10 risks known from risk identification step (the blue bubbles)Risk criteria already known and agreed internally.Risk analysis is the process of deciding what the risk level rating is based on measuring likelihood and consequence and applying the risk matrix.After we have determined the answer, the risk and its level is recorded in the risk register.
12Risk analysis - what to consider 1.The way in which likelihoods, consequences and combined risk level are expressed should reflect the nature of the risk, the information available and the purpose of the risk assessment and importantly, be consistent with the risk criteria.2.The reliability of estimates should be considered and communicated to decision makers and other stakeholders. Factors such as divergence of expert opinion, uncertainty, availability, quality, quantity and ongoing relevance of information used should be clearly stated. Any critical assumptions, and the sensitivity of the risk analysis to these assumptions, should be disclosed.3.Risk analysis can be undertaken with varying degrees of detail, depending on the risk, the purpose of the analysis, and the information, data and resources available.4.Risk analysis must include the effect of existing controls (Risk Levels will be expressed for Residual and Inherent risk in the Risk Register). Controls and measuring their effect will be covered in Week 7.
13Risk analysis bias and error▪Risk analysis endeavours to provide an accurate and unbiased measure of the risk.▪Analysis can be qualitative, quantitative or a combination.▪Estimation error will always be present regardless of method. The level of estimation error, and critical assumptions made, should be communicated to decision-makers.▪When doing risk analysis, biases such as conservatism, optimism, perception biases and data/modelling errors should be minimised.▪Risk aversion is embedded in the Risk Matrix by how the risk levels are assigned to the pairwise combinations of likelihood and consequence. To avoid double-counting and inaccuracy, the estimates of risk levels should be risk-neutral and bias-neutral.
14Methods of risk analysis▪Qualitativeanalysis uses expert opinions and judgmentto assess the likelihood and consequence according to the descriptive scales contained in the risk criteria.▪Quantitativeanalysis uses data, theoretical models and numeric computationsto quantify numerically likelihood and consequence.▪Note, sometimes in practice a hybrid approach may be applied to quantitative estimates. eg. the risk estimate from a model predicting the likelihood of borrower default may be adjusted by a loans officer for factors not otherwise captured by the model, or it may include subjective inputs.
You've reached the end of your free preview.
Want to read all 59 pages?
- Three '16
- Economic Capital, risk manager, CFaR, risk principles