Risk Management, Threat Modelling & Business Continuity
CISSP CISSP · Security and Risk Management
What risk actually is
Risk is the possibility of loss, and the CISSP defines it with a simple relationship: risk arises when a threat exploits a vulnerability in an asset, producing an impact. A threat is any potential danger — a hacker, a flood, a careless employee. A vulnerability is a weakness the threat can exploit — an unpatched server, a door left unlocked, a process with no backup. An asset is anything of value: data, systems, people, reputation. Risk management is the disciplined process of identifying these elements, assessing the risk they create, and deciding what to do about it. The exam expects you to think in this structure: every risk question can be broken down into threat, vulnerability, asset and impact, and the correct answer is usually the one that addresses the right element.
Risk management follows a cycle: identify risks, assess them, treat them, and monitor the results. Treatment has four options, and the exam tests them constantly: mitigate (reduce the risk with controls), accept (acknowledge the risk and live with it, formally documented and approved at the right level), transfer (shift the financial impact to someone else, usually through insurance or outsourcing — note that transferring the financial impact does not transfer accountability), and avoid (eliminate the risk by stopping the activity that creates it, such as decommissioning a vulnerable legacy system). A fifth option sometimes appears: deter, which discourages the threat rather than removing the risk. The choice depends on the risk level and the cost of the control: a control should never cost more than the loss it prevents, and accepting a low risk is often the right business decision.
Quantitative risk analysis: the numbers
Quantitative analysis expresses risk in money, and the exam tests the formulas directly — these are among the few calculations you must be able to do. The chain is: Asset Value (AV) is what the asset is worth. Exposure Factor (EF) is the percentage of the asset's value lost in a single incident. Single Loss Expectancy (SLE) = AV × EF. Annualized Rate of Occurrence (ARO) is how often the incident is expected per year (a once-in-ten-years event has an ARO of 0.1). Annualized Loss Expectancy (ALE) = SLE × ARO. ALE is the number that drives decisions: if a control costs less than the ALE it eliminates, it is worth buying; if it costs more, it is not. A related concept is total risk versus residual risk: total risk is the risk before controls, and residual risk is what remains after controls are applied. Residual risk can never be zero, and the organisation must formally accept it.
| Formula | Meaning |
|---|---|
| SLE = AV × EF | Expected loss from one incident |
| ALE = SLE × ARO | Expected loss per year |
| Control value = ALE before − ALE after − control cost | Whether a control is worth buying |
Worked example: a database server worth $200,000 would lose 40% of its value in a flood (EF = 0.4), so SLE = $80,000. Floods of that severity occur once every 20 years (ARO = 0.05), so ALE = $4,000 per year. A $50,000 flood barrier is not justified on ALE alone; a $2,000-per-year insurance premium might be. The exam also tests the limits of quantitative analysis: it requires data that often does not exist (how often will a specific attack succeed?), and it struggles with intangible losses such as reputation damage. That is why most organisations use qualitative or hybrid methods in practice.
Qualitative risk analysis
Qualitative analysis uses judgement rather than numbers: risks are rated on scales such as low/medium/high or 1–5 for likelihood and impact, often plotted on a risk matrix. It is faster, cheaper, and works when data is scarce, but it is subjective and harder to use for cost-benefit decisions. Common qualitative techniques include brainstorming and interviews, the Delphi method (anonymous expert estimates refined over rounds until consensus emerges), and structured risk matrices. The exam often asks which method suits a situation: quantitative when you need a cost-benefit justification for an expensive control; qualitative when data is unavailable or the organisation needs a fast, broad assessment. A hybrid approach — qualitative screening followed by quantitative analysis of the highest risks — is common in mature programs.
Threat modelling
Threat modelling is the structured process of identifying what can go wrong with a system, and it is most valuable early — during design, when fixes are cheap. The best-known method is Microsoft's STRIDE, which categorises threats as Spoofing, Tampering, Repudiation, Information disclosure, Denial of service, and Elevation of privilege. Each STRIDE category maps to a security property it violates: spoofing violates authentication, tampering violates integrity, repudiation violates non-repudiation, information disclosure violates confidentiality, denial of service violates availability, and elevation of privilege violates authorisation. Other methods include PASTA (a risk-centric, seven-stage process), attack trees (diagramming the paths an attacker could take to reach a goal), and DREAD (rating threats by Damage, Reproducibility, Exploitability, Affected users and Discoverability). The exam tests STRIDE most often, and the key insight is that threat modelling is a design activity — doing it after deployment is better than never, but far more expensive.
Business impact analysis and continuity requirements
Domain 1 covers the requirements side of business continuity (the planning and recovery side sits in Domain 7). The Business Impact Analysis (BIA) is the process of identifying critical business functions, the resources they depend on, and the impact if they are disrupted. The BIA produces the numbers that drive all continuity planning: Maximum Tolerable Downtime (MTD) — the longest a function can be unavailable before the damage is unacceptable; Recovery Time Objective (RTO) — the target time to restore a function, which must be less than the MTD; Recovery Point Objective (RPO) — the maximum acceptable data loss measured in time (an RPO of one hour means backups at least hourly); and Work Recovery Time (WRT) — the time needed to test and verify systems after recovery. The relationship RTO + WRT must be less than MTD is a classic exam point. The BIA also identifies dependencies: a function cannot recover until its inputs — power, network, suppliers, people — are available, and the recovery sequence must respect that order.
Supply chain risk management
Modern organisations depend on suppliers, and every supplier is a potential attack path — the exam has increased its emphasis on this area. Supply chain risk management means assessing the security of vendors and service providers before engaging them (through questionnaires, audits, and certifications such as SOC 2 reports or ISO 27001), writing security requirements into contracts (including the right to audit, breach notification timelines, and data handling obligations), and monitoring suppliers continuously rather than once. Hardware and software supply chains carry additional risks: counterfeit components, compromised software updates, and malicious code inserted upstream. The exam's framing is governance: supply chain risk is a risk management problem, not a purchasing problem, and the correct answer usually involves assessment, contractual controls and monitoring rather than blind trust or total avoidance.
Risk registers, appetite and reporting
A risk register is the organisation's record of identified risks, their likelihood and impact, their owners, and their treatment status. Each risk has an owner — the person accountable for managing it — and that owner should be a business leader, not the security team, because risk decisions are business decisions. Risk appetite is how much risk the organisation is willing to accept in pursuit of its objectives; risk tolerance is the acceptable variation around that appetite for a specific risk. Senior management sets the appetite, and the security program operates within it. The exam tests the distinction between appetite (strategic, set by leadership) and tolerance (operational, applied to individual risks), and expects you to know that exceeding tolerance requires escalation, not silent acceptance.
Exam mindset for risk
Risk questions reward the answer that follows the process: identify, assess, then treat — in that order. If a scenario jumps to buying a control before the risk has been assessed, that is usually the wrong answer. Learn the ALE formulas cold; they are the most reliable calculation questions on the exam. And remember that accepting risk is a legitimate treatment — the correct answer is not always "implement a control".
Exam traps to watch for
- Buying a control before assessing the risk — the process is identify, assess, treat, in that order. A control chosen before assessment is a guess.
- Confusing SLE and ALE: SLE is one incident, ALE is per year. The exam often gives you one and asks for the other.
- Thinking risk transfer removes accountability: insurance or outsourcing moves the financial impact, not the responsibility.
- Forgetting that RTO + WRT must be less than MTD — a recovery plan that restores systems after the maximum tolerable downtime has already passed is a failed plan.
Practice questions
Test yourself on this guide. Answers are checked instantly, with a full explanation for each question.
Q1.A web server valued at $100,000 would lose 60% of its value if destroyed by fire. Fires of this severity occur on average once every 25 years at this location. What is the Annualized Loss Expectancy (ALE)?
Explanation: SLE = AV × EF = $100,000 × 0.60 = $60,000. ARO = 1/25 = 0.04 per year. ALE = SLE × ARO = $60,000 × 0.04 = $2,400 per year. The $60,000 figure is the SLE (one incident), not the ALE. The $40,000 figure uses the wrong exposure factor (40% instead of 60%), and $1,500 comes from using an ARO of 0.025 (once in 40 years) instead of 0.04.
Exam objective: CISSP Domain 1: Understand and apply risk management concepts (quantitative analysis)
Q2.A security team has identified a vulnerability in a legacy application. No patch exists, the application is critical to operations, and replacing it would cost $500,000. The assessed ALE for exploitation is $8,000 per year. What is the MOST appropriate risk treatment?
Explanation: Mitigation with compensating controls is the balanced answer: the application cannot be decommissioned (it is critical) or patched (no patch exists), so the risk must be reduced through other controls. Decommissioning would avoid the risk but destroy a critical business function — an unacceptable trade. Transfer through insurance moves financial impact but does not reduce the likelihood of exploitation or the operational damage. Accepting the risk with no action ignores that cheap compensating controls can reduce it substantially; acceptance is for risks that cannot be reduced cost-effectively, not for risks nobody has tried to reduce.
Exam objective: CISSP Domain 1: Understand and apply risk management concepts (risk treatment)
Q3.During a threat modelling exercise using STRIDE, the team identifies that an attacker could modify transaction records in transit between a web server and a database. Which STRIDE category does this threat belong to?
Explanation: Tampering is the STRIDE category for unauthorised modification of data, whether in transit or at rest — it violates integrity. Spoofing is impersonating another identity, which is not what is described. Repudiation concerns the inability to prove an action occurred, which is a different threat. Information disclosure is unauthorised reading of data; the scenario describes modification, not disclosure.
Exam objective: CISSP Domain 1: Understand and apply threat modelling concepts and methodologies
Q4.A business impact analysis determines that the order-processing system has a Maximum Tolerable Downtime (MTD) of 24 hours, a Recovery Time Objective (RTO) of 12 hours, and a Work Recovery Time (WRT) of 16 hours. What does this indicate?
Explanation: The total time to restore and verify the function is RTO + WRT = 12 + 16 = 28 hours, which exceeds the 24-hour MTD — the business would suffer unacceptable damage before the system is fully back. An RTO below the MTD is necessary but not sufficient; the verification time must also fit within the MTD. The RPO is a separate measure (data loss), and its absence does not make this particular arithmetic wrong. WRT is measured from the point systems are restored, not from the start of the outage.
Exam objective: CISSP Domain 1: Identify, analyse, and prioritise Business Continuity (BC) requirements
Q5.An organisation is evaluating a new cloud service provider that will host customer data. Before signing the contract, what is the MOST important security step?
Explanation: Supply chain risk management starts before the contract is signed: assess the provider's security (questionnaires, audits, certifications) and bind the requirements contractually, including audit rights and breach notification timelines. Price negotiation is a commercial concern, not a security step. Insurance transfers financial impact but does not verify the provider's controls. A pilot tests functionality, not the provider's security posture, and does nothing about contractual protections.
Exam objective: CISSP Domain 1: Understand and apply supply chain risk management concepts
Q6.A risk assessment team has no historical loss data and limited time, but needs to prioritise risks across 50 business units. Which risk analysis approach is MOST appropriate?
Explanation: Qualitative analysis is designed for exactly this situation: it uses structured expert judgement and likelihood/impact matrices to rank risks when hard data is unavailable, and it scales quickly across many units. Quantitative analysis requires loss data and frequency estimates that do not exist here, so its precision would be false. Monte Carlo simulation is a quantitative technique with the same data problem, plus far more effort. Refusing to assess at all leaves the organisation with no prioritisation whatsoever.
Exam objective: CISSP Domain 1: Understand and apply risk management concepts (qualitative vs quantitative)
Q7.A company’s board has stated it will accept up to $250,000 of annual loss from fraud in pursuit of faster payment processing, but no single fraud event may exceed $50,000. Which statement BEST describes these two figures?
Explanation: Risk appetite is the broad, strategic amount of risk the organisation is willing to accept in pursuit of its objectives ($250,000 per year); risk tolerance is the acceptable variation for a specific risk or event ($50,000 per incident). Reversing them confuses the strategic with the operational. Residual risk is what remains after controls are applied, which is a different concept from either figure.
Exam objective: CISSP Domain 1: Understand and apply risk management concepts (risk appetite and tolerance)
Q8.A penetration test discovers that an organisation’s public website leaks customer data through an API. The vulnerability has existed for two years. The CISO’s FIRST action should be to:
Explanation: The first priority is to understand the scope and stop the bleeding: determine what data was exposed and contain the vulnerability so the leak stops. Notifying customers before knowing what was exposed produces inaccurate notifications and may violate breach-notification rules that require specific facts. Blaming an individual does nothing to protect customers. A full risk assessment is valuable afterwards, but it is slower than containment and does not stop the active leak.
Exam objective: CISSP Domain 1: Understand and apply risk management concepts