Next: Introducing market risk, Previous: User guide, Up: User guide [Contents][Index]
Risk management is both art and science. The science part can be seen as quantitative risk measurement, selection of stochastic models and pricing functions, as well as setting up scenario generation and aggregation. Risk measurement follows a straigthforward approach and is only constraint by operational questions, e.g. requirement and test management, computational limitations, reporting timelines and maintenance effort. Calculation automation and calculation performance are two of the key design principles of OCTARISK. The dependencies between the modules and configuration capabilities are covered in See Developer guide description of all classes and functions.
The art part covers more the qualitative aspects of risk management, e.g. how can risk assessment contribute to reaching investment goals, which types of risk should be selected, which risk types can be neglected (known unknowns), where are blind spots (unknown unknowns) and what is the remaining model risk. Risk management can be both highly specific to the investment business (e.g. internal models) as well as more or less standardized (e.g. Solvency II Standard Model) to allow for easy regulatory comparison between companies. In the following some ideas on individual risk management are given and a road map for running an appropriate risk model is outlined.
Before one should think about risk management, a clear investment objective or goal has to be specified. These goals could be (early) retirement, buying a house in some years, repaying debt or even just accumulating as much wealth as possible. Even unspecific goals are better than no goals at all. Once the goal and responsibilities are set, the following aspects should be covered in an written investment policy statement (IPS):
These points are valid both for individual persons as well as for institutions or trusts. Only after thourough specification of an IPS it makes sense to think about risk management.
In the light of the IPS, risk management is a steering tool for reaching the investment goals. Ideally, risk management is part of the continuous feedback loop between risk assessment, interpretation and analysis and adjusting the asset allocation or adjusting the investment goal. Possible purposes could be risk assessment of speculative investments, e.g. equity option short selling. Here risk management serves as input for limit setting and will directly trigger investment actions (e.g. derisking or allowing for further short selling). Another purpose is asset-liability management (ALM), where the asset allocation should closely match the risk profile of the liabilities to reach immunization (e.g. against inflation and interest rate risk for a retirement portfolio). The most basic purpose could be yearly risk assessment of aggregated risk figures of the asset portfolio (e.g. VaR or expected shortfall) and comparing to the risk appetite set in the IPS. Depending on the outcome, adjustments in asset allocation could be triggered. If no liability positions are involved, the time period between risk assessment could be significantly lowered towards once a year.
Once the purpose of risk management was clarified, all possible risk types should be evaluated and the most specific for the portfolio should be selected. For a pure long only equity portfolio based on index funds there is no need to cover interest rate or inflation risk, while concentration risk (counterparty risk, issues risk) could be worth analyzing. For an ALM portfolio interest rate and inflation risk will be the most prominent risk types. Assessing more risk types leads to increased maintenance effort (e.g. calibration) and complexity to analyse risk contributions.
The number is risk types covered is both dependent on the underlying asset and liability risk profile and only limited by available maintenance effort. If risks diversify away or have only minimal impact, they could be left out.
New risks could emerge at any point in time (e.g. political, climate or country specific risks). Maybe it is worth analysing global risk and wealth reports (e.g. from McKinsey or World Economic Forum), which most often include assessment of severity and probability of different risks.
Full internal model calibration based on historic time series or personal expert judgement vs. taking expert judgement (e.g. standard deviation and correlation) from Solvency II Standard Formula. Open points to considers are availablility and quality of historic time series of risk factors and quality and update frequency of expert judgement. Even commercial time series are limited in quality for more exotic risk factors and require special treatment..
Risk management is requiring a holistic view or the whole personal situation beyond investing in the stock market. One should think about the worst possible scenario which would lead to undesired outcome (reverse stress testing), e.g. liquidty needs rise dramatically (relative persons require help during crisis, one is stuck in foreigh country during Covid-19 crisis and cannot fly back), stock, commodity and bitcoin markets fall dramatically in weeks (contagion effect - everything devaluates), and internet banking is not reachable due to lock downs and power outages. Do you have enough money, gold and food to live appropriately? Assessing these kind of scenarios does not require to take precautionary measures - but thinking about it can help to comfortably an dactively decide to do nothing against it.
Proper presentation of results can be helpful to educate clients or family members and provide arguments for taking / not taking certain actions and increases condfidence. OCTARISK assists in this part with providing high quality graphics.
One should specify certain points in time when to review the approach (e.g. on a yearly basis after consulting the newest global risk report).
A full working example of a long only equity, real estate, gold and bond portfolio incl. savings and pension plans can be found under WORKING_FOLDER. The calibration was performed with shocks and correlations according to Solvency II Standard Formula. The reporting output contains LaTeX tables with key figures and graphics highlighting the risk profile of the portfolio.
The following steps could be performed to determine a proper risk model and to periodically reassess the assumptions:
| Step for each Portfolio | Long Only Investment | Asset-Liability Retirement | Speculative Stock Option Trading |
|---|---|---|---|
| Risk types to cover | Depending on assets, e.g. Equiy, FX and Real Estate | EQ, RE, IR, Inflation | Option valuation requires EQ and EQ Vol |
| Risk figures to assess | Stress tests could be enough, VaR measures could further assist | VaR / ES measures are recommended to capture complex IR and Inflation dependencies | VaR / ES and stress tests incl. qualitative considerations about liquidity, concentration and counterparty risk would be required |
| Calibration | If only stress testing (e.g. EQ -50%) is required, a spread sheet calculation would be enough. No calibration requirements. | Risk factor shocks calibration (both standard deviation and correlation) could be taken from Solvency II Standard Formula (e.g. even taking symmetric equity adjustment into account), in addition inflation risk shall be calibrated by historic inflation expectation curves. Correlations set by expert judgement. | Full internal model calibration is required, to select risk factors for equity and equity vol. A full through-the-cycle calibration of shock sizes and correlation should be performed on historic time series (e.g. starting before the financial crisis and covering Covid-19 scenario). Moreover severe historic stresses should be evaluated (e.g. Flash Crashs, Black Friday, 9/11 etc.) |
| Inverse stress testing: specify an upper barrier of loss where you would freak out and think about possible scenarios | Which stock market shock would lead to this really hurtful loss? | Which inflation expectation, political decisions of quantitative easing would lead to this loss? | Increasing margins, liquidity drain for personal reasongs, stock market loss and equity volatility spike: which combination could lead to this loss? |
| Other risks | Concentration risk of the portfolio (all money at one internet bank - fraud / ransom / IT bugs)? | Political risks (tax changes, inflation target changes), Country and political risks, local risks of e.g. own real estate required during retirement, health and longevity risks, government pension related risks | Counterparty risk (stock exchange or issuer of warrants), margin calls, issues risk of broker, and model risk (e.g. pricing models of options adequate to capture real world risk profile?) |
| Further risk indicators | Dependencies of counterparty (Swap party), issuer (of ETF), country (where custodian bank is) etc. to determine concentration risks, ESG score | Dividend / coupon payments of portfolio (e.g. to cover monthly expenses) | Qualitative assessment of calibration result and historic time series appropriateness (e.g. hourly trading vs. daily time series) |
Next: Introducing market risk, Previous: User guide, Up: User guide [Contents][Index]