would work in London or Frankfurt still needs three things done differently to work in the Gulf. We treat all three as core scope, designed in from the first build rather than added once the standard model is live.
Islamic finance is not a variant — it is the model. The GCC holds roughly half of the world’s Islamic finance assets, and in most Gulf banks Shariah-compliant products are not a side book; they are a large share of the balance sheet, and in the full-fledged Islamic banks they are all of it. The finance model has to reflect that Murabaha and Ijara generate profit and rental income rather than interest, that Mudaraba and Musharaka are profit-and-loss-sharing arrangements that carry their own capital treatment and their own risk profile, that unrestricted and restricted investment accounts must be segregated with profit distributed to account holders on actual returns, and that sukuk are asset-backed instruments with their own accounting. SAMA maintains ECL and reporting frameworks specifically for full-fledged Islamic banks and for Shariah windows, and the AAOIFI standards intersect with IFRS in ways that shape both the disclosure and the profitability view. A consolidation and profitability model that treats Islamic products as conventional lending with the labels changed will misstate both profit and capital — and will not survive a Shariah audit.
One group, several regulators. A Gulf banking group with subsidiaries in more than one country reports into more than one central bank — SAMA, the CBUAE, the QCB, the CBK, the CBB, the CBO — each with its own returns, its own Pillar 3 framework, its own quarterly templates filed within tight windows, and its own national discretions on the Basel and IFRS 9 treatment. FCCS is where a single group consolidation can carry the local regulatory views alongside the group IFRS view, so the group close and the several local filings come from one reconciled source rather than from parallel spreadsheets that drift apart between quarters.
The regional foundations that every Gulf EPM model needs. Bilingual Arabic-English metadata on the dimensions, so the model reads correctly to the teams and the regulators who use it. Alignment to both the Hijri and Gregorian calendars where reporting and disclosure cycles require it. The correct tax treatment for a bank’s ownership structure, including the zakat-and-income-tax split that applies to mixed Saudi and foreign ownership. And, because banking data is among the most sensitive personal and financial data there is, in-region residency: for SAMA and NCA obligations the workload sits in an in-Kingdom OCI region — Jeddah or Riyadh — with residency enforced by where the data physically lives, not asserted in a contract. These are not enhancements bolted on at the end. In a bank, they are the conditions under which the model is allowed to exist.
The five ways these implementations fail in the Gulf
1. The ECL engine is treated as the finance system. A bank stands up a capable IFRS 9 risk platform and assumes the provisioning problem is solved. But the provision the risk engine calculates is never reconciled to the number finance books and consolidates, because no one owns the bridge between them. The gap surfaces at audit or examination, when finance reporting and risk reporting show different expected credit losses for the same book, and neither function can fully explain the difference. The fault is not the engine; it is the absence of the finance layer that was supposed to receive its output.
2. Multi-country group consolidation lives in spreadsheets. A group with banks under two or three regulators consolidates the group result by hand each quarter, reconciling different local provisioning treatments and intercompany positions in linked workbooks. The close runs long, the local regulatory filings are prepared separately from the group accounts, and the two views diverge in ways that take days to investigate. The group has a strong risk function and no consolidation platform, and it feels the difference every quarter-end.
3. Islamic products are modelled as conventional lending. A bank with a large Shariah book runs its profitability and capital on a model that treats Murabaha profit as interest, ignores the profit-and-loss-sharing character of its investment accounts, and applies conventional capital treatment across the board. The reported profitability by product is wrong, the capital position is misstated, and the profit distributed to investment account holders does not tie to the model. The problem is not visible until a Shariah audit or a sharp analyst asks why the numbers do not behave like an Islamic bank’s should.
4. The cost of risk is a flat assumption. The budget carries provisioning as a single percentage, unchanged across the plan. Then a sector shock — an oil-price move that stresses corporate exposures, a correction in a concentrated real-estate book — pushes a tranche of loans from Stage 1 to Stage 2, the provision on those facilities switches from a twelve-month to a lifetime basis, and the cost of risk jumps in a quarter the plan never contemplated. Because the staging sensitivity was never modelled, the P&L surprise arrives with no warning and lands straight on capital.
5. Capital is an output, not a constraint. The strategic plan is built on growth and margin, looks strong on the income statement, and is approved. Only later, when risk-weighted assets are loaded against it, does the plan breach the capital buffer SAMA expects the bank to hold. Had the plan modelled RWA and the capital ratios forward, the constraint would have shaped the plan; instead it invalidates it after the fact, and the bank re-plans under pressure.
Implementation scope and sequence
For a GCC bank building the finance-side performance layer around an existing ECL engine, the work has a predictable shape. It is sequenced deliberately: consolidation and disclosure first, because they are the regulatory floor, then planning and profitability on top of a stable consolidated foundation.
| Phase | Focus | Typical duration |
|---|---|---|
| Requirements to zero-level understanding | Map the risk-finance boundary, the regulators in scope, the Islamic product set, and the ECL feed | 3–4 weeks |
| Dimension and metadata design | Entity, regulator view, product and Shariah classification, bilingual metadata, calendar alignment | 3 weeks |
| FCCS build | Group consolidation, intercompany, currency, IFRS 9 provision input, regulatory views, disclosures | 6–8 weeks |
| Cost-of-risk and capital planning (EPBCS) | Provision forecasting with staging sensitivity, RWA and capital-ratio planning by scenario | 4–6 weeks |
| Profitability and transfer pricing (PCMCS) | Product, branch, and customer profitability; funds- or profit-transfer pricing | 4–6 weeks, where in scope |
| Integration | ECL engine, core banking, and GL feeds; reconciliation design | 3–4 weeks |
| Validation and parallel run | Reconcile ECL to booked provision and group result against a known period before go-live | 3–4 weeks |
Where the base consolidation environment is itself unstable, the honest sequence is to remediate that before building planning and profitability on top of it — a judgement worth making openly at the requirements stage rather than discovering during the parallel run.
The questions a bank CFO or CIO should ask
- Does the provision our risk engine calculates reconcile, on a defined bridge, to the provision finance books and consolidates?
- Do our group accounts and our several local regulatory filings come from one governed source, or from parallel spreadsheets?
- Does our plan model the cost of risk with stage-migration sensitivity, or as a flat assumption?
- Do we plan risk-weighted assets and capital ratios forward, so the growth plan is tested against the capital constraint before approval?
- Does our profitability view net the cost of risk and the cost of capital, or only direct cost?
- Are our Islamic products modelled on their own economics, or as conventional lending relabelled?
- Can we produce SAMA and other GCC regulatory disclosures from the same source as the management accounts?
- Where does the banking data physically reside, and does that satisfy SAMA and NCA residency expectations?
Related Insights
- Oracle EPM, BI and automation for banking
- Business Intelligence and Analytics for Banking and Financial Services in the GCC and Egypt: The Definitive 2026 Guide
- Oracle FCCS for Multi-Entity Organisations in the GCC and Egypt: The 2026 Implementation Guide
- Oracle EPM and Hyperion consulting for GCC and Egypt enterprises
Frequently Asked Questions
Q: Can Oracle EPM calculate IFRS 9 expected credit loss? No, and it should not. The ECL calculation — probability of default, loss given default, exposure at default, staging, and the significant-increase-in-credit-risk logic — belongs in a dedicated risk or ECL engine. Oracle EPM consumes that result and does the finance work downstream of it: consolidating the provision, producing the disclosures, planning the cost of risk and capital, and analysing profitability. Proposing that a planning tool compute credit losses is proposing the wrong tool for the job.
Q: Where does Oracle EPM fit versus a risk platform like OFSAA? They sit either side of a clean boundary. The risk platform answers how much credit loss the bank expects and why. Oracle EPM answers what that does to the consolidated result, the regulatory and statutory disclosures, next year’s plan, the capital position, and profitability by product and segment. The reconciliation between the two — the ECL calculated versus the provision booked — is itself a finance responsibility that lives in the EPM layer.
Q: Can it consolidate a banking group across several GCC regulators? Yes — this is one of the strongest reasons a Gulf group needs FCCS. A single group consolidation can carry the local regulatory views for SAMA, the CBUAE, the QCB, and their peers alongside the group IFRS view, so the group close and the local filings come from one reconciled source. The alternative — parallel spreadsheets per regulator — is exactly what drifts apart between quarters and draws examiner attention.
Q: Does Oracle EPM support Islamic banking? It has to, and in the GCC that is a design requirement rather than an option. The model must reflect profit and rental income rather than interest on Murabaha and Ijara, the profit-and-loss-sharing character and capital treatment of Mudaraba and Musharaka, the segregation and profit distribution of investment accounts, and the accounting for sukuk — consistent with SAMA’s frameworks for Islamic banks and Shariah windows and with the AAOIFI standards where they intersect with IFRS. A model that treats Islamic products as conventional lending will misstate profit and capital.
Q: How do you plan the cost of risk and capital? Cost of risk is forecast in the planning platform with stage-migration sensitivity built in, so the plan shows what happens to provisions when exposures move from a twelve-month to a lifetime basis under a stress scenario. Capital is planned by modelling risk-weighted assets forward and deriving the resulting ratios, so the growth plan is tested against the buffer the regulator expects before it is approved rather than after.
Q: Is IFRS 9 provisioning a finance or a risk responsibility? Both, at a boundary. Risk calculates the expected loss; finance books it, consolidates it, discloses it, and plans it. The strongest arrangement treats the risk engine as the source of the ECL and Oracle EPM as the place where that number becomes a consolidated, disclosed, planned result — with an explicit reconciliation between the two so the functions report the same figure and can explain any legitimate difference.
About Loop Wise Solutions
Loop Wise Solutions is an enterprise performance consultancy based in Cairo, serving organisations across the Arab world. We specialise in Oracle EPM, Business Intelligence, Intelligent Automation, and Business and Technical Consultancy — helping finance and operations leaders plan better, report faster, and make decisions with confidence.
Our consultants have delivered consolidation, close, planning, and profitability models inside financial-services enterprises across Saudi Arabia, the UAE, and the wider Gulf, in environments where SAMA and its regional peers are a primary audience and where Islamic finance is core to the balance sheet. We treat the regional foundations — multi-regulator consolidation, Islamic product economics, bilingual and Hijri reporting, and in-Kingdom data residency — as core scope in every engagement, not a later phase. The people who design the model are the people who build it, and we stay in partnership through the first close and planning cycles rather than leaving at go-live.
If your bank has invested in the ECL engine but not in the finance layer that turns its output into consolidated, planned, and disclosed numbers, that is the gap worth closing before the next examination or the next plan. We would be glad to look at where your architecture stands today.
Contact: contact@loop-wise.com | Website: www.loop-wise.com
Where performance meets precision.
