Balance horizons and risk appetite
6 min read
Lesson objectives
- Construct a mix preserving cash and specialist reserves under category specific risk appetite.
Please sign in to use all the tools
Once you sign in, the reading tools are yours: larger or smaller text, typeface, line spacing, focus mode, highlighting and bold, and notes saved to your account and gathered in My study.
An organisation can prefer stability while still running experiments. Ambition for growth does not authorise disruption of its core service. Risk appetite describes the types and extent of exposure an organisation is willing to take to achieve its objectives. Risk capacity asks what loss or interruption it can actually absorb without losing its essential function. Willingness can exceed the limits imposed by cash, reputation or team capability. Begin the portfolio discussion with an objective and distinguish willingness, capacity and binding constraints. The UK Orange Book places appetite within board leadership, planning and decisions. Labels such as cautious or ambitious do not carry a fixed numerical meaning that can simply be imported into any organisation.
Turn a broad statement into decision conditions. We welcome innovation does not tell a portfolio manager whether an untested service may be released to every user. A fictional organisation could say: we accept the cost of bounded experiments to learn about a new service model, while protecting the existing core service from interruption. Translate that into an experimental funding cap, participant group, rollback threshold and permissible test period. A specified tolerance can then identify the deviation from desired performance that triggers action, such as an agreed failure rate. Define how the terms are used locally. Differences in terminology between frameworks should not become an unstated change in what teams believe they have permission to do.
Appetite can differ by risk category. An organisation may accept commercial uncertainty in a new service while remaining highly cautious about safety or confidential information. A single medium rating conceals that difference. Describe the exposure profile: loss of experiment funding, service interruption, poor adoption and dependence on one supplier. Identify who can accept each exposure and which treatments must precede acceptance. A reversible experiment can reduce the consequence of failure without increasing the chance of success. A recovery plan therefore does not prove that customers want the product. It prevents a learning exercise from becoming an irreversible commitment. A large forecast benefit must not erase a binding restriction that the board has no authority to override.
Time horizon adds another dimension. Preservation initiatives maintain an existing capability. Improvement initiatives enhance relatively understood performance. Exploration tests a more distant opportunity. There is no universally correct allocation percentage. The board chooses a mix using organisational condition, obligations and resources, making explicit what is sacrificed when funds move between categories. Reserving money for exploration is ineffective if operations consume all analytical expertise. Check both money and people, including the time operating units need to absorb change. Delaying an experiment by two months while its opportunity remains available may be preferable to running it alongside a core transition that needs the same specialists. An allocation is meaningful only when it can be executed.
Test the mix using scenarios that describe connected changes, as well as individual sensitivities. If a shared supplier slips by two months while funding income falls, can the organisation maintain its minimum service? Can experiments be stopped before they consume cash reserved for recovery? A scenario is not a forecast. It exposes a choice that works only when every assumption goes well. Distinguish a real reserve from a number labelled contingency. Uncommitted cash may be available quickly; scarce expertise cannot necessarily be summoned at short notice. Funds already locked into noncancellable contracts should not be described as freely available flexibility. Agree the response order before disruption so that negotiation does not have to start from nothing when pressure arrives.
Expected monetary value alone cannot justify crossing an exposure limit. A ten percent chance of losing six million gives an expected loss of 0.6 million. The organisation may nevertheless be unable to survive the six million loss if it occurs. The average does not change the plausible consequence or guarantee service continuity. Review the probability estimate, its source and sensitivity, then examine capacity and ways of limiting exposure. Do not add the failure probabilities of unrelated projects and call the total the probability of portfolio failure. That calculation needs a defined event and relationships between events. Interdependence will be analysed later; the leadership task here is to reject a comfortable number that conceals a threat to the organisation's essential function.
Lesson summary
Balance requires usable reserves and scenarios, not spending percentages or average exposure alone.