OneNym.idDiscuss deployment
English

THE BOARDROOM FIREWALL / Chapter 3

Cyber Risk Is Just Business Risk

Cyber risk is often treated as mysterious and technical, which leads either to handing it entirely to specialists or to spending out of fear. Chapter 3 shows how to weigh it with the same tools leaders already use for floods, recalls and lost customers.

This page summarises the chapter. The full text is in the book.

Two questions: how likely, and how bad?

Every risk has two parts: how likely it is to happen and how bad it would be if it did. A small kitchen leak is likely but cheap; a major fire is rare but severe. Cyber risks sort the same way. A staff member clicking a suspicious link is almost certain at some point but usually contained by other defences, while a ransomware attack that stops production for two weeks is less likely but could cost millions. Asking both questions separately balances the security team's focus on worst cases against finance's focus on what has never happened before.

Precision is not on offer; nobody can say the chance of ransomware next year is exactly 7 per cent. The aim is to be roughly right rather than precisely wrong. A heat map, with likelihood on one axis and impact on the other, helps a leadership team see the whole picture on one page. At Meridian Foods, the security manager places eight scenarios on such a grid. Ransomware stopping production planning and the warehouse for more than three days, and payment fraud through fake supplier bank changes, sit in the red corner, and the CEO notices that the ransomware risk now looks bigger than the recall risk the board spent months discussing.

Risk appetite and tolerance are board decisions

No active company can eliminate risk, so the real question is how much risk to carry, and in pursuit of what. Risk appetite is the board's broad statement of attitude; risk tolerance is the specific, measurable line beyond which the company must act. Plain statements, such as refusing any risk that could stop production beyond a stated number of days, let teams make decisions without escalating every question. Appetite can also differ by type of harm: a food company may accept short website outages but have no appetite for anything that could affect food safety records.

Setting appetite is a board-level judgement, made with management's help, just as it is for financial or market risk. Meridian's board debates three draft statements for half an hour and changes a three-day limit on production stoppage to five days after the CFO explains what three would cost. That trade-off between the cost of protection and the level of comfort is the conversation the chapter wants boards to have.

Bring cyber into enterprise risk management

In many companies cyber risk lives on a separate IT register written in technical language, while the enterprise register carries a single line for it whose rating has not changed in years. As a result, cyber is not compared fairly with other risks, and owners such as the head of operations may not realise that a cyber attack is now the likeliest cause of a long production disruption. The fix is to write the main cyber scenarios in business terms, rate them on the same scale as other risks, give each a named executive owner and review them in the same committee.

At Meridian, ransomware is assigned to the operations executive, loyalty-data theft to the head of marketing, payment fraud to the CFO and the supplier scenario to the head of procurement. When the marketing head asks why she should own a cyber risk, the CEO explains that her decisions about what the loyalty scheme collects, shares and keeps shape the risk more than any technology. For specialists, the chapter introduces quantitative methods such as FAIR and suggests quantifying only the top three to five scenarios while keeping the heat map for wider communication.

Four choices, and no such thing as zero risk

Every risk decision is one, or a mix, of four options: accept, reduce, transfer or avoid. Accepting is legitimate when it is conscious, signed off by someone with authority and reviewed over time. Reducing covers most security spending. Transferring, mainly through insurance or contracts, moves money but not harm, and insurers increasingly expect risks to be reduced first. Avoiding, by stopping the activity that creates the risk, is often the most effective and most overlooked choice, because data you do not hold cannot be stolen.

In a single afternoon, Meridian's leadership chooses to reduce and transfer ransomware risk, formalises call-back checks and a second approver for payments above 50,000, stops collecting the year and day of customers' birth dates because marketing only uses the month, and formally accepts website defacement with a review date. The chapter stresses that zero risk cannot be bought. Residual risk always remains; risk appetite tells you when to stop spending, and boards that understand this judge incidents by the quality of the response rather than the mere fact that something happened.

Putting risk in money

A simple method: pick one well-defined scenario, estimate how often it might happen and what it would cost in a typical and a bad case, then combine the two. A scenario expected once every ten years that would cost 3 million represents about 300,000 of expected loss a year. That is not a forecast, but a way to compare risk with the cost of reducing it. Meridian's illustrative table shows payment-fraud controls costing almost nothing, data reduction being cheap and effective, and a ransomware investment of about 250,000 a year set against an expected loss of about 400,000.

Averages hide extremes, assumptions must be written down so they can be challenged, and some harms convert to money only roughly. Even so, a rough number with stated assumptions is far more useful than none. Meridian's next budget request is a short paper covering scenarios, likely costs, proposed actions and remaining risk, and the chair remarks that it was the first security discussion in years where she understood what she was voting for.

KEY POINTS

Key points

  • List your top cyber scenarios in business terms, with a likelihood and impact for each.
  • Ask the board to agree a short written cyber risk appetite.
  • Move the main cyber scenarios onto the enterprise risk register with named executive owners.
  • Record an explicit accept, reduce, transfer or avoid decision for each major risk.
  • Look for avoidance first: stop collecting data you do not need.
  • Express your largest cyber risk in money, with assumptions written down, and compare it with the cost of reducing it.

QUESTIONS

Common questions

How do you assess cyber risk in business terms?
Ask two questions about each scenario: how likely is it, and how bad would it be if it happened? Describe the scenario in business terms, such as ransomware stopping production planning for more than three days, and rate it on the same scale as other enterprise risks. Precision is not required; the aim is to be roughly right and able to compare risks.
Who should set a company's cyber risk appetite?
The board, with management's help. Management and the security leader can describe the risks and propose options, but deciding how much risk the company will carry in pursuit of its strategy is a board-level judgement, as it is for financial or market risk. Clear appetite statements let security teams make decisions without escalating every question.
What are the four ways to treat a cyber risk?
Accept, reduce, transfer or avoid. Accepting must be a conscious, signed-off and periodically reviewed decision; reducing covers most security spending; transferring through insurance or contracts moves money but not harm; and avoiding means stopping the activity that creates the risk. Avoidance, such as collecting less personal data, is often the cheapest and most overlooked option.
How do you calculate the expected loss from a cyber risk?
Estimate how often a defined scenario might occur and what it would cost when it does, then multiply the two. A scenario likely once every ten years that would cost 3 million represents about 300,000 of expected loss per year. This is not a forecast but a way to compare the risk with the cost of reducing it, and boards should look at the bad case as well as the average.
Should cyber risk be on the enterprise risk register?
Yes. The main cyber scenarios should be written in business language, rated on the same scale as other risks and owned by the executive whose part of the business would be harmed. Keeping cyber only on a separate technical IT register prevents fair comparison with other risks and hides its impact from other risk owners.