Loading field map
Commercial banking has always revolved around a fundamental tension: how to attract funding, extend loans, manage risk, and remain profitable in a competitive environment. Over the past century, banks have adopted different frameworks—each with its own assumptions about the proper way to balance these objectives. This overview traces six frameworks that have shaped commercial banking, focusing on their distinctive commitments and the relationships between them.
For the first half of the twentieth century, commercial banks operated under the asset-management framework. Banks treated deposits as largely given—passive and stable—so they focused on the asset side of the balance sheet: selecting safe, liquid investments. The dominant doctrine was that banks should lend short term, self-liquidating loans (the real bills idea) and hold government bonds for liquidity. Regulatory constraints and a conservative culture reinforced this passivity. Liquidity risk was managed entirely through asset choice: hold enough liquid assets to cover expected deposit outflows. This framework assumed that a bank's success depended on its portfolio of assets, not its ability to attract new funds.
In the 1960s, the liability-management revolution overturned the passive-assumption of asset-management. Banks discovered that they could actively manage their liabilities by issuing negotiable certificates of deposit (CDs) and borrowing in the money market. Instead of waiting for deposits to arrive, banks could now purchase funds to meet loan demand. This shift reduced reliance on liquid asset buffers and allowed banks to grow faster. Yet both frameworks address liquidity risk: asset-management does so from the asset side, liability management from the liability side. They coexist today as complementary funding strategies: most banks still hold some liquid assets, but active liability management has become standard infrastructure.
The 1980s brought new pressures: corporate borrowers began bypassing banks via commercial paper markets, and interest rate deregulation squeezed margins. In response, transaction banking emerged as a fee-based, scale-oriented model. It treats banking as a series of standardized, automated transactions—cash management, trade finance, payments—with minimal relationship-specific customization. The framework assumes that competition is primarily on price and operational efficiency, and that information about customers is easy to codify. This is a sharp contrast with the earlier funding-centric models; transaction banking earns fees, not interest spread, and often services customers without lending.
At nearly the same time, relationship banking developed as a deliberate alternative. Rather than automating transactions, relationship banks invest in collecting soft, proprietary information about borrowers through repeated interactions and personalized service. They offer customized loans and advice, often to opaque small and medium enterprises that cannot access securities markets. The framework assumes that information asymmetry is severe and that long-term ties reduce monitoring costs. Relationship banking coexists with transaction banking: the former thrives where information is hard to standardize, the latter where scale and efficiency matter. Their tension persists in debates about the value of relationship lending versus arm's-length transactions.
The risk-management framework crystallized in the late 1980s with the Basel I capital accord, but it is far more than a regulatory response. It introduced a comprehensive language for measuring and managing risk: value-at-risk (VaR), risk-adjusted return on capital (RAROC), and credit scoring models. Banks began to treat risk as a quantifiable commodity that could be priced, hedged, and allocated across business lines. This framework integrates the asset and liability sides of the balance sheet under a single analytical lens. It absorbs earlier asset-liability management techniques but extends them into a bank-wide risk infrastructure. Today, risk-management is the dominant logic in commercial banking: regulators, rating agencies, and investors all demand risk-based quantification. Yet it does not replace the other frameworks; it provides the coordinating language in which they operate. For example, loan pricing in relationship banking now incorporates risk models, and originate-to-distribute pipelines depend on risk ratings.
Starting in the 1990s, securitization enabled banks to originate loans and then sell them to investors rather than holding them to maturity. This originate-to-distribute (OTD) model fundamentally altered the incentive structure: a bank's profit shifted from collecting interest over the loan life to earning origination and servicing fees. In breaking the hold-to-maturity assumption of relationship banking, OTD also weakened monitoring incentives because the bank no longer bore the credit risk. The framework relies on a liability-management funding infrastructure (short-term borrowing to fund securitization pipelines) and on risk-management quantification (credit scores, loan ratings) to certify loan quality. The 2007–2009 crisis exposed these weaknesses, leading to regulatory reforms that forced banks to retain some risk (risk retention rules). OTD survives but in a more constrained form.
All six frameworks remain active in contemporary commercial banking, each occupying a specific domain. Liability management funds the balance sheet. Transaction banking serves large corporations needing efficient payments. Relationship banking serves small businesses and mid-market firms. Risk-management provides the measurement infrastructure that regulators and internal committees rely on. Originate-to-distribute channels mortgage and consumer loans. The leading frameworks today—risk-management, transaction banking, and relationship banking—agree that banks must hold capital against risk and that funding sources must be diversified. They disagree vigorously about the extent to which risk can be fully quantified: transaction banking and risk-management favor models and data, while relationship banking insists that soft information cannot be captured in scores. This productive tension continues to drive innovation in commercial banking, preventing any single framework from becoming dogma.