•   Centralized, transparent funds transfer pricing (FTP) gives ALM direct influence over pricing, planning, and business performance.
•   Sustainable profitability requires relationship-level economics that combine funding, credit, capital, and customer value.
•   Effective balance sheet analytics must translate into timely decisions, not just reports, by embedding business context and ownership.
•   AI should enhance accessibility and insight generation while validated risk engines remain the system of record.
•   Best-in-class organizations operate from a single governed data foundation with shared assumptions across risk, finance, and business teams.
•   Competitive advantage comes from ranking opportunities consistently across funding, risk, and capital to optimize scarce resources and improve capital allocation.

What are the defining characteristics of organisations that have successfully transformed ALM from a reporting function into a strategic partner, and how can others replicate this shift?

The banks we've seen make this shift didn't get there on better analytics. What changed was what ALM was allowed to own, and the clearest marker is funds transfer pricing. Where FTP is owned centrally, run transparently, and set at the product and segment level, ALM has a seat in pricing and planning, because it owns the number every line of business is measured against. Where it's a black box maintained elsewhere, ALM stays close to the limits and far from the decisions, whatever the models can do. The other thing that changes is the lens. The questions the CFO is asking — which segments to grow, where margin holds without buying volume, where the next unit of capital does the most work — don't sit inside rate and liquidity risk alone. They run across credit, capital, and profitability at once. The teams that have made the jump are able to answer them in the room where those discussions occur rather than in the materials that are circulated afterward. A good place to start is what types of tangential analytics you can pull into funds transfer pricing discussions, and shortening the time it takes to iterate on analyses — some institutions have even been able to shorten it so much that they can run some things live in the ALCO meeting.

What best practices enable firms to effectively connect balance sheet analytics with customer, product, and segment-level economics to drive more sustainable profitability?

The work is measuring profitability at the right level of granularity, at the relationship level rather than the chart of account level. Most teams know the portfolio average hides as much as it shows; the harder part is doing something about it. The funding piece is what comes out of ALM, through matched-maturity FTP set at the instrument level and with thoughtful behavioral parameters so it reflects how those clients actually behave rather than when the instrument contractually matures. That splits net interest margin into pieces someone can be accountable for: the spread the business earns on its assets, the value of the deposits it brings in, and the rate and liquidity residual that stays with Treasury. But a relationship's economics don't stop at funding. The same customer carries a probability of default, an expected loss, and a capital charge. Whether to grow it or reprice it isn't clear until all of that is really understood. That's when it gets interesting, the loan that looked great on spread can turn out to lose money once you count the rest of the customer around it, while the thin-margin loan a bank was ready to pass on, is actually the one anchoring a sticky, low-beta operating deposit. Sustainable profitability comes from acting on that ranking, and the ranking only holds up when it's built across funding, credit, and capital together instead of inside any one of them.

How can institutions ensure that balance sheet insights are translated into timely, actionable business decisions rather than remaining as analytical outputs?

This is the question our session was built around. Most balance sheet analysis stalls in the space between producing a number and acting on it, the synthesis gap between information and a decision. A lot of why it stalls is that many decisions at banks actually span domains: a rate move changes debt coverage, weaker coverage moves the probability of default and the capital supporting the loan, and that resets the price the loan should carry. No single system owns that entire chain, so the number tends to arrive only after someone assembles it by hand. Instead of that, the analysis that closes the gap shows up already measured against the bank's own limits, with the escalation it would trigger attached, and on the committee's own cadence. It carries its interpretation with it, so what lands in front of people is the implication and not one more sensitivity run. Timing counts as much as form: an intraday read on what a hedge or a deposit-rate move does to EVE and net interest income is useful before the ALCO votes and academic a week later. And it has to land with someone who can act, since an insight without an owner rarely becomes a decision.

What does "good" look like when leveraging AI and advanced analytics to enhance decision-making speed and quality, while maintaining governance and control?

Good keeps a clear line between the analytics and the AI. The engines a bank already trusts, the ones that have been through validation and exams, stay the system of record for each risk, and the AI doesn't recompute them. Its use is connective: putting the funding, credit, capital, and profitability views into one picture and letting someone question that picture in plain language and get an answer they can stand behind. That connective role is where the speed comes from, and what separates it from a good demo is that the work is always shown. The numbers trace back to the engine and the data behind them, the same question returns the same answer, and the decision stays with a person rather than the model. All of it sits inside the model-risk governance already in place, so speed doesn't come at the expense of control. The case worth designing against is the confident, well-formed answer that happens to be wrong, and the value of good analytics is that they make it easy to catch.

What are the key components of a best-in-class integrated risk, finance, and business intelligence framework, and how do these capabilities drive better capital allocation and competitive advantage?

Everything we've described sits on one governed foundation, so risk, finance, and the front line all work from the same entities and the same numbers as of the same date. That takes away the familiar situation where two functions show up defending different numbers for the same exposure. On top of that, the economics get computed once and reused, so a single deal's numbers reconcile up to the portfolio and up to the plan without anyone rebuilding them along the way. All three views run on the same set of assumptions, so the PD behind the impairment number is the same one credit analytics is using, and a single scenario moves credit, provisions, net interest income, EVE, and capital together. Above all of it, there's a decision layer that hands the business the combined answer no single engine produces on its own. That's where capital allocation gets better, because when every deal carries the same kind of return, adjusted for capital and funding, the bank can rank what's competing for the next scarce dollar and keep steering it toward the best use. The trusted-data layer has become table stakes. What's scarce now, and where we've chosen to work, is the last mile from trusted context to a decision a committee can act on.