目录
AI Treasury / Liquidity / ALM Forecasting / Stress Evidence Architecture Playbook
适用对象: CBAP-level Financial Retail PM / Senior BA / Treasury Product Owner / ALM Architect / Risk Data Product Lead / AI Governance / Model Risk / Enterprise Architect / Board MI Lead。
目标: 把 treasury、liquidity forecasting、deposit beta/runoff、cash-flow ladder、ALM、IRRBB、FTP、portfolio scenario、stress testing、contingency funding、human committee、board MI 和 data lineage 设计成可落地、可运营、可审计的 AI 产品架构能力。
核心观点: Treasury AI 的交付物不是预测数字, 而是 forecast-to-action system: 数据血缘、模型证据、压力假设、解释、人工挑战、行动记录和董事会管理信息。
0. Boundary And Disclaimer
本文是学习、作品集和架构设计材料, 不构成法律意见、监管解释、模型验证结论、流动性充足性结论、资本/财务确认、投资建议、融资执行建议、会计意见、审计意见或董事会治理意见。
正式项目中的适用范围、监管口径、报告义务、模型治理要求、客户数据使用、披露和委员会权限由 Legal、Compliance、Regulatory Affairs、Model Risk、Treasury、Finance、Risk、Internal Audit 和管理层权责方确认。本文只提供产品与架构设计框架。
Product boundary:
AI 可以预测、解释、模拟、生成候选情景、提醒异常、起草 ALCO/board MI 初稿。
AI 不应独立批准或执行 funding action、asset sale、hedging、FTP policy change、customer pricing、contingency funding activation、监管报告结论或董事会风险偏好变更。
任何影响资金、流动性、客户、财务报告、监管报告、风险偏好或重大经营决策的动作都必须进入 human committee、dual control、limit check 和 evidence process。
1. Executive Framing
弱项目通常这样定义:
Build an AI model to forecast deposits and liquidity.
成熟项目应该这样定义:
Build a governed treasury intelligence and action-evidence platform
that connects cash-flow forecasts, deposit behavior assumptions,
ALM/IRRBB scenarios, FTP incentives, stress testing, CFP actions,
committee decisions and board MI.
Executive one-liner:
This is not a predictive analytics dashboard. It is a controlled liquidity decision system with evidence.
1.1 Steering Committee Questions
哪些 treasury / ALM 决策会被 AI 输出影响: daily liquidity, ALCO, FTP, pricing, portfolio, CFP, board MI?
每个预测的 grain、horizon、cadence、owner、allowed action 和 prohibited use 是否清楚?
Deposit beta/runoff 是按什么 segment、scenario、rate regime 和 relationship evidence 定义?
Cash-flow ladder 是否区分 contractual、behavioral、operational、contingent 和 intraday liquidity?
ALM / IRRBB 的 NII/EVE 结果如何使用 AI assumptions, 谁批准 assumption change?
FTP 是否把 funding risk、liquidity premium 和 contingent liquidity cost 反馈到产品/业务线激励?
压力情景、人工覆盖和 committee challenge 是否能被重放?
Board MI 的每个数字是否可以追溯到数据、模型、情景、阈值、行动和 owner?
2. Source Anchors
访问日期: 2026-06-30。
Source-to-control pattern:
official source
-> control objective
-> product requirement
-> technical control
-> evidence artifact
-> owner
-> monitoring metric
3. Target Operating Model
Treasury AI 要作为 cross-functional control product 管理。
3.1 Operating Forums
Forum Primary purpose AI product responsibility Treasury daily management cash positioning, funding needs, collateral movement within approved limits daily forecast, intraday alerts, limit status, action evidence ALCO liquidity, ALM, IRRBB, FTP, balance-sheet strategy scenario pack, assumption changes, NII/EVE, FTP and portfolio impacts Treasury Risk / Market Risk independent challenge, limits, stress review challenger outputs, limit breach evidence, issue tracking Model Risk Committee model inventory, validation, ongoing monitoring model pack, validation evidence, monitoring, change approval AI Governance / AIMS forum enterprise AI policy, controls, incident, performance AI RMF / ISO 42001-aligned control evidence Board Risk / Audit Committee oversight, risk appetite, management action challenge board MI data product and evidence lineage
3.2 RACI
Activity Treasury ALM Risk Model Risk Data Technology Legal/Compliance Internal Audit Forecast contract approval A C C C C C C I Deposit assumption owner C A C C C I I I Stress scenario library A A A C C C C I Model validation C C C A C C I I Data lineage and quality C C C C A R I I Action execution A C C I I R C I Board MI sign-off A A A C C C C I Control testing I I C C C C C A
Legend: A = accountable, R = responsible, C = consulted, I = informed.
4. Reference Architecture
1. Source and lineage layer
core banking, deposits, loans, payments, cards, GL, treasury trades,
securities, collateral, market data, pricing, CRM, legal entity, limits
2. As-of treasury data products
deposit balance and flow facts
contractual cash-flow schedules
behavioral cash-flow features
payment and settlement calendars
rate curve and competitor rate products
collateral and liquidity source inventory
3. AI / analytics layer
baseline forecast
deposit beta / runoff
cash-flow forecast
anomaly / early-warning
scenario generation
challenger models
explanation services
4. Risk and finance engines
liquidity ladder
stress survival horizon
ALM / IRRBB NII and EVE
FTP allocation and simulation
portfolio scenario and limit impact
5. Decision workbench
Treasury dashboard
ALCO scenario workbench
action recommendation with constraints
override and committee challenge
dual-control action register
6. Evidence and MI plane
forecast contracts
model inventory and validation
scenario versioning
assumption register
data lineage graph
action log
board MI metric contracts
audit export
Architecture rule:
If a forecast can change a liquidity action, it must have a contract, owner, lineage,
model-risk status, scenario context, decision boundary and evidence trail.
5. Core Product Modules
5.1 Treasury Data Product Layer
Module What it provides Evidence Deposit balance and flow mart daily/intraday balances, inflows/outflows, product, segment, legal entity source mapping, cutoff, reconciliation, restatement log Cash-flow schedule service contractual loan/security/funding/coupon/maturity flows source contract, schedule version, exception handling Payment and settlement calendar ACH/wire/card/check/payroll/tax/holiday settlement timing calendar version, rail owner, holiday logic Rate and market data product Fed/market curves, product rates, competitor rates, spreads source, timestamp, curve construction Collateral and liquidity source inventory unencumbered securities, eligibility, haircut, operational availability custody source, encumbrance, legal entity Customer concentration graph customer group, relationship, top depositor exposure, concentration entity resolution logic, group owner Limit and risk appetite registry liquidity, IRRBB, concentration, model, data and action limits threshold owner, approval date, review cadence
5.2 Forecast And Scenario Layer
Module Product behavior Baseline model suite naive/seasonal, business rules, statistical and ML baselines for each forecast contract Deposit beta/runoff engine segment and scenario-level behavioral assumptions with uncertainty Cash-flow forecast engine contractual + behavioral + operational + contingent forecasts Early-warning monitor anomalous flow, concentration movement, rate sensitivity, competitor gap, event signals Scenario library base, idiosyncratic, market-wide, combined, reverse stress, operational and intraday Challenger framework independent model or rules that challenge primary output Explainability service driver bridge, segment contribution, scenario sensitivity, action implication
5.3 Decision And Evidence Layer
Module Product behavior ALCO workbench compare scenarios, limits, NII/EVE, liquidity ladder, FTP impacts and actions Action register proposed action, owner, approval forum, constraint check, execution status, closure evidence Overlay workflow human adjustment with reason, scope, expiry, approver and impact Committee challenge log challenge question, response, decision, residual risk owner Board MI generator metric contracts, thresholds, status, drivers, actions and evidence links Audit evidence export selected sample can reconstruct source-to-decision chain
6. Decision Gates
Gate 0: Use-Case Eligibility
Question Pass condition Evidence Which decision will the forecast influence? Decision and owner documented Forecast Contract Could the output affect funding, pricing, liquidity, ALM, FTP or board MI? Impact tier assigned Use Case Risk Tier Is AI needed, or can deterministic rules / existing ALM process answer it? Alternatives recorded Alternatives note What actions are prohibited? Explicit action boundary Action Boundary Card Who owns applicability and policy interpretation? Legal/Compliance/Model Risk owners named Governance routing record
Gate 1: Data Readiness
Question Pass condition Evidence Are sources reconciled to GL / authoritative systems where needed? Reconciliation pass or issue accepted Reconciliation report Is event time separate from available time? As-of design implemented Data lineage record Are product/segment/legal-entity mappings stable? Mapping versioned and approved Mapping catalog Are restatements and corrections tracked? Change log exists Restatement log Are sensitive data and customer-level signals governed? Data governance route completed Data use record
Gate 2: Model And Assumption Design
Question Pass condition Evidence Is there an honest baseline? Baseline model defined Baseline benchmark Are outputs probabilistic where action risk requires it? Interval/quantile output available Forecast output spec Is deposit beta/runoff segmented? Segment and scenario rationale documented NMD Assumption Register Are challenger models or expert overlays available? Challenger/overlay workflow active Challenger report Are limitations explicit? Model card and limitation log approved Model risk pack
Gate 3: Stress And ALM Integration
Question Pass condition Evidence Are scenarios versioned and approved? Scenario library entry approved Scenario Definition Card Does liquidity ladder include contingent and operational flows? Full flow taxonomy implemented Ladder specification Do NII/EVE outputs use approved assumptions? ALM assumption linkage exists ALM assumption trace Is FTP impact calculated where incentives matter? FTP simulation integrated FTP impact report Are legal-entity/currency constraints visible? Constraint layer enabled Constraint report
Gate 4: Release And Governance
Question Pass condition Evidence Can ALCO challenge the output? Workbench shows drivers, alternatives, limitations ALCO pack sample Are human overrides controlled? reason, approver, expiry and impact captured Overlay record Are actions dual-controlled? approval and execution separated Action register Can board MI be traced to evidence? metric contracts and lineage links exist Board MI lineage sample Is monitoring live? performance, drift, data quality, issue aging active Monitoring dashboard
7. Required Artifacts
Artifact Minimum contents Forecast Contract decision use, grain, horizon, cadence, owner, outputs, allowed actions, prohibited use, evidence Treasury Data Contract source systems, fields, event time, available time, quality rules, reconciliation, retention NMD Assumption Register beta, runoff, average life, repricing rate by product/segment/scenario, owner, approval Scenario Definition Card scenario type, horizon, rate path, runoff shock, funding shock, collateral haircut, operational assumption Model Card purpose, design, data, performance, limitations, validation, monitoring, approved use ALCO Workbench Spec metrics, scenarios, explanations, limits, action options, challenge fields FTP Impact Report products/business lines affected, funding cost, liquidity premium, contingent liquidity cost Overlay Record reason, scope, duration, approver, quantitative impact, expiry and post-review Action Register trigger, proposed action, approval forum, execution status, owner, closure evidence Board MI Metric Contract definition, source lineage, threshold, owner, cadence, decision use, action link Evidence Binder source-to-decision graph, data snapshot, model/scenario version, committee record, action log
8. Data Contracts
8.1 Deposit Flow Contract
Field Definition account_id stable account key or governed token customer_group_id relationship / household / commercial group, versioned legal_entity booking entity product_type standardized deposit taxonomy event_time when balance-affecting event occurred available_time when event was available to forecast system amount signed inflow/outflow amount balance_after account balance after event rate_paid current paid rate at event time pricing_event_id linked campaign / product rate change if applicable channel branch, online, wire, ACH, internal transfer, sweep quality_status pass, warning, excluded, reconciled
8.2 Cash-Flow Forecast Output Contract
Field Definition forecast_id unique forecast run forecast_contract_id governing contract as_of_time data cutoff scenario_id base or stress scenario time_bucket intraday/day/week/month bucket legal_entity / currency liquidity transfer constraints flow_type contractual, behavioral, operational, contingent p10 / p50 / p90 probabilistic outputs driver_ref explanation artifact model_version model and feature version limitation_flag known caveat for use
8.3 Decision And Action Contract
Field Definition trigger_id limit breach, warning, scenario result or committee request proposed_action funding, pricing, collateral, portfolio, CFP, monitoring action_boundary read-only, recommend, committee approval, dual-control execution owner accountable decision owner approval_forum Treasury, ALCO, Risk, Board, other approved forum rationale structured reason and evidence links constraint_check legal entity, currency, collateral, operational, policy execution_status proposed, approved, rejected, executed, closed closure_evidence execution confirmation or reason for no action
9. Model And Scenario Selection
Use case Preferred model family Required controls Daily cash position rules + statistical / ML forecast data freshness, cutoff, reconciliation, exception alerts Intraday liquidity event-driven forecast + payment rail calendar low-latency monitoring, operational dependency, fallback Retail NMD runoff survival / hazard, hierarchical, gradient boosting, probabilistic TS segment logic, beta/runoff register, challenger, stress overlay Commercial deposit concentration concentration graph + anomaly detection + relationship features top depositor lineage, group mapping, human review Rate sensitivity / beta elasticity models, regime-aware regression, causal sensitivity rate campaign leakage check, competitor rate context Loan prepayment borrower behavior model + rate incentive ALM assumption approval, validation, scenario response Commitment drawdown historical utilization + stress overlays contingent liquidity evidence, expert challenge ALM NII/EVE existing ALM engine with AI assumption challenger assumption trace, NII/EVE bridge, model risk controls FTP simulation rule/curve-based engine + scenario inputs methodology transparency, business-line incentive review Board MI narrative source-grounded LLM summarizer approved data only, citations, human review, no unsupported conclusion
Selection rule:
Use the simplest model that can support the decision, uncertainty, stress behavior,
explainability and evidence requirement.
10. Forecast-To-Action Matrix
Signal Example threshold Required action Owner Evidence Green Forecast within normal band, data quality pass normal monitoring Treasury dashboard snapshot Watch segment runoff rising but within appetite analyst review and commentary Treasury / ALM driver bridge Amber P90 runoff or survival horizon breaches internal watch threshold ALCO / Treasury Risk review ALCO owner scenario pack, challenge log Red stress survival horizon below appetite or severe concentration movement management escalation, CFP option review Treasurer / CRO action register, committee record CFP trigger approved contingency threshold reached activate approved CFP workflow authorized management forum CFP action evidence Data quality block source reconciliation fails for material input block board MI or label limitation Data Owner / Technology quality issue log Model limitation model drift or interval coverage failure restrict use, apply approved overlay, remediate Model Owner / Model Risk model issue record
Action principles:
A red signal without an action owner is not a control.
A model limitation without a use restriction is not transparent governance.
A human overlay without expiry becomes shadow methodology.
A board metric without source lineage is weak MI.
11. Evidence And Control Checklist
Control objective Product control Evidence Owner Forecast used for approved purpose Forecast Contract enforced in UI/API contract id on every forecast output PM / Treasury Data is as-of correct event_time and available_time captured lineage and cutoff report Data Owner Data quality is known completeness/freshness/reconciliation rules quality dashboard and issue log Data / Technology Deposit assumptions are governed NMD Assumption Register approval and version history ALM Model is validated for intended use model card, validation and challenger validation report and issue log Model Risk Stress scenario is reproducible scenario library with parameters scenario definition card Risk / Treasury AI explanation is grounded explanation links to data/scenario/model refs driver decomposition artifact Product / Data Science Committee challenge is real challenge questions and responses captured ALCO / Risk minutes Committee Secretary / Owner Human overlay is controlled reason, approver, expiry, impact overlay record Treasury / Risk Actions are authorized dual control and limit checks action register Treasury Operations Board MI is traceable metric contracts and lineage links board MI lineage pack MI Owner Issues are closed action owner and due date remediation log Risk / Audit
12. Monitoring Metrics And KRIs
Metric Why it matters forecast error by horizon and segment 看模型在哪里失效, 不只看全行平均 prediction interval coverage 检查 P90 是否真的覆盖 tail behavior directional accuracy around stress windows 检查预警价值 calibration by product and legal entity 防止某些实体/产品持续偏差 challenger disagreement rate 衡量模型不确定性和有效挑战
12.2 Liquidity And ALM KRIs
Metric Decision use stress survival horizon by scenario liquidity appetite and CFP readiness top depositor concentration movement concentration risk and early warning rate-sensitive deposit share pricing and runoff vulnerability beta / runoff assumption drift ALM and FTP assumption review contingent liquidity exposure CFP and FTP charge review NII / EVE sensitivity bridge ALCO balance-sheet strategy usable liquidity by legal entity/currency transferability and constraint visibility collateral encumbrance and haircut sensitivity stress funding capacity
12.3 Governance Metrics
Metric Why it matters model validation status by use case release and continued use data quality exception aging MI reliability human override rate and reason mix model trust, business change, shadow policy action aging after amber/red signal management responsiveness evidence completeness score audit and board readiness expired assumption count ALM discipline unresolved model limitations use restriction and residual risk
13. Board MI Pack Design
Board pack should show decision-useful facts, not raw model outputs.
13.1 Minimum Board View
Section Content Executive liquidity status green/amber/red, key change since last period, decision requested Liquidity buffer and survival base and stress survival horizon, usable liquidity by constraint Deposit behavior beta/runoff by key segment, top depositor concentration, digital/high-rate segment movement ALM/IRRBB NII/EVE sensitivity and assumption changes FTP / portfolio incentives major products or business lines where incentives changed CFP readiness tested capacity, action execution time, dry-run findings AI/model risk model status, limitations, challenger disagreement, validation issues Management actions approved, rejected, executed, overdue, residual risk owner
13.2 Board MI Tile
Topic: 30-day liquidity outlook under idiosyncratic deposit confidence stress
Status: Amber
Metric: stress survival horizon 43 days, down 11 days from prior quarter.
Drivers: P90 runoff increase in digital money-market and high-yield promotional cohorts; competitor-rate spread widened; no material change in top depositor concentration.
ALM impact: higher retention pricing improves runoff outcome but reduces projected NII; EVE sensitivity unchanged within appetite.
FTP impact: promotional deposit stability credit reduced; contingent liquidity charge increased for affected campaign balances.
Controls: challenger model confirms direction; primary model magnitude higher; 60-day expert overlay approved with expiry.
Management action: ALCO approved targeted retention for operational relationship customers and a CFP operational dry-run.
Evidence: data cutoff, model version, scenario id, overlay record, ALCO decision and action owner linked.
Decision requested: note amber status and challenge CFP dry-run completion timeline.
14. Tabletop Scenarios
Scenario A: Digital Deposit Run Triggered By Confidence Event
Element Design Trigger unusual digital outflow, call center concern, media/news signal, top depositor inquiry AI role early warning, segment contribution, P90 runoff forecast, scenario comparison Human action Treasury Risk review, ALCO escalation, CFP option assessment Evidence event sources, customer segment movement, scenario version, action log Failure to test AI overreacts to noisy NLP signal or misses concentration transfer
Scenario B: Rapid Rate Hike And Competitor Pricing Shock
Element Design Trigger market curve shift and competitor rate spread widening AI role beta update, migration forecast, NII/liquidity trade-off Human action pricing committee and ALCO review Evidence rate data source, pricing campaign linkage, beta assumption change Failure to test broad promotional pricing improves balances but destroys margin and attracts unstable funds
Scenario C: Payment Rail / Settlement Disruption
Element Design Trigger ACH/wire/card settlement delay or treasury platform degradation AI role intraday liquidity impact, operational cash-flow reroute options Human action incident command, treasury operations, BCP invocation Evidence payment event logs, BCP record, liquidity source availability Failure to test liquidity exists on paper but cannot be moved operationally
Scenario D: Collateral Haircut And Wholesale Funding Stress
Element Design Trigger market liquidity shock, securities haircut change, rollover pressure AI role haircut sensitivity, funding source availability, survival horizon Human action collateral strategy, funding diversification review Evidence collateral inventory, encumbrance, haircut scenario, funding action feasibility Failure to test stress assumes asset sale that is not operationally or market-feasibly executable
Scenario E: Model Drift During Benign Period
Element Design Trigger low error in aggregate but poor interval coverage for key segment AI role monitoring flags calibration gap and challenger disagreement Human action restrict use for affected segment, approve overlay, start remediation Evidence backtest, interval coverage, model limitation log, overlay expiry Failure to test good portfolio-level accuracy masks material risk pocket
15. Implementation Roadmap
Phase 1: Foundation, 0-30 Days
Workstream Deliverable Scope prioritized use cases: daily liquidity, NMD runoff, ALCO stress pack, board MI Governance forecast contracts, action boundaries, RACI, committee routing Data source inventory, data contracts, lineage gaps, reconciliation requirements Baseline simple rules/seasonal baseline for key cash-flow and deposit forecasts Evidence draft evidence graph and board MI metric contracts
Phase 2: Pilot, 31-60 Days
Workstream Deliverable Deposit behavior segment beta/runoff prototype with uncertainty Liquidity ladder contractual + behavioral + operational + contingent flow view Scenario base, idiosyncratic, market-wide and combined scenario cards ALCO workbench scenario comparison, drivers, limits, override capture Model risk model cards, validation plan, challenger baseline, monitoring metrics
Phase 3: Controlled Release, 61-90 Days
Workstream Deliverable Production data automated lineage, quality gates, reconciliation and data issue workflow Decision workflow action register, committee challenge log, dual-control integration ALM/FTP NII/EVE bridge and FTP simulation for selected products Board MI traceable board tile and evidence export Assurance tabletop scenario, dry-run, issue remediation and release sign-off
Phase 4: Scale
Workstream Deliverable Portfolio multi-entity, currency, product and business-line expansion Stress reverse stress, intraday stress, operational stress and CFP dry-run library Governance AIMS integration, internal audit testing, continuous monitoring Optimization scenario-aware portfolio/pricing/funding option analysis under constraints
16. Anti-Patterns
Anti-pattern Why it fails Better practice "AI predicts deposits" as project scope 预测没有决策边界 use-case-specific forecast contracts Single dashboard for Treasury, ALM, Board 不同决策需要不同 evidence grain role-specific MI from same data product Model-first delivery 忽略 committee, actions, controls forecast-to-action workflow first P50 forecast as liquidity truth tail risk is the point quantiles, stress and survival horizon Deposit beta as one scalar hides segment, rate and relationship behavior segmented assumption register Manual ALCO pack no reproducible evidence metric contracts and lineage LLM-generated narrative without source grounding plausible but unauditable grounded summary with evidence links and human review No challenger model no effective challenge honest baseline and challenger dashboard Unlimited human overrides shadow methodology controlled overlay with reason, expiry and post-review CFP actions not tested plan may not execute under stress dry-run evidence and operational readiness
17. Interview And Portfolio Language
Q1: What would you build first for AI treasury liquidity forecasting?
30 秒版本:
I would build the forecast contract and evidence spine first, not the most advanced model. The first release should define decision use, grain, horizon, data lineage, baseline forecast, deposit behavior assumptions, scenario definitions, ALCO challenge workflow and board MI metric contracts.
Q2: How do you prevent AI from becoming a black box in ALM?
30 秒版本:
I separate AI assumptions from ALM engines. AI can challenge NMD beta, runoff, prepayment and scenario assumptions, but approved ALM assumptions remain versioned governance objects. ALCO sees driver bridge, challenger comparison, uncertainty, limitations and override records before using them in NII/EVE decisions.
Q3: What is a good evidence pack for stress testing?
30 秒版本:
A good pack includes scenario definition, data cutoff, model and feature versions, assumptions, overlays, liquidity ladder output, ALM impact, FTP impact, limit status, committee challenge, management actions and post-event backtesting. The key is source-to-action traceability.
Q4: How would you connect FTP to AI forecasts?
30 秒版本:
I would use AI to estimate behavior and stress sensitivity, then feed approved assumptions into FTP. Stable operational deposits may receive funding benefit only when stability evidence supports it; promotional or rate-sensitive balances receive lower stability credit and potentially higher contingent liquidity charges. FTP becomes the incentive layer for liquidity risk.
Q5: How do you phrase this for senior stakeholders?
30 秒版本:
The goal is not to automate Treasury. The goal is to give Treasury, ALCO and the board a faster and better-evidenced view of liquidity risk, assumption changes and management options, while preserving human decision rights and auditability.
18. Final Checklist
Question Good answer Can every forecast output identify its contract? yes, forecast_contract_id is attached Can the team replay a board liquidity metric? yes, metric -> data -> model -> scenario -> decision -> action Are deposit assumptions segmented and approved? yes, NMD register has owner, version and evidence Are stress scenarios reproducible? yes, scenario cards are versioned Can ALCO challenge model outputs? yes, drivers, alternatives and limitations are visible Are overrides controlled? yes, reason, impact, approver, expiry and review exist Are actions authorized and tracked? yes, action register links to trigger and evidence Is model risk live after release? yes, monitoring, drift, backtest, issue and limitation tracking run continuously Is exact regulatory applicability owned by the right functions? yes, Legal/Compliance/Model Risk own formal applicability and conclusions
Memory card:
AI Treasury Playbook =
forecast contracts
+ as-of data lineage
+ deposit behavior assumptions
+ liquidity ladder
+ ALM / IRRBB / FTP integration
+ stress evidence
+ human committee
+ board MI
+ continuous model risk monitoring.