Tendencia: Análise Técnica de Forex
4 stars based on
The insurance activity is based on the statistical treatment of uncertainty. It therefore behoves all stakeholders to analyse the highest possible number of possible situations or results that might impinge on their current or future financial position and jeopardise their solvency and stability.
To that end, and following a methodology used by financial institutions, the application of stress tests in risk management might help to ascertain whether or not insurers have the financial wherewithal and operational flexibility for assuming losses that might arise from a host of more or less atypical scenarios.
In recent years, especially after the international financial crises and the appearance of Basel II, supervision authorities and financial institutions have been taking a growing interest in the question of measuring and monitoring vulnerabilities of the financial system, doing so by applying techniques that gauge the effects caused by changes in the main variables affecting this system.
These techniques, known as stress tests,were first used in the nineties of last century and have since become widespread in the finance world. Along the same lines the International Association of Insurance Supervisors, IAIS drew up a global supervision framework under the name Solvency II IAIS, whose aim is to keep insurance markets efficient, fair, firm and stable, thereby guaranteeing the protection of policyholders.
Solvency II updates the regulatory requirements for evaluating and supervising the financial situation of insurance companies and their internal mode of operation, taking its cue from the guidance papers drawn up by IAIS IAIS, It also requires insurers to undertake regular stress testing for a wide range of adverse scenarios to ascertain the adequacy of capital resources in case technical provisions have to be increased.
Among the principles referred to therein, stress tests are deemed to be especially relevant to the following ones:. The concept of stress testing takes in a set of mathematical and statistical techniques for quantifying the vulnerability of a financial or insurance institution to any significant change in the economic environment it operates in, with the aim of ascertaining potential losses that might affect the solvency thereof.
This article analyses how stress tests should be conducted as part of all-in insurer risk management and how these tests may help to establish and maintain the minimum capital and solvency requirements laid down by Solvency II.
Stress testing is a generic term normally used to describe a set of techniques applied in financial and insurance institutions, the aim of which is to ascertain their potential level of economic and financial vulnerability in the event of certain exceptional but plausible scenarios.
Stress tests help insurance companies to manage their risks and maintain enough financial resources to cover them, doing so by identifying and quantifying different complex scenarios based on expected future financial positions. These tests estimate the quantitative impact of adverse exceptional but plausible disturbances on those variables that affect the results and solvency of an insurer or group of insurance companies IAIS, Generally, in the insurance context, the term stress testing includes two types of analysis: The objective is to ensure that insurers have sufficient data for a better understanding of the vulnerabilities they face under unlikely but not impossible atypical conditions.
These stresses might be financial, operational, legal, liquidity-based or be related to any other risk that might have an adverse economic impact on the insurer. Scenario testing, for its part, quantifies the effect of a simultaneous move in diverse risk factors.
The scenarios may be based on significant market events occurring in the past or on estimates of the consequences of a an event or possible variations on market conditions that have not yet occurred.
The first type are called historical scenarios and the second, hypothetical scenarios KPMG, Historical scenarios reflect the changes in risk factors that have occurred at given moments of history. The simplest way of defining these scenarios is to identify specific periods of time days or months that were particularly extreme in terms of volatility or variability of risk factors and then observing their effects on insurers.
The main advantage of historical scenarios is that the risk-factor changes brought under the spotlight actually occurred rather then being arbitrarily chosen; this grants them a certain aura of trustworthiness on the premise that history usually repeats itself.
Furthermore, this method is very transparent because the past events and consequences under study and brought to bear on the present usually have well-known outcomes.
The drawback is precisely the fact that they are based on the assumption that past events will reoccur in the future and ipso facto that the risk factors will always behave the same; this is not necessarily so. Another disadvantage of this method is that it is hard to assess new products when by definition they could not possibly have been involved in the past effects and variations under study.
Hypothetical scenarios, on the other hand, are based on a set of shocks that are thought to be plausible but have not yet in fact occurred, thus covering the lacunae of the previous method. It should be borne in mind here, however, that hypothetical scenarios are always based on some hypothesis of historical behaviour. For example, any estimation of the contagion effects of future financial crises will be based on the quantified effects of similar crises or situations that have occurred in the past.
The particular usefulness of this type of methodology lies in its analysis of risk situations that, although they have not actually occurred, are plausible. This makes it possible to estimate their effects and gauge whether insurers could cope with them, even though such situations are different from those dealt with on a daily basis. The probable maximum loss PML determines the effect on the company of a combination of changes in the risk factors determining the market in which it operates, causing a loss in the assets possessed.
Conversely, extreme value theory uses a series of statistical techniques based on asymptomatic behaviour and distributions, stochastic processes and limit laws to identify and model extreme observations or outliers. Its object is to ascertain how extreme the greater or lesser magnitude of a random phenomenon might be, i. When designing specific stress tests for the insurance market, consideration needs to be given first of all to such factors as scope, type of analysis, risks analysed, variables subjected to perturbation, size of the perturbation and timeframe over which the effects of said perturbations are to be measured.
The immediate consequence of this decision-taking process is that each insurer designs its own stress tests in light of its own risk profile and the specific characteristics of its business. This is likely to lead to variation among insurers in terms of the extent and nature of the tests performed.
More specifically, stress-test design should consider those events deemed to be relevant in terms of their impact and likelihood. On the basis of the IAIS guidance paperstress tests should be designed with the following risks in mind: The concept of insurance risk is traditionally broken down into three risk categories: When drawing up stress tests to evaluate this type of risk, consideration should be given to the following risk-associated factors: Another component of the insurance risk is the catastrophe risk, defined as losses deriving from catastrophic or aggregated claims accumulations.
For stress-test purposes the assessment of this type of risk takes in the following aspects: Additionally, the liquidity risk implies consideration of the following: This risk may be covered through guarantees and other financial instruments, such as derivatives.
Credit risk is made up by the risk of omission or direct payment default, indirect credit, liquidation, sovereign risk, counterparty risk and retrocession or migration risk. The factors to be taken into account in stress tests to assess risks of this type are basically the following:.
From the insurance point of view, and in business terms, this risk can be defined as the risk deriving from changes in margins, in costs or turnover.
In operational terms it may derive from situations of fraud, errors, legal problems, etc. The factors allowing the insurer to assess operational risk within stress tests are the following: In any case, when designing stress tests, the insurer has to demonstrate at least that the operational risks have been considered and that there are suitable plans and procedures for dealing with them in any adverse situation.
The assessment of group risks involves the following factors: The design of stress tests, after selection of the risk factors to be analysed, should be a group effort across the whole firm, including risk managers, financial personnel, actuaries and business line managers. It should also take on board other points of view outside its normal operational sphere, such as insurance supervisors, external consultants, the accountant and actuarial professions, reinsurers and rating agencies.
As regards the frequency of stress tests, due consideration should be given to the time patterns of the effects produced by the abovementioned factors, but they should be conducted at least annually, reflecting the new characteristics of importance to the insurer and its portfolio trend. When market conditions are changing rapidly, supervisors may require quarterly stress testing but with fewer details than the yearly tests.
Furthermore, stress tests should examine the effects and impact that different time horizons will have on business plans, strategic risks and future operating requirements, so the time horizon needs to be long enough for the effects of the stress to be fully evident, for management to act and for the results to emerge.
In the case of certain risks this may call for stress testing over a complete economic cycle. Various modelling techniques are used in stress testing. In general, however, they are based on static or dynamic modelling and deterministic or stochastic approaches. A simple example of a static deterministic stress test is where an insurer, in determining its appropriate capital level, examines the effects of loss ratios on its balance sheet. The loss ratio is the risk variable, and the impact on net assets is the resultant exposure.
Such tests do not take into account the actual probabilities of the different loss ratios occurring. Stochastic models are more advanced techniques. They are based on probabilities that predict how key financial parameters interact with each other over time, and generate a distribution of outcomes based on simulations of those parameters in the future.
One of the advantages of stochastic modelling is that it provides an indication of the range and the likelihood of different financial outcomes.
This is useful in achieving a particular level of confidence in the solvency level, e. Stochastic models are useful, and at times essential, where the insurance contracts contain both embedded options and financial guarantees. An example of a stochastic risk measurement technique is Value at Risk VaR. VaR models, often used in banks, provide a probability-based boundary on likely losses for a specified holding period usually 10 days or 1 year and a given confidence level. More specifically,VaR was defined by Sharpe in However, statistical models such as VaR have a limited ability to accurately capture what happens in exceptional circumstances or extreme events, since statistical inference is imprecise without a sufficient number of observations and, in any event, is based on extrapolation of past experience into an unknown future.
It is also important to bear in mind here that risks are not completely dependant or independent, so it is crucial for the insurer to analyse such correlations as they might bear with each other, doing so with the aim of assessing the possible effects thereof on the stress tests and the working hypotheses used. In this case, it might be well worthwhile to apply in stochastic models the probability distributions deriving from copula theory.
Insurers should analyse these correlations with a certain regularity to modify the initial assumptions and keep them in line with changing circumstances. This is because experience has shown that, in adverse situations, the correlation levels, initially low, may rise sharply for example, if an insurer is affected by a large-scale catastrophe there might be a knock-on effect on other parties dependent on said insurer, such as reinsurers, intermediaries, other service providers and the counterparties of the capital market.
An example of such tail-dependency would be where there are two risks that are usually uncorrelated, but where an extreme event for one risk may lead to greater loss from the other risk than would otherwise have occurred under normal circumstances.
This situation is obvious when a catastrophe coincides with a stock market collapse exactly as occurred with the 11 September terrorist attack. Once the models to be used in the stress test have been defined, they should then be reviewed by individuals not engaged in the development or regular use thereof, nor involved in the corresponding business decisions, with the aim of gauging their precision.
Thus, when stochastic models are used, an insurer should stress-test the working assumptions to pinpoint errors and ascertain how sensitive the results are to the assumptions and the model parameters. There should also be a process in place for ongoing analysis, monitoring, identification, documentation and auditing of such changes as might occur in the models due to modifications in the initial situation and changes in modelled results from one period to the next.
One way of assessing the accuracy of a model used in the tests is back testing. The aim of this method is to show that actual model results, over a set period of time, fall within the expected range for said model. This is done by the reconstruction, based on historical records, of situations that occurred in the past, using the rules defined by a given strategy. This gives rise to a series of statistics such as net profit or loss, the period of time or time interval during which the test was conducted, the universe or set of assets or liabilities included in the analysis, the volatility measurements, the average or percentage gain or average loss, the exposure or percentage of invested capital or capital subject to risk in the insurance market, the loss ratios, annual return on capital and risk-adjusted return on capital which can then be used to gauge whether said strategy has been effective.
The underlying hypothesis of this test is that any strategy that worked well in the past is probably going to work well in the future, and vice versa. The trouble is that for many insurance-sector risks the validity of these checks are limited, for example when assessing situations involving legal changes or when analysing catastrophic events.
Stress tests take in a wide variety of tools for estimating the main vulnerabilities any insurance company faces in its daily activity.
Though fairly clear in theory, stress tests are more complicated in practice due to the difficulties of determining all the following: The complexity of the statistical techniques to be used and the accuracy of the estimates will depend on the quality of input information and the presence of any variables that might be difficult to measure.
The growing use of these tests, especially in the financial sector, could encourage risk managers to see stress tests as a way of mitigating the economic fallout of a disaster.
From this point of view stress tests could give a false sense of security to those using them, for the application of these tests currently depends on a structure of probabilities that makes them completely random.
It is particularly important to remember that stress tests are considered to be a way of verifying the VaR or random forecast models. When stress tests are incorporated into risk models, however, these models can then be validated by back testing to ascertain whether or not the model used is in keeping with a series of historical records. International Actuarial AssociationA global framework for insurance solvency assessment.
Research Report of the Insurance solvency assessment working party. Towards a common structure and common standards for the assessment of insurer solvency www.
International Association of Insurance SupervisorsInsurance core principles and methodology. Anales del Instituto de Actuarios.