Loading field map
A bank's balance sheet is a promise stretched across time. Deposits can be withdrawn tomorrow, but mortgages run for decades. A corporate loan may default, while the certificate of deposit that funded it must be repaid at par. Coordinating these mismatched maturities, interest-rate sensitivities, and liquidity risks is the task of asset liability management (ALM). Over the past century, ALM has moved from a passive assumption that assets alone determine safety to an integrated discipline that simulates the entire balance sheet under uncertainty. Each major framework emerged because its predecessor could not handle a new pressure—rising rate volatility, competitive funding markets, regulatory capital constraints, or the demand for enterprise-wide risk measurement.
For the first half of the twentieth century, bankers treated liabilities as given. Deposits were stable, interest rates moved slowly, and regulation capped the rates banks could pay. Under the Asset-Management paradigm, safety meant holding the right mix of assets: short-term self-liquidating commercial loans, government securities, and cash reserves. The liability side was a passive pool of deposits that bankers assumed would stay put. This framework prescribed that a bank's only active decision was how to allocate its asset portfolio. Liquidity risk was managed by keeping a buffer of highly liquid assets—a practice that later regulators would codify into liquidity coverage requirements. The paradigm worked well in a low-volatility environment, but it could not address what happened when depositors began shopping for higher yields or when interest rates started to swing.
By the 1960s, deposit markets had become competitive. Large depositors moved funds to the highest bidder, and banks discovered that liabilities were not fixed. The Liability Management framework turned the Asset-Management paradigm on its head: banks could actively manage their funding side by issuing certificates of deposit, borrowing in the federal funds market, and using Eurodollar deposits to raise funds when loan demand exceeded deposit growth. This was not a refinement of the older view but a replacement of its core assumption. ALM now meant balancing two sides of the balance sheet, not just one. The framework gave banks flexibility to grow without waiting for deposits to arrive, but it also introduced new vulnerabilities. If a bank relied on purchased funds that could disappear overnight, a funding crisis could unfold quickly. The savings-and-loan crisis of the 1980s would expose exactly this weakness.
The interest-rate volatility of the late 1970s and early 1980s made the Liability Management framework's informal approach to rate risk untenable. Banks needed a systematic way to measure how much their net interest income would change when rates moved. Gap Analysis provided that tool. It grouped assets and liabilities into maturity buckets—for example, 0–3 months, 3–12 months, 1–5 years—and calculated the difference, or gap, between rate-sensitive assets and rate-sensitive liabilities in each bucket. A positive gap meant more assets repricing than liabilities in that period, so net interest income would rise if rates increased; a negative gap meant the opposite. The framework was simple enough for any bank to implement, and it gave ALM officers a concrete number to manage. But Gap Analysis had a serious limitation: it treated all items within a bucket as if they repriced at the same moment, ignoring differences in timing within the bucket. It also measured only income sensitivity, not the change in the economic value of equity. Banks that relied solely on Gap Analysis could be misled about their true exposure.
Duration Gap Analysis addressed the blind spots of Gap Analysis by shifting the focus from income to market value. Duration measures the weighted-average time to receive a security's cash flows, expressed in years; it also approximates the percentage change in price for a given change in interest rates. By calculating the duration of assets and the duration of liabilities, a bank could compute a duration gap: the difference between the two, adjusted for leverage. A positive duration gap meant that the market value of equity would fall if rates rose; a negative gap meant equity would rise. This framework absorbed the maturity-bucket logic of Gap Analysis but replaced its static income focus with a present-value sensitivity that linked ALM directly to shareholder value. Duration Gap Analysis coexisted with Gap Analysis through the 1980s and 1990s—many banks used both, one for short-term income planning and the other for long-term value protection. Yet duration itself had limits: it assumed parallel shifts in the yield curve and did not handle options such as prepayments or early withdrawals. Those limitations pushed the field toward more flexible simulation tools.
In 1988, the Basel Committee on Banking Supervision introduced a capital adequacy framework that fundamentally changed ALM. For the first time, external regulation prescribed how much capital a bank must hold against its assets, based on risk-weighted categories. The Basel Accords framework did not replace internal ALM models; it overlaid them with binding constraints. A bank could measure its interest-rate gap however it liked, but it also had to meet minimum capital ratios calculated from risk-weighted assets. Basel II (2004) added a three-pillar structure—minimum capital, supervisory review, and market discipline—and allowed sophisticated banks to use internal models for credit risk. Basel III (2010–2017) introduced liquidity standards that directly affected ALM: the Liquidity Coverage Ratio (LCR) required banks to hold enough high-quality liquid assets to survive a 30-day stress scenario, and the Net Stable Funding Ratio (NSFR) required stable funding for long-term assets. These metrics revived the Asset-Management paradigm's emphasis on liquid asset buffers, but now as a regulatory floor rather than a voluntary practice. The Basel framework remains active today, and its standardized approach to capital and liquidity coexists in tension with banks' own internal ALM models.
By the 1990s, ALM had accumulated a collection of separate tools—Gap Analysis for income, duration for equity value, stress tests for liquidity, and regulatory ratios for capital. Dynamic Financial Analysis (DFA) integrated these into a single stochastic simulation framework. DFA models project the entire balance sheet forward under thousands of random interest-rate, credit, and economic scenarios, producing distributions of earnings, capital, and liquidity. It absorbed the earlier gap and duration tools as inputs rather than standalone methods. A DFA model might simulate how a rising-rate environment affects net interest income (the Gap Analysis question), the market value of equity (the duration question), and the LCR (the Basel question) all at once. The framework is computationally intensive and requires significant modeling expertise, so it is used mainly by large banks and insurance companies. DFA remains active today, and its relationship with the Basel framework is one of living disagreement. Basel's standardized metrics are designed for comparability across banks and simplicity of supervision; DFA's bank-specific simulations are designed for internal risk management and strategic planning. Regulators encourage DFA for internal purposes but require Basel metrics for reporting, creating a dual system where banks must reconcile two different views of their own risk.
Today, the Basel Accords framework and Dynamic Financial Analysis are the two leading frameworks in ALM, and they agree on several fundamentals: both recognize that ALM must cover interest rate, liquidity, and credit risk simultaneously; both use scenario analysis and stress testing; and both tie capital adequacy to risk measurement. Their disagreements run deeper. Basel's standardized approach treats all banks of a given type similarly, while DFA insists that each bank's risk profile is unique. Basel sets fixed thresholds (e.g., LCR ≥ 100%), while DFA produces probability distributions that management can interpret flexibly. Basel is backward-looking in its reliance on historical risk weights, while DFA can incorporate forward-looking scenarios. The field is moving toward greater integration—regulators now require internal liquidity stress tests alongside the LCR, and DFA models increasingly incorporate regulatory constraints as boundary conditions. The central question for the next generation of ALM is whether standardized regulation or internal simulation will drive the most important decisions, or whether a hybrid approach can preserve the strengths of both.
A student looking at the timeline should see not a simple march of progress but a series of responses to specific pressures: stable deposits gave way to competitive funding, simple income gaps gave way to market-value sensitivity, internal models gave way to regulatory overlays, and fragmented tools gave way to integrated simulation. Each framework remains visible in current practice—the Asset-Management paradigm's liquid asset buffers live on in the LCR, Gap Analysis survives in earnings-at-risk reports, duration is embedded in DFA models, and Basel metrics constrain every decision. ALM today is a discipline of managing these layers simultaneously, knowing that no single framework captures the whole picture.