ICAAP Forum examines capital allocation and balance-sheet trade-offs

Written by Russell Pereira

Oct 12, 2026
7 min

The Banking Academy's inaugural ICAAP Forum examined how changing economic conditions affect bank capital, profitability and liquidity, with a simulation exploring the consequences of lending and funding decisions.

Banks' decisions on lending, funding and capital allocation are closely connected, but their consequences depend on economic conditions and management assumptions. Expanding lending may support income while increasing funding requirements and risk exposures. Attracting deposits may improve funding availability but raise costs, while retaining additional capital provides protection against losses at the expense of alternative uses.

These considerations formed the basis of The Banking Academy's inaugural Internal Capital Adequacy Assessment Process (ICAAP) Forum on 9 October 2026. The roundtable combined discussion of regulatory capital assessment, economic conditions and profitability with a banking simulation. Emmanuel Daniel, founder of TAB Global introduced the forum as a quarterly programme intended to "exchange notes and rehearse economic scenarios". Brian Lo, the forum's chairman and a former DBS treasury executive, examined ICAAP's role in management decisions. Banking analyst Nick Dove reviewed economic conditions and capital allocation, while Anoop Rai and Mark Flood demonstrated the ProBanker simulation.

Capital assessments extend beyond regulatory reporting
Lo examined ICAAP within the Basel supervisory framework. ICAAP is the internal process through which banks assess whether they hold sufficient capital against their risks, including those not fully captured by minimum regulatory requirements. It forms part of Pillar 2, the supervisory review process under the Basel framework.

Lo argued that ICAAP should not be treated solely as an annual reporting exercise. Lending, funding and liquidity decisions depend on assumptions that can change between formal assessments. Drawing on earlier financial crises and interest rate movements, he discussed how market conditions can alter both a bank's exposures and its options for responding. Longer-dated bonds, for example, may lose market value when interest rates rise, but the consequences also depend on whether the bank needs to sell and whether sufficient market liquidity is available. "To whom? Who's going to take from you?" Lo asked, questioning the assumption that securities can always be sold readily when market conditions deteriorate.

The connection between stress testing and management decisions is reflected in established supervisory expectations. The Basel Committee on Banking Supervision's stress-testing principles call for results to inform capital planning, liquidity assessments, risk appetite and wider business decisions. Lo's discussion concerned how banks apply these principles as conditions change, providing the context for Dove's subsequent analysis of profitability, credit growth and funding costs across markets.

Economic differences complicate capital allocation
Dove examined changes in interest rates, lending growth, profitability and funding costs across major markets. He highlighted "funding costs not necessarily following policy rates", particularly as government borrowing influences competition for longer-term financing. Banks therefore face different lending opportunities and funding constraints even under broadly similar economic conditions.

Comparing Indonesia, Thailand and the Philippines, Dove observed that Indonesian lending growth had remained relatively strong despite higher real interest rates, while Thailand's lower rates had not produced comparable credit expansion. He used the comparison to question how much lending activity can be explained by interest rates alone, given differences in borrower demand, economic conditions and banks' willingness to extend credit.

Dove also examined choices facing banks with capital above their immediate operating and regulatory needs. Additional lending may support earnings but increase exposure to credit losses, while retaining larger buffers limits resources available for growth or shareholder distributions. He warned that banks could "put their feet too hard on the loan growth accelerator" when capital positions are strong and recent credit losses have been limited.

Daniel noted that capital allocation decisions differ between institutions, while Dove acknowledged that banks with profitable growth opportunities may have stronger reasons to retain capital than those with fewer opportunities to deploy it. Lo also raised the influence of regulatory changes following previous crises. These considerations framed the balance-sheet decisions examined in Rai's subsequent simulation.

Lending and deposit decisions create competing balance-sheet effects
Rai introduced ProBanker, a training simulation in which a hypothetical bank operates under specified economic and regulatory assumptions. The exercise covers lending, deposits, investment securities, borrowing and capital, with outcomes reflected in financial statements and performance measures.

One exercise involved reducing loan interest rates by 0.3 percentage points to encourage borrowing while increasing deposit rates by one percentage point to attract funding. The changes were intended to stimulate lending and attract deposits, but lower lending rates and higher funding costs could also compress net interest margins. The resulting profitability and liquidity position depends on customer responses and the composition of the balance sheet.

Rai also introduced decisions involving securities, provisions for potential loan losses and wholesale borrowing. The simulation assessed their effects on profitability, liquidity and capital adequacy, including net interest margins and regulatory capital ratios.

A further feature was the treatment of discount-window advances. Rai explained that the model uses these when a bank ends a period with a residual funding shortfall, obtaining funds at a penalty rate. This is a simplified mechanism within the simulation, not an assumption that central bank funding would automatically be available to an actual institution. It illustrates how lending or investment decisions may generate additional income while creating funding pressures if the liabilities needed to support asset growth have not been adequately considered.

The exercise demonstrated how decisions intended to improve one financial measure can affect others. Its outcomes were generated under simplified assumptions rather than representing forecasts or assessments of actual banks.

Questions examine the limits of simulation
The demonstration raised questions about the model's treatment of liquidity, securities portfolios, capital requirements and funding shortfalls, including whether ProBanker could support more detailed analysis within individual institutions.

Rai explained that the simulation simplifies certain regulatory calculations to focus on relationships between banking activities. In the exercise discussed, for example, it did not include a separate capital requirement for market risk. The model therefore differs from the more comprehensive calculations required for a bank's ICAAP.

Flood, who previously worked at the US Federal Reserve, distinguished the educational exercise from institutional risk management. "Training is a very different exercise," he said, explaining that actual regulatory stress testing requires considerably greater detail and precision. The simulation allows mistakes and their consequences to be examined, whereas risk management in an operating bank is intended to prevent such outcomes.
Flood identified unexpected discount-window borrowing as one indication that asset growth may have been pursued without sufficient attention to funding requirements. A shortfall within the simulation provides an opportunity to examine the relationship between assets, liabilities and liquidity.

The exchanges clarified the simulation's scope. It offers a way to examine the relationship between profitability, capital and liquidity under specified conditions, but does not replace institution-specific stress testing, governance or internal capital assessment.

Recurring scenarios will extend the balance-sheet exercise
The quarterly programme will extend the introductory exercise through further economic scenarios and successive rounds of simulated bank management, as outlined by Daniel and Rai. Decisions made in one period will influence the financial position entering the next, allowing different lending, funding and capital-allocation choices to be examined over time. These results remain conditional on the model's assumptions rather than reflecting the circumstances of an individual institution.

As economic conditions change, the choices facing banks may shift even when their starting capital positions remain similar. The relationship between credit demand, funding costs and potential losses will influence whether capital is retained, deployed or distributed. The usefulness of recurring scenario assessment will depend on how banks interpret those relationships within their own regulatory requirements and operating constraints.