Digital Finance, Sustainability and the Future of Global Treasury

Introduction

International financial management is being reshaped by digital technologies, environmental pressures and rising expectations of responsible corporate conduct. The traditional treasury function—focused primarily on cash, foreign exchange, liquidity and financial risk—is becoming a digitally connected control centre. Modern global treasurers must understand fintech platforms, real-time payment infrastructure, blockchain-based settlement, central bank digital currency experiments, cloud treasury systems, data analytics and artificial intelligence (AI). At the same time, they must evaluate cybersecurity, data protection, model risk, climate exposure and environmental, social and governance (ESG) performance.

Digitalisation does not eliminate the fundamental objectives of treasury. A multinational enterprise still seeks to preserve liquidity, reduce financing and transaction costs, manage currency and interest-rate risk, ensure compliance and support strategic investment. Digital technologies change how these objectives are achieved: information can be processed faster, payments can be automated, exposures can be identified more accurately and controls can be embedded into workflows. However, digitalisation also creates new risks, including cyberattacks, technology concentration, fraud, algorithmic bias, data misuse and operational dependency on external providers.

Sustainability is similarly not an optional public-relations issue. Climate change can affect asset values, supply chains, insurance costs, borrowing conditions, exchange-rate pressures and the continuity of operations. Treasury therefore contributes to both financial resilience and the financing of an organisation’s transition to a lower-carbon economy.

Learning Objectives

By the end of this chapter, students should be able to:

1. Explain the role of fintech, blockchain, digital currencies and real-time payments in international treasury.
2. Evaluate treasury management systems, data analytics and responsible AI applications.
3. Identify cybersecurity, model-governance and data-ethics requirements.
4. Analyse how ESG factors and climate risk influence international financial decisions.
5. Apply integrated governance principles to the design of a future-oriented global treasury.

1. The Changing Architecture of Global Treasury

A multinational treasury traditionally performs five connected functions:

• Liquidity management: ensuring that subsidiaries and the parent company have sufficient funds at the right time and place.
• Foreign-exchange management: measuring and managing transaction, translation and economic exposure.
• Funding: selecting bank loans, bonds, commercial paper, intercompany finance and equity-related sources.
• Financial risk management: controlling interest-rate, commodity, counterparty and settlement risks.
• Governance and control: establishing policies, authorisations, segregation of duties, reporting and compliance.

Digital treasury adds a sixth dimension: information architecture. Treasury decisions depend on the quality, timeliness, accessibility and security of data. A payment instruction, forecast or hedge recommendation is only as reliable as the underlying data and the controls surrounding its use.

A useful conceptual chain is:

Data → information → analysis → decision → transaction → monitoring → learning

Weakness at any stage can produce financial loss. For example, a system may execute foreign-exchange trades efficiently but still create losses if subsidiary forecasts are incomplete or if a currency exposure is incorrectly classified.

Important terminology

Fintech refers to technology-enabled financial products, services and infrastructures. Examples include digital wallets, application programming interfaces (APIs), automated lending, embedded finance and electronic foreign-exchange platforms.

Open banking is the regulated or consent-based sharing of financial data and payment functionality through secure APIs. It can improve visibility but introduces third-party, consent and privacy risks.

Embedded finance integrates financial services into a non-financial commercial platform. A global marketplace, for example, may provide foreign-exchange conversion and supplier payments within its own procurement system.

Tokenisation represents an asset or claim digitally on a distributed ledger. Tokenisation may improve transferability and settlement efficiency, but the legal rights attached to a token must be clearly established.

2. Fintech and Digital Payment Infrastructure

Fintech affects treasury in three principal ways. First, it creates new channels for moving money. Second, it improves access to financial information. Third, it automates decisions and controls.

Real-time payments

Real-time payment systems transfer funds within seconds or minutes and normally operate continuously, including outside traditional banking hours. Their benefits include improved working-capital visibility, faster supplier settlement, reduced reliance on cash and enhanced customer experience. For multinational firms, they may also reduce the time between collection and deployment of funds.

The risks are equally important. Real-time payments compress the period available for detecting fraud, correcting errors or recalling funds. A fraudulent payment that would previously have been reviewed overnight may now be irrevocable before detection. Treasury policy should therefore include beneficiary verification, transaction limits, strong authentication, segregation of duties and continuous monitoring.

APIs and data connectivity

APIs allow systems to exchange structured data. A treasury management system (TMS) can use bank APIs to retrieve balances, issue payment instructions and obtain transaction confirmations. Connectivity reduces manual reconciliation and can provide near-real-time cash visibility.

API governance should address authentication, authorisation, encryption, availability, version control, vendor dependency and incident response. A successful integration is not merely a technical connection; it is a controlled business process.

Digital foreign-exchange markets

Electronic trading platforms provide price comparisons, execution records and automated access to liquidity providers. Algorithmic execution may reduce transaction costs, but algorithms must be monitored for inappropriate trading behaviour, poor liquidity assumptions and excessive market impact. A firm should distinguish between an algorithm that executes an approved hedge and one that independently determines the firm’s risk appetite. The latter requires much stronger governance.

3. Blockchain, Distributed Ledgers and Digital Currencies

A blockchain is a type of distributed ledger in which transactions are recorded across a network according to agreed rules. Consensus mechanisms determine how the network accepts new records. Cryptographic techniques help protect the integrity of the ledger, while smart contracts can automate actions when specified conditions are met.

Potential treasury applications

1. Cross-border settlement: Distributed ledgers may reduce the number of intermediaries involved in selected transactions.
2. Trade finance: Digital records can connect invoices, shipping documents, insurance and payment conditions.
3. Collateral management: Tokenised securities may allow more efficient movement and monitoring of collateral.
4. Intercompany reconciliation: A shared ledger can reduce disputes about balances and transaction status.
5. Programmable payments: A payment may be released automatically when delivery or documentation conditions are verified.

These applications should not be treated as automatically superior to conventional systems. A blockchain may improve shared visibility but perform poorly where transactions require confidentiality, rapid reversal or high throughput. Governance, legal enforceability and interoperability remain decisive.

Digital currencies

Digital currencies include privately issued cryptocurrencies, stablecoins and central bank digital currencies (CBDCs). A stablecoin attempts to maintain a stable value relative to an asset such as a national currency, but stability depends on reserve quality, redemption arrangements, governance and regulation. A CBDC is a digital liability of a central bank, whereas commercial-bank money is a liability of a commercial bank.

For international treasury, possible advantages include faster settlement, reduced correspondent-banking friction and improved access to foreign markets. Risks include exchange-rate volatility, reserve or issuer risk, sanctions exposure, operational failure, privacy concerns and uncertain legal treatment. A treasury should assess the legal claim represented by a digital instrument rather than focusing only on its technological design.

Smart contracts

A smart contract is software that executes agreed logic on a digital platform. It is not necessarily a complete legal contract. A smart contract may automatically transfer funds when an external data source reports that a shipment arrived, but the reliability of that external source—the “oracle”—is critical. Errors in code may be executed at scale and may be difficult to reverse.

4. Treasury Management Systems and Data Analytics

A TMS integrates cash management, payments, forecasting, funding, risk management, accounting interfaces and reporting. Enterprise resource planning systems usually capture operational transactions, while a TMS specialises in treasury processes. The boundary varies by organisation.

A robust TMS should provide:

• centralised exposure and liquidity data;
• role-based access and approval workflows;
• bank connectivity and payment controls;
• audit trails;
• forecasting and scenario analysis;
• hedge and debt administration;
• reconciliation and exception management; and
• resilience through backup, recovery and continuity arrangements.

Cash forecasting

A simple cash forecast for period t can be represented as:

Here, Cₜ is closing cash in period t, Cₜ₋₁ is opening cash, Iₜ is operating and investing inflows, Oₜ is operating and investing outflows, and Fₜ is net financing cash flow. All terms must be measured in the same currency and time period. The identity is an accounting-style cash-flow relationship, not a forecasting method by itself. Forecast quality depends on the assumptions used to estimate inflows, outflows and financing.

Forecast accuracy can be evaluated using mean absolute error:

where Aₜ is actual cash flow, Fₜ is forecast cash flow and n is the number of observations. MAE is expressed in currency units and treats positive and negative errors equally after taking absolute values. It is useful for comparing forecasting performance, although it does not show whether forecasts are systematically optimistic or pessimistic.

Data analytics

Descriptive analytics explains what has happened, diagnostic analytics investigates why it happened, predictive analytics estimates what may happen, and prescriptive analytics recommends action. Examples include:

• identifying dormant bank accounts;
• detecting unusual payment beneficiaries;
• forecasting foreign-currency receipts;
• estimating liquidity under stress scenarios; and
• recommending the timing or source of funding.

Analytics should support—not replace—professional judgement. A model trained on stable historical data may fail during a war, pandemic, sanctions event or sudden interest-rate shock. Treasury therefore needs scenario analysis and human review in addition to statistical prediction.

5. Artificial Intelligence and Responsible Automation

AI can assist treasury with invoice classification, cash-flow forecasting, fraud detection, document extraction, anomaly detection, credit assessment and natural-language reporting. Machine-learning systems may identify relationships that are difficult to specify through conventional rules.

Responsible AI requires at least six controls:

1. Purpose limitation: the model must have a clearly defined business purpose.
2. Data quality: input data should be accurate, relevant, timely and representative.
3. Explainability: users should understand the factors materially influencing an output.
4. Validation: performance should be tested before deployment and periodically thereafter.
5. Human accountability: a responsible person must own decisions and exceptions.
6. Monitoring: drift, bias, error rates, cyber manipulation and unexpected outcomes must be observed.

Model governance

Model risk is the possibility that a model produces an incorrect or unsuitable result because of flawed data, assumptions, coding, implementation or use. Governance should include a model inventory, documented purpose, approval thresholds, independent validation, version control, performance metrics, change management and retirement procedures.

For a forecasting model, back-testing compares previous forecasts with outcomes. A model should not be judged only by average accuracy; tail errors and performance during stressed conditions may be more important to treasury.

Generative AI introduces additional concerns. It may produce plausible but incorrect explanations, expose confidential information through inappropriate prompts or reproduce hidden biases. It should not be allowed to originate or approve high-value payments without controlled verification.

6. Cybersecurity, Data Ethics and Operational Resilience

Digital treasury expands the attack surface. Threats include phishing, ransomware, credential theft, business-email compromise, payment diversion, insider abuse, API exploitation and attacks on cloud providers.

A layered control system includes:

• multi-factor authentication;
• least-privilege access;
• separation of payment preparation and approval;
• privileged-access monitoring;
• encryption in transit and at rest;
• secure API gateways;
• independent confirmation of new beneficiaries;
• transaction velocity and anomaly limits;
• immutable or protected audit logs;
• offline or segregated backups; and
• tested incident-response and business-continuity plans.

Cybersecurity is a financial-management issue because a cyber incident can create direct cash loss, market exposure, regulatory penalties, operational interruption and reputational damage.

Data ethics

Treasury data may include employee information, supplier details, bank-account identifiers, credit information and commercially sensitive forecasts. Ethical data management requires:

• lawful and transparent collection;
• purpose limitation;
• data minimisation;
• accuracy and correction mechanisms;
• retention limits;
• controlled sharing;
• protection against unauthorised re-identification; and
• respect for individual rights under applicable privacy regimes.

Data should not be collected merely because storage is inexpensive. A firm must be able to explain why data is needed and how decisions based on it affect stakeholders.

7. ESG and Sustainable Finance in International Treasury

ESG analysis examines environmental, social and governance factors that may affect financial performance, stakeholder relationships and long-term legitimacy. Treasury is directly involved because it selects banks, arranges funding, manages liquidity and evaluates counterparties.

Environmental factors

Relevant environmental issues include greenhouse-gas emissions, energy use, water stress, biodiversity loss, pollution and exposure to physical climate hazards. Treasury may face these through:

• higher insurance and financing costs;
• disruption to production or logistics;
• declining collateral values;
• carbon taxes or regulatory changes;
• commodity-price volatility; and
• reduced access to capital for high-transition-risk activities.

Social and governance factors

Social factors include labour standards, human rights, community impact, health and safety, and supply-chain practices. Governance includes board oversight, anti-corruption controls, tax conduct, executive incentives, audit quality and cybersecurity.

A firm with weak governance may face higher counterparty risk even when its balance sheet appears strong. A bank’s conduct, sanctions controls and exposure to controversial activities may also matter when selecting financial partners.

Sustainable finance instruments

Green bonds finance eligible environmental projects. Sustainability-linked loans or bonds generally connect financing terms to specified sustainability performance targets. Transition finance supports credible movement from a high-emission business model toward a lower-emission one.

Treasury must avoid “greenwashing,” in which environmental claims are exaggerated or unsupported. Instruments should have clear use-of-proceeds rules or performance indicators, reliable reporting and appropriate external review where required. The financial benefit should not be separated from the credibility of the sustainability framework.

Climate-risk framework

Climate risk is commonly divided into:

• Physical risk: damage or disruption caused by floods, storms, heat, drought or sea-level rise.
• Transition risk: losses arising from policy, technology, market or consumer changes during movement toward a lower-carbon economy.
• Liability and reputational risk: legal claims or loss of trust connected with environmental harm or misleading claims.

A practical treasury process is:

1. map assets, suppliers, customers, banks and collateral by geography;
2. identify physical and transition exposures;
3. estimate effects on cash flows, funding access and asset values;
4. perform scenario analysis;
5. define mitigations, such as diversification, insurance, hedging or refinancing; and
6. monitor indicators and disclose material information.

Climate scenario analysis is not a precise prediction. It explores possible outcomes under specified assumptions. Its value lies in revealing vulnerabilities and testing strategic resilience.

8. Integrated Decision-Making: A Practical Framework

A future-oriented treasury can evaluate a digital or sustainable-finance initiative through seven questions:

1. Strategic fit: Does it support the organisation’s international operating model?
2. Economic value: Are expected benefits greater than implementation, control and transition costs?
3. Risk: What new market, liquidity, operational, cyber, legal and reputational risks arise?
4. Data: Are the data sources reliable, lawful and sufficiently representative?
5. Control: Can the activity be authorised, monitored, audited and reversed where necessary?
6. Sustainability: Does it reduce or manage material environmental and social risks?
7. Resilience: Can the firm continue if a provider, network, market or system becomes unavailable?

This framework prevents technology enthusiasm from replacing financial analysis. A payment system that is faster but impossible to audit may be unsuitable. A sustainable loan with weak performance indicators may create reputational exposure. An AI forecast with impressive average accuracy may still be unsafe if it fails during stressed conditions.

Applied Case: Meridian Components plc

This is an illustrative case.

Background

Meridian Components plc, headquartered in the United Kingdom, operates manufacturing subsidiaries in Mexico, Poland, Vietnam and South Africa. It receives customer payments in euros and US dollars, pays suppliers in several currencies and maintains multiple bank accounts. Its treasury department relies on spreadsheets, daily bank files and manual email approvals.

Problem

The company experiences three problems. First, cash forecasts are frequently inaccurate because subsidiaries submit inconsistent information. Second, a payment-diversion fraud results in a loss after an employee receives a convincing supplier-bank-change email. Third, a proposed sustainability-linked loan is criticised by investors because its emissions target is vague and difficult to verify.

Analysis

The board proposes a cloud TMS, API bank connectivity, AI-assisted forecasting, real-time payments and a sustainability-linked revolving credit facility. Treasury assesses the proposal using the integrated framework.

The TMS could centralise balances and establish approval workflows. APIs could reduce manual entry, but access tokens and provider concentration would require strong controls. AI could improve forecasts if data definitions were standardised first. Real-time payments could accelerate supplier settlement but would increase the need for beneficiary verification and transaction monitoring.

The sustainability-linked facility would be credible only if the performance target were measurable, material and independently verifiable. Meridian therefore links the margin adjustment to a defined reduction in verified emissions intensity, establishes a baseline year, specifies calculation boundaries and commits to annual assurance.

Approach and solution

Meridian implements the changes in stages:

1. It creates a common chart of treasury data and a formal data-owner structure.
2. It introduces multi-factor authentication, dual payment approval and callback verification for beneficiary changes.
3. It deploys a TMS with role-based access, audit logs and disaster-recovery testing.
4. It uses AI forecasts only as decision support, with treasury review and override documentation.
5. It limits real-time payments to approved corridors and applies value and velocity thresholds.
6. It conducts quarterly model validation and cyber-fraud simulations.
7. It renegotiates the sustainability-linked facility with independently assured indicators.

Outcome

The company obtains better daily visibility, reduces manual reconciliation and detects unusual payments more quickly. Forecast performance improves, but the treasury committee continues to use stress scenarios because the AI model performs less reliably during an unexpected commodity-price shock. The sustainability-linked facility gains credibility because its target, methodology and assurance process are transparent.

Learning points

Technology creates value only when combined with standardised data, clear accountability and resilient controls. AI should augment professional judgement rather than obscure it. Sustainable finance requires measurable commitments, not broad environmental language. Finally, real-time capability must be matched by real-time control.

Challenges and Limitations

Digital finance remains subject to several limitations. Interoperability between banks, platforms and jurisdictions is incomplete. Regulation differs across countries, particularly for digital assets, data transfers and AI. Cloud concentration may create systemic dependency on a small number of providers. Cybersecurity controls become obsolete as attack methods change. Smaller subsidiaries may lack the skills needed to manage advanced systems.

Sustainable-finance classifications can also differ across jurisdictions, making comparisons difficult. ESG ratings may disagree because providers use different definitions, weights and data. Climate models contain uncertainty, especially over long horizons. Firms should therefore avoid presenting scenarios as precise forecasts.

Finally, automation can create an illusion of objectivity. A model may reproduce historical discrimination or amplify poor assumptions while appearing mathematically sophisticated. Governance must remain a human and institutional responsibility.

Best-Practice Principles for the Future Global Treasury

1. Treat data as a controlled financial asset.
2. Integrate cybersecurity into treasury design from the beginning.
3. Use automation for repetitive, rule-based work while preserving human accountability for material decisions.
4. Maintain independent validation of important models and algorithms.
5. Test systems under liquidity, market, cyber and climate stress.
6. Establish clear ownership for ESG data and sustainability claims.
7. Diversify critical providers and maintain tested fallback arrangements.
8. Document assumptions, exceptions, approvals and overrides.
9. Review legal and regulatory requirements in every relevant jurisdiction.
10. Evaluate digital and sustainable-finance projects according to long-term resilience, not only short-term cost reduction.

Chapter Summary

Digital finance is transforming international treasury through fintech platforms, APIs, real-time payments, electronic markets, distributed ledgers, digital currencies, cloud TMSs and analytics. These technologies can improve speed, transparency, forecasting and operational efficiency, but they also create cyber, legal, operational and counterparty risks.

Blockchain and smart contracts may improve shared records and automate settlement, yet their usefulness depends on legal enforceability, data quality, interoperability and reliable external information. Digital currencies require analysis of issuer or central-bank liability, convertibility, privacy, regulation and settlement risk.

Treasury management systems provide integrated control over liquidity, payments, funding and financial risk. Analytics and AI can support forecasting, anomaly detection and decision-making, but responsible use requires explainability, validation, monitoring, human accountability and protection against bias and data misuse.

ESG and climate risk influence cash flows, funding access, collateral, counterparties and operational resilience. Physical and transition risks should be identified through mapping, scenario analysis and mitigation planning. Sustainable-finance instruments are credible only when their targets, methodologies and reporting are material, measurable and verifiable.

The future global treasury is therefore not simply faster or more automated. It is an integrated function in which technology, finance, cybersecurity, ethics, governance and sustainability are managed together.

Key Terms

Artificial intelligence: Technology that performs tasks requiring functions associated with human intelligence, such as prediction or classification.

API: A controlled software interface that allows systems to exchange data or instructions.

Blockchain: A distributed ledger that records transactions according to agreed technical and governance rules.

CBDC: A digital form of central-bank money.

Climate risk: Financial risk arising from physical climate effects or the transition to a lower-carbon economy.

Data minimisation: Collecting and retaining only data necessary for a defined purpose.

Distributed ledger: A shared database maintained across multiple network participants.

Embedded finance: Financial services integrated into a non-financial commercial platform.

ESG: Environmental, social and governance considerations relevant to organisational performance and responsibility.

Fintech: Technology-enabled financial products, services or infrastructure.

Greenwashing: Misleading or unsupported claims about environmental benefits.

Liquidity management: The process of ensuring sufficient funds are available when required.

Model risk: Risk arising from incorrect models, data, assumptions, implementation or use.

Open banking: Consent-based sharing of financial data or services through secure interfaces.

Real-time payment: A payment processed and made available within a very short period, often continuously.

Smart contract: Software that executes predefined logic on a digital platform.

Sustainability-linked finance: Financing whose terms are connected to specified sustainability performance.

Tokenisation: Digital representation of an asset, right or claim.

Treasury management system: An integrated system supporting treasury data, transactions, controls and reporting.

Transition risk: Financial risk caused by policy, technology, market or behavioural changes associated with decarbonisation.

Review Questions

Short Answer Questions

1. What are two benefits and two risks of real-time payments for multinational firms?
2. Distinguish between a blockchain and a central bank digital currency.
3. Why is model validation important in AI-assisted treasury forecasting?
4. Define physical climate risk and transition risk.
5. What is the purpose of segregation of duties in digital payment controls?

Long Answer Questions

1. Explain how a treasury management system can improve international liquidity management while creating new operational dependencies.
2. Evaluate the potential uses and limitations of blockchain and smart contracts in cross-border trade finance.
3. Discuss the governance requirements for responsible AI in global treasury.
4. Explain how ESG and climate considerations can affect funding, counterparties and financial risk.
5. Develop a cybersecurity and data-ethics policy for a multinational treasury using API connectivity and cloud systems.

Analytical / Application Questions

1. A multinational firm is considering real-time payments for all supplier transactions. Analyse the control changes required before implementation and recommend whether a phased rollout is appropriate.
2. A forecasting model has a low average error but performs poorly during currency crises. Explain why the model may still be unsuitable and propose a validation and stress-testing plan.
3. Design an evaluation framework for a sustainability-linked loan that combines financial value, emissions-target credibility, data assurance, regulatory compliance and reputational risk.

Multiple Choice Questions with Answers

Question 1

Which function is most directly associated with a treasury management system?

A. Product design and advertising

B. Liquidity, payments and financial-risk administration

C. Employee recruitment only

D. Production-line scheduling

Correct Answer: B. Liquidity, payments and financial-risk administration

Question 2

What is a principal risk of real-time payments?

A. Payments can always be recalled automatically

B. Fraudulent transactions may be completed before detection

C. They eliminate the need for authentication

D. They cannot operate outside banking hours

Correct Answer: B. Fraudulent transactions may be completed before detection

Question 3

A central bank digital currency is best described as:

A. A digital liability of a central bank

B. An unregulated corporate loyalty point

C. A physical banknote stored electronically

D. A mandatory cryptocurrency issued by every bank

Correct Answer: A. A digital liability of a central bank

Question 4

In the cash-flow identity C_t = C_t-1 + I_t − O_t + F_t, what does F_t represent?

A. Forecast error

B. Fixed assets

C. Net financing cash flow

D. Foreign-exchange volatility

Correct Answer: C. Net financing cash flow

Question 5

Which control best supports segregation of duties in payment processing?

A. One employee prepares, approves and releases every payment

B. Payment preparation and approval are assigned to different authorised users

C. All payment passwords are shared across the treasury team

D. Beneficiary changes are accepted without verification

Correct Answer: B. Payment preparation and approval are assigned to different authorised users

Question 6

What is the main purpose of model validation?

A. To guarantee that a model can never be wrong

B. To assess whether the model is suitable, accurate and properly implemented

C. To remove the need for human review

D. To increase the number of variables regardless of relevance

Correct Answer: B. To assess whether the model is suitable, accurate and properly implemented

Question 7

Which is an example of transition risk?

A. Flood damage to a factory

B. A cyclone interrupting port operations

C. A carbon regulation reducing the value of a high-emission asset

D. A server outage caused by a power failure

Correct Answer: C. A carbon regulation reducing the value of a high-emission asset

Question 8

What does data minimisation require?

A. Collecting all potentially useful data indefinitely

B. Collecting and retaining only data necessary for a defined purpose

C. Deleting all financial records immediately

D. Sharing data with every platform provider

Correct Answer: B. Collecting and retaining only data necessary for a defined purpose

Question 9

Which feature most improves the credibility of a sustainability-linked financing target?

A. A vague statement about becoming greener

B. An unverifiable target selected after the loan is issued

C. A measurable, material target supported by reliable reporting or assurance

D. Avoiding disclosure of the baseline year

Correct Answer: C. A measurable, material target supported by reliable reporting or assurance

Question 10

Why can an AI model with good average forecasting accuracy still be unsafe for treasury?

A. Average accuracy proves that stress performance is always strong

B. It may fail during unusual events such as crises or market shocks

C. Treasury models never use historical data

D. AI models cannot produce numerical forecasts

Correct Answer: B. It may fail during unusual events such as crises or market shocks