Risk assessment is one of the most important steps economic analysis because in order to manage risk, it must first be identified, analyzed and evaluated.
Risk analysis - this is the application of a system of special knowledge for the study of economic phenomena and processes in conditions of uncertainty and conflict in order to obtain qualitative and quantitative information necessary for making managerial decisions.
The purpose of risk analysis is getting necessary information for making managerial decisions to the extent possible to predict and appropriately protect the company from the consequences of risky events.
Usually risk analysis is carried out in the following sequence:
1) definition of internal and external factors, increasing or decreasing the degree a certain kind risk;
2) analysis of the identified factors;
3) assessment of a certain type of risk;
4) establishment of an acceptable degree of risk;
5) analysis of individual operations in relation to the selected degree of risk;
6) development of measures to reduce the degree of risk.
For risk analysis, quantitative and qualitative methods are used. Qualitative Risk Analysis involves identifying sources and causes of risk in processes and works, identifying areas and types of risk, identifying practical benefits and possible negative consequences that may arise in the process of implementing projects (works, processes) that contain risk. Most scientists involved in the problem of risk assessment state that qualitative analysis is the most difficult step general analysis degree of risk.
Qualitative risk assessment is the process of identifying and identifying risks that require a quick response. This risk assessment determines the importance of the risk and chooses how to respond. The availability of information that accompanies them helps to prioritize the different risk categories. A qualitative risk assessment is an assessment of the conditions for the occurrence of risks and the determination of their impact on an object using standard methods and means. The main task of a qualitative assessment is to identify possible types of risks, as well as factors affecting the level of risks in the implementation of a certain type of activity. At this stage importance has an identification of all possible circumstances and a detailed description of all possible risks.
The most risk analysis framework provides:
Identification of individual types of risk;
Choice of risk analysis technique;
Establishment of risk factors and their significance;
Creation of a risk action mechanism model;
Establishing the relationship between individual risks and the cumulative effect of their action;
Formation of a common portfolio of economic risks.
As a result of the analysis, it is necessary to obtain a clear understanding of all possible risks, potential risk areas have been identified, and Negative consequences or additional benefits that may arise from the implementation of a particular solution.
Qualitative risk assessment of an enterprise is carried out in the following areas:
The nature of the business;
External environment;
quality management;
Nature of activity;
Stability of work, stability;
The financial condition of the organization, etc.
As a rule, a qualitative risk analysis is carried out in two stages. The first stage involves comparing the expected positive results(income) of choosing a specific direction entrepreneurial activity With possible consequences(losses, damages), among which are: financial, material, temporary, social, implementation, environmental and moral and psychological. they are compared with the expected results that the enterprise will receive in the development of a certain area of activity.
At the second stage, the influence of decisions taken by the managers of the enterprise at the stage of developing a strategy, the interests and behavior of other business entities are determined, since they do not function in isolation, but are elements of the market. During this stage of the analysis, those subjects are also identified who will benefit from the occurrence of a certain type of risk.
For a qualitative risk assessment most commonly used in practice expert methods, based on a subjective assessment of expected performance parameters. The most common method used in the analysis process is method of expert assessments, the essence of which is to obtain the necessary information about the risks that may threaten the organization's activities, based on the processing of the opinions of experienced professionals and experts.
It is advisable to use this method when solving problems that cannot be formalized, when the incompleteness and unreliability of information do not allow the use of accurate methods for assessing the level of risk.
Level of risk - this is an assessment of the ratio of the scale of expected losses to the value of the property of the enterprise, as well as the likelihood of these losses. For instance, financial condition Enterprises can be assessed by experts with a division into high, medium and low levels of risk in the context of the following components:
Using Credits:
a. the level of risk is high if the enterprise cannot carry out current activities without the use of borrowed funds;
b. the level of risk is medium, if the company needs investment loans for development, business expansion;
v. a low level of risk, when the company does not take loans or uses them rarely.
Level of own working capital:
a. the risk is high if there are problems (deficit) of own working capital;
b. the risk is medium if the equity ratio working capital equal to the standard;
v. the risk is low if the ratio of own working capital is higher than the normative one.
Liquidity of assets:
a. a high level of risk is associated with large excess stocks of raw materials and materials, finished products in stock, overdue receivables.
Probability of bankruptcy: a. high;
b. exists; v. low.
Profitability:
a. low profitability compared to the industry average means high risk.
Level of receivables:
a.60% of current assets, turnover period of more than 180 days - high risk;
6.40-60% of current assets, turnover period 30-60 days - medium risk;
v. less than 40% of current assets, turnover period less than 30 days - low.
Financial investments of the enterprise:
a. if the share of financial investments in assets is large, and the return on assets is lower than the profitability of the main activity, respectively, the enterprise was engaged in risky speculation, did not meet the expectations of high profitability.
The main goals of using expert assessments in qualitative risk analysis are:
Forecasting the course of development of an event or phenomenon today and in the future. According to risk analysis and assessment, this is identifying the sources and causes of risk, predicting the actions of competitors, identifying all possible risks, assessing the likelihood of risk events occurring, assigning coefficients of relative importance and ranking risks, identifying ways to reduce risk, etc.;
Drawing up action scenarios;
Formation of a complete set and qualitative assessment of options using various ways risk reduction, or a combination thereof, and the like.
The advantages of this method are the speed of obtaining information for timely management decisions and relatively low costs. The disadvantage is relatively high level subjectivity and, as a result, the lack of confidence in the reliability of the estimates obtained.
So, the main results of a qualitative risk analysis are: identification of specific risks and their causes, analysis and cost equivalent of the hypothetical consequences of the possible realization of certain risks, proposals for measures to minimize damage, and, finally, their cost estimate.
It should be noted that a qualitative risk analysis also implies its quantitative result, that is, the analysis process should include not only a description of specific types, identification possible causes their occurrence, analysis of the expected consequences of their implementation and proposals for minimizing the identified risks, but also a valuation of possible losses and all measures to minimize the identified risks.
Increasing the reliability of expert assessments requires appropriate procedures for selecting experts according to many criteria and quantitative methods for processing their conclusions.
The results of the analysis serve as important input for the implementation quantitative analysis risk, which provides for the numerical determination of individual risks, as well as the total risk of the enterprise.
The need to identify and highlight significant risk factors, improve the efficiency of the management process, the existence of the possibility of choosing a specific economic decision in the aggregate alternatives makes it necessary to supplement the analysis with a quantitative one.
The purpose of quantitative analysis is getting numeric expression individual risks with the definition of the characteristic of probability and possible losses. The quantitative value of the risk level is often defined as a certain function of the product of the indicators of the consequences of a risky situation and the probability of its occurrence. To do this, a set of scenarios is formed and for certain types of risk, distribution functions for the probability of losses occurring can be constructed depending on their size.
Quantitative Methods provide for risk assessment in absolute and relative terms. In absolute terms, the risk is measured by the frequency or size of possible losses in monetary terms.
In relative terms, the risk is measured by various dimensionless indicators, is the ratio of two or more indicators.
Frequency the occurrence of a certain level of losses is determined by the formula:
where R - the frequency of occurrence of a certain level of losses; P - the number of occurrences of a particular level of losses; pzag - the total number of cases in a statistical sample that covers everything, including accounts and successful transactions.
Determining the degree of risk in absolute values it is advisable to use relative to the characteristics of certain types of losses, and relative - when comparing the predicted level of losses with the real, industry average, average for the market segment, etc.
By doing quantitative analysis risk, the most common and universal are such methods:
Statistical method;
Method of analogies;
Decision tree method;
Analytical method.
Let's consider the methods in more detail.
Statistical Method is based on the study of statistics of losses and profits that have taken place in a given or similar enterprise in order to determine the likelihood of an event, to establish the magnitude of the risk. Probability refers to the possibility of obtaining a certain result.
The main task of statistical risk assessment methods is to determine the probability of occurrence of a single adverse event based on statistical research available data on the activities of a specific risk object (organization) in the past. In the simplest case quantitative risks activities are evaluated using the indicators of dispersion, standard deviation, coefficient of variation.
In absolute terms, the degree (degree) of risk (the degree of expected failure to achieve the goal) can be defined as the product of the probability of failure (unwanted consequences) by the magnitude of these undesirable consequences (losses, payments, etc.):
where I AM, - the magnitude of the risk;
B, - amount of losses (losses)
G., - probability of undesirable risks.
Probability (R) calculated with a sufficient degree of accuracy on the basis of statistical data. To evaluate the actual data, all probabilities are assumed to be the same and are determined as follows:
The average expected value is related to the uncertainty of the situation, it is expressed as a weighted average of all possible outcomes V, where is the probability of each outcome (R) used as the frequency or weight of the corresponding value (V).
Then the expected, most probable amount of losses (losses, incomes, profits) will be:
where P - number of cases;
In and - the amount of losses (losses, incomes, profits) in / "-th case; re - the probability of occurrence of /"-th case.
The average expected value is a generalized quantitative characteristic and does not allow making a decision in favor of any option. For the final decision, it is necessary to measure the fluctuations of indicators, that is, to determine the degree of fluctuation possible outcome. The fluctuation of the possible outcome is the degree of deviation of the expected value from medium size. To determine it, the variance or standard deviation is usually calculated.
Dispersion (<т) - это взвешенное среднее из квадратов отклонений действительных результатов от среднего значения:
The mean deviation is calculated when conducting statistical tests of various hypotheses, as well as to identify relationships between random variables. This statistic is the most common type of variance used in calculating risk scores.
Average deviation is determined by the formula:
The coefficient of variation (d) is the ratio of the standard deviation of income and the relative value of expected income (expenses):
The coefficient of variation allows you to compare the fluctuations of signs that have different units of measurement. The higher the coefficient of variation, the stronger the fluctuations of the signs. Based on the value of the coefficient of variation, you can use the following scale to assess the risk:
0.0-0.1 - minimal risk;
0.1-0.25 - low risk;
0.25-0.50 - acceptable risk;
0.50-0.75 - critical risk;
0.75-1.0 - catastrophic risk.
Statistical methods for quantitative risk assessment are considered to be one of the most common. Their advantages include the simplicity of mathematical calculations, and their disadvantages include the need for a large number of observations (the larger the data array, the more reliable the risk assessment).
Method of expert assessments. This method involves the analysis by experts of two groups of factors - quantitative and qualitative. Expert assessments are a logical presentation and conclusions of specialists regarding a particular economic phenomenon or process. Unlike the statistical method, this method can be used in conditions of scarcity and even complete absence of information. This is its significant advantage over other methods.
Another advantage of the method of expert assessments lies in the possibility of using the experience of experts in the process of analyzing situations and taking into account the influence of various qualitative factors. Formally, the procedure for peer review most often consists of the following. The management of the organization develops a list of evaluation criteria in the form of expert (questionnaire) sheets. For each criterion, appropriate weighting coefficients are assigned, not notified to the experts. Further, for each criterion, answer options are composed, the weights of which are also unknown to the experts. Experts, conducting the analysis, analyze the object of study and mark the selected answer. The completed expert sheets are processed accordingly (using statistical (computer) information processing packages, and the result (or results) of the examination is obtained.
In practice, both individual and group expert assessments are used.
The advantage of individual expertise is the speed of obtaining information for decision-making and relatively low costs. The disadvantage is the high level of subjectivity, and, accordingly, the decrease in the reliability of the obtained expert assessments.
Panel reviews tend to be less subjective and decisions based on them are more likely to be implemented. It is believed that in the case of solving a problem under conditions of uncertainty, the opinion of a group of experts is more justified than one expert.
To ensure conditions favorable for the formation of objective opinions by experts, it is desirable to adhere to the following principles:
Independence of the formation by experts of their own judgments about the object of study;
Convenience of working with the questionnaire (questions must be formulated in generally accepted terms, exclude any semantic ambiguity, etc.);
Logical correspondence of questions to the structure of the object of study;
Acceptable time spent on answering the questionnaire, convenient time for assessment;
Maintaining the anonymity of responses;
Providing experts with all the necessary information.
Depending on the specifics of the expert survey, the object of study and the methodology used for expert data, expert assessments may have a different scale of measurement.
When diagnosing the collected expert data, in accordance with the purpose of the study and the adopted models, it is necessary to submit the information received from the experts in a form convenient for making a management decision (arrange objects, indicators, factors, etc.), and also determine the consistency of actions experts and the reliability of expert assessments. So, for example, the risks identified during the diagnostic process must be presented in order of their importance (degree of impact), risk reduction options - in the order of their preference. The most common ordering methods are ranking, direct evaluation, sequential comparison, pairwise comparison.
To assess the consistency of experts' opinions, the concordance coefficient is most often used, the value of which allows one to draw a conclusion about the reliability of the estimates.
Concordance coefficient is determined by the formula:
where <Уф - the actual variance of the final (ordered) estimates given by the experts;
° max - variance of the final (ordered) estimates, provided that the opinions of experts completely coincide;
a¡ - the total score obtained by the G "th object;
T - the number of objects under study;
P - the number of experts.
The value of the concordance coefficient varies from 0 to 1. If the value of the coefficient is zero, there is no connection between the assessments of various experts, that is, there is no agreement of opinions. If the value is equal to one, then there is complete agreement between the opinions of experts. For a simplified assessment, it is customary to consider the opinions of experts as agreed, if] ¥> 0.5, and well matched if th ¥> 0.7 .
Intuitive characteristics, based on the knowledge and experience of an expert, in most cases give fairly accurate estimates. Expert methods allow you to quickly and without large time and labor costs to obtain the information necessary for the development and adoption of a management decision.
The effectiveness of the application and the reliability of risk diagnosis using the method of expert assessments largely depends on the competence and number of selected experts, the quality of the factors (criteria) of the accuracy and unambiguity of the wording. These circumstances often limit the wide application of this method.
Systems of estimating coefficients and, if necessary, scales of weights for these coefficients;
Scales for assessing the values of the obtained indicators;
In the process of developing a rating system, the problem of choosing a benchmark for comparison arises (comparison with a conditional benchmark enterprise is quite possible, but the choice of such a benchmark requires clarification for each type of risk). It is not necessary to choose the best among a large number of enterprises, it is more convenient to simply select the best ones from the list of indicators of enterprises and form a standard for comparison from them.
When using this method, the rating scale must be pre-formed and consist of a minimum of ranking values, as a rule, on a "good" or "bad" scale. It is obvious that such a ranking system of the rating method of risk assessment does not allow to determine its degree with sufficient accuracy. However, this method has found wide application in practical risk assessment for the following reasons. Firstly, this method does not provide for the analysis of large data arrays. Secondly, the application of this method involves the parallel ranking of the result obtained on a certain scale. Thirdly, the rating method does not require the user to have special mathematical training, but only skills in elementary financial calculations. To improve the adequacy of this method, in some cases it is advisable to use corrective parameters, various weightings, as well as recalculation taking into account the riskiness of certain financial transactions when calculating coefficients.
Quite effective in the rating assessment is the presence of certain reference values for comparison, it involves the calculation of the rating by the method of distances or differences between real and reference values. When determining the final rating, some value is ranked, and distances.
However, studies show that this method does not make it possible to take into account all the subtleties of a particular situation and often leads to erroneous results, which is explained by the limits of the estimated coefficients, the possibility of their application only in specific conditions, the moment of action of unaccounted factors, and the possibility of choosing the wrong standard.
analogy method. The essence of the analogy method lies in the analysis of all available data on objects that have a high degree of similarity with the estimated one. This is done in order to calculate the probabilities of the occurrence of losses. The method uses a database of similar objects to identify common dependencies and transfer them to the object under study.
The analogy method finds the greatest application in assessing the risk of frequently recurring projects, for example, in construction, the insurance business, etc.
This method is used when all other risk assessment methods are unacceptable. Sources of information can be very diverse: statistical and accounting reports, published reports of partner enterprises and competitors, information from government agencies, and the like.
The feasibility of using the analogy method lies in the fact that it can be used in case of identifying the degree of risk of new areas of business activity, when there is no statistical information.
The disadvantages of the method include the failure to take into account the time factor in risk assessment and the need for only complete and reliable information. In the case of using the analogy method, it is always necessary to correct the results of the risk assessment for the degree of similarity of phenomena or processes.
Decision tree method. A "decision tree" is a schematic representation of the problem of making a managerial decision, it has the form of a graph, the vertices of which represent certain states in which the need arises for choice, and the branches of the tree represent various events (decision, consequences, operations) that can take place in each individual condition. Each branch of the "tree" is assigned individual numerical characteristics. As branches, for example, the amount of payment (financial flow) and the probability of its implementation, characterizing the level of its risk, are considered.
A decision tree enables a manager to evaluate various activities, relate financial results to them, adjust according to their likelihood, and then compare alternatives. In the process of preparing a decision, various options are considered that can be accepted, as well as, for each option, situations that may occur. In almost all cases of decision-making, the leader evaluates the likelihood or possibility of an event. The probability ranges from 1 if the event will occur, to 0 if it will definitely not occur.
The method is based on the assumption that the financial flows of each branch of the "tree" are not correlated with each other.The central concept of the decision tree method is to determine the expected value for each of the alternatives or options, which is the sum of the possible values multiplied by their probabilities.
In the analysis, based on determining the probability of a positive result on each "branch", actions are selected with the one with the best characteristic, that is, the most positive expected value.
This method is useful in the following cases:
There are a limited number of alternatives or strategy options available to choose from with certain probabilities of their occurrence;
The results of the decision taken depend on which alternative is chosen and what events actually take place.
The disadvantages of the method include the possibility of choosing the wrong scenario for the development of events, the need for complete and reliable information on each of the alternatives.
Analytical method. The analytical method of risk assessment is a specific combination of statistical evaluation and the principles of expert analysis. Usually, it is carried out in several stages.
At the first stage, preparation for analytical processing of information is carried out, it includes:
1) determination of the key parameter in relation to which the assessment of a particular area of business activity is carried out (for example, sales volume, profit margin, profitability, etc.);
2) the selection of factors that affect the activities of the organization, and therefore on the key parameter (for example, the inflation rate, political stability, the degree of implementation of contracts, etc.);
3) calculation of the values of the key parameter at all stages of the production process (development, introduction into production, production, liquidation of this area of activity).
The sequences of expenses and receipts formed in this way make it possible to determine not only the overall economic efficiency of the studied area of activity, but also to reveal its significance at each of the stages.
At the second stage, diagrams of the dependence of the selected resulting indicators on the value of the initial parameters are constructed. Comparing the obtained diagrams with each other, it is possible to single out those main indicators that have the greatest impact on a given type (or group of types) of entrepreneurial activity.
At the third stage, critical values of key parameters are determined. In this case, the break-even point can be calculated, showing the minimum allowable sales volume to cover the expenses of the enterprise, the minimum rate or mass of profit, and the like.
During the fourth stage, based on the obtained critical values of key parameters and factors influencing them, possible ways to improve the efficiency and stability of the enterprise, and, consequently, ways to reduce the degree of risk, are analyzed.
The advantage of the analytical method is that it combines both the possibility of a factor-by-factor analysis of the parameters that affect the risk, and the identification of possible ways to reduce its degree by influencing them.
Analytical methods also include sensitivity analysis, the method of adjusting discount rates based on risk, the method of equivalents, the method of scenarios, and others.
Sensitivity analysis is reduced to the study of the dependence of some initial indicator on the variation of the values of the indicators involved in its determination. Using this method provides answers to the following questions:
To what extent can the value of one or more input quantities deviate from the given values, provided that the initial indicator does not go beyond the permissible limits?
How much will the value of the effective indicator change for a given deviation of one or more input values from their predetermined values?
Discount rate adjustment method taking into account the risk is the simplest and therefore the most used in practice. Its main idea is to adjust some basic discount rate, which is considered risk-free or minimally acceptable. The adjustment is carried out by increasing the amount of risk reward.
Via method of reliable equivalents adjustment of the expected values of the flow of payments is carried out by introducing special reduction factors in order to bring the expected receipts into the amounts of payments, the receipt of which is practically beyond doubt, and the value of which can be reliably determined.
Scenario Method allows you to combine the study of the sensitivity of the resulting indicator with the analysis of probabilistic estimates of its deviations. Using this method, you can get a fairly clear picture for various variants of events. It is a logical development of the sensitivity analysis method, since it takes into account the simultaneous change in several factors.
Since each of the considered methods is not without drawbacks, it is necessary to use several different methods in practice. Of course, the results obtained by different methods will differ, but an analysis of the differences between them will allow us to identify factors that are taken into account in some methods and not taken into account in others, affect the accuracy of the assessment and the reliability of the results obtained. An analysis of the differences in the results, in comparison with the risk factors taken into account, will make it possible to identify existing trends in the development of future events in terms of the risk of certain types of activities, and this will make it possible to more accurately predict the degree of risk of achieving the planned results.
Methods of expert assessments, in many cases plays a decisive role in making important management decisions.
Methods for obtaining quantitative expert estimates
1. Direct quantification
Direct quantitative assessment is used both in the case when it is necessary to determine the value of the indicator measured quantitatively, and in the case when it is necessary to assess the degree of comparative preference of various objects.
In the first case, each of the experts directly indicates the value of the indicator for the assessed object. This may be a specific numerical value of the indicator for the object being assessed, for example: the cost of a residential apartment, etc.
If an expert finds it difficult to indicate a specific value of an indicator, he can indicate the range in which the value of the indicator being evaluated lies.
In the second case, when the comparative preference of objects is assessed by one or another indicator, the quantitative assessment indicated by the expert determines the degree of their comparative preference.
It is necessary to agree in advance that, say, a higher value of the estimate corresponds to a more preferable alternative. Sometimes it is more expedient to quantify the relative preference of objects in points, using specially designed point scales.
2. Midpoint method
This method can be used in the expert evaluation of the numerical values of indicators of a quantitative nature.
3. Method Churchman - Akof
The Churchman-Akof method is used to quantify the comparative preference of alternative options and allows for adjustment of estimates given by experts.
If a, is less preferable than the sum of the remaining alternatives, then it is compared with the sum of the remaining alternatives, except for the last one.
If alternative variant a at some step turned out to be preferable to the sum of other alternative variant and this ratio is confirmed for estimates, then ax is excluded from further considerations.
This process continues until all alternatives have been sequentially reviewed.
In practical application, in the case of a sufficiently large number of compared alternatives, some adjustments can be made to the method to reduce its complexity.
So, for example, the sum of the largest number of alternative options can be immediately determined with the rejection of less preferred options, which is less than Yes, etc.
4. Lottery Method
According to this method, for any triple of alternatives ax, a2, a3, ordered in descending order of preference, the expert indicates such a probability p, in which the alternative ar is equivalent to a lottery, in which the alternative ax occurs with probability p, and the alternative ab - with probability 1 - p.
Methods for obtaining qualitative expert assessments
Sometimes the specifics of the objects of expert evaluation are such that experts find it difficult to give quantitative estimates of the values of the evaluated indicators or the object as a whole, and
in some cases, such assessments are simply unjustified and do not allow obtaining sufficiently reliable expert information.
In these cases, it is often much more justified to use methods of qualitative assessment of objects of expertise.
There are also situations when the nature of expert information is such that quantitative assessments in the usual sense are practically impossible.
1. Expert classification
This method is advisable to use when it is necessary to determine whether the alternative options being evaluated belong to established and accepted classes, categories, levels, varieties, etc. (hereinafter referred to as classes).
It can also be used when the specific classes to which the evaluated objects should be assigned are not predetermined. The number of classes into which the evaluated objects are divided may not be predetermined.
It can only be established after the completion of the classification procedure.
2. Method of paired comparisons
The paired comparison method is one of the most common methods for assessing the relative preference of alternatives.
In the practice of using the method of paired comparisons, one often encounters inconsistency and even inconsistency in expert assessments.
In these cases, it is necessary to conduct a special analysis of the results of the examination.
We also note that with a sufficiently large number of alternative options being evaluated, the procedure for pairwise comparison of all their possible pairs becomes laborious for an expert. In this case, it is advisable to use the appropriate modifications of the method of paired comparisons.
3. Ranking of alternatives
A fairly common procedure is also the direct ranking by the expert according to the preference of the evaluated alternatives.
In this method, the expert is presented with alternative options selected for comparative evaluation, but preferably no more than 20-30 for their ordering by preference.
If there are more alternative options, then it is advisable to use the appropriate modifications of the ranking method.
4. Method of preference vectors
This method is more often used when it is necessary to obtain a collective expert ranking. The expert is presented with the entire set of evaluated alternatives and is asked for each alternative to indicate how many, in his opinion, other alternatives are superior to this one.
This information is represented as a vector, the first component of which is the number of alternatives that are superior to the first, the second component is the number of alternatives that are superior to the second, and so on.
The preference vector method is relatively labor-intensive and can be used taking into account the nature of the expertise.
This method can also be applied in the case when the expert has difficulties in using other methods for assessing the comparative preference of alternative options.
In collective expertise conducted using the preference vector method, it is advisable to calculate the resulting collective ranking, which reflects the collective point of view of all experts.
5. Discrete expert curves
If the goal is to develop forecasts or analyze the dynamics of changes in indicators characterizing the object of development and adoption of a managerial decision, then it is advisable to use discrete expert curves.
When constructing a discrete expert curve, a set of characteristic points is determined at which a change in the trend of changing the values of the indicator from the parameter under consideration is observed or expected, as well as the value of the indicator at the characteristic points.
In the areas between the characteristic points, it is assumed that the values of the indicator change linearly, i.e., two adjacent characteristic points of the curve can be connected by straight line segments.
If there are good enough reasons to determine the non-linear changes in the values of the indicator on the sections of the curve between adjacent characteristic points, it makes sense to move from discrete expert curves to expert curves.
When constructing expert curves, segments of straight lines can be replaced by segments of non-linear curves or curves constructed directly by experts.
Note, however, that we do not always have information that allows us to reliably judge the behavior of the curve in the sections between the characteristic points.
In addition, processing the results of expert assessments, and in particular the determination of the resulting collective expert assessment, is more reliable for discrete expert curves.
The use of expert curves makes it possible to more clearly and reliably present various scenarios for the development of a situation, which is often necessary when developing forecasts.
Expert curves can be effectively used both in the analysis of a decision-making situation and directly in the development and adoption of managerial decisions.
Expert assessments and measurement methods.
An expert's assessment is a quantitative and/or qualitative measurement of the relevant indicator.
Methods for obtaining quantitative expert estimates
Direct Quantification
The expert indicates a specific numerical value or range of the desired parameter.
midpoint method.
It is used with a large number of alternative options.
First, the most (A) and least (I) preferred alternatives are selected. Next - an intermediate option - M, the assessment of which divides the segment A - Z in half.
Methods for obtaining qualitative expert assessments
Expert classification (determination of belonging to a group, variety, category - tea variety)
Ranking of alternatives - ordering of compared alternatives according to the degree of preference for a certain feature.
The method of expert curves is used to obtain forecasts based on the analysis of the dynamics of indicators characterizing the object of the experiment (graph, extrapolation). To prevent errors in extrapolation, based on the generalized opinion of experts, points are determined on the graph at which a change in the development trend of the indicator should be expected.
The tools of the above methods are qualitative and quantitative scales.
19. Methods for evaluating the qualities of experts and the formation of expert commissions.
court method.
The procedure is similar to a court session. Some of the experts are accusing, some are defending, and some are the jury, rendering the final verdict. Expert assessments are subjective, so the key issue is the selection of experts.
Evaluation of the qualities of an expert.
Expert- a specialist, a professional, whose assessments and judgments the decision maker considers it useful to take into account in the decision-making process.
The formation of the composition of the expert commission depends on:
specific decision-making situation;
Opportunities for the organizers of the examination to attract highly qualified specialists for work;
Opportunities for specialists to participate in the work of the expert commission.
There is no generally accepted unified methodology for assessing the qualities of an expert.
You can rely on:
Professional knowledge;
Experience and efficiency in the composition of expert commissions.
Methods for evaluating the qualities of an expert:
A priori– do not use information about the results of the expert's participation in previous examinations (self-assessment, mutual assessment, expert list method, documentary (questionnaire) method).
a posteriori- use information about the results of the expert's participation in previous examinations (the method of paired comparisons of objects that are most preferable for this expert, the method of deviation from the resulting group assessment).
Test- involves a special test.
The test must be specially designed for specific objects of peer review
The true values of the estimated parameters (correct answers) should not be known to the expert
A scale should be developed to determine the accuracy of the assessments given by the expert.
The probability of randomly guessing the correct estimate should be very small.
The use of test methods makes it possible to evaluate such an important professional quality of an expert as the reproducibility of expert assessments. Conduct a series of similar tests and evaluate the results. The stability of the assessments indicates the professional competence of the expert.
Any changes in the conditions or results of the organization (industrial enterprise, trading company or bank), associated with a particular management decision alternative, can only be determined when there are corresponding indicators that can be measured. But we are talking not only about quantitative measurements in the usual sense for us, but also about qualitative assessments that make it possible to judge the dynamics of ongoing changes. Therefore, speaking of experts' assessments, we understand them as quantitative or qualitative measurements of the relevant indicators.
Depending on the objectives of the examination, experts can assess the value of real estate, the expected level of inflation, changes in the exchange rate, the reliability of the issuer of the security, the rating of the bank, etc. The objects of assessment and the estimated indicators are very diverse. If at the same time the expert must measure the value of a quantitative indicator (give a quantitative estimate), he can indicate a specific numerical value or an interval in which it lies. If the expert must determine the value of an indicator that can only be assessed qualitatively (give a qualitative assessment), then he can use verbal-numerical scales for this.
Thus, expert information can be both quantitative and qualitative. Consider the most common methods for measuring it separately for each of these two cases.
Methods for obtaining quantitative expert estimates
Direct quantification. The expert directly indicates the numerical value of the indicator for the assessed object (alternative), for example, the cost of an apartment or land plot, the estimated market capacity, etc. If it is difficult to indicate a specific value of an indicator, the expert can determine the range in which this value is located.
midpoint method. This method is used when there are many alternatives to be evaluated. At the beginning, the expert quantifies the most preferable alternative - f(а1) and the least preferable - f(а2). Next, he must choose the third alternative a3, the estimate of which f(a3) is located in the middle between the values of f(a1) and f(a2) and is equal to half of their sum. Then the expert determines the estimate of alternative a4, the value of which - f(a4) - is located in the middle between the values of f(a1) and f(a3), and alternatives a5 - f(a5), the value of which is the average value between f(a3) and f (a4), and so on, until all compared alternatives have been evaluated.
Methods for obtaining qualitative expert assessments
Method of expert classification. This method is appropriate to use when it is necessary to determine whether the alternatives being evaluated belong to established and accepted classes, categories, levels, varieties, etc., for example, assessment of tea varieties by tasters or determining the category of a hotel. If the classes are set in advance, then the expert is sequentially presented with alternative options and he determines to which class this alternative belongs. If the classes are not established, then the expert is presented with alternative options in pairs and asked to determine whether they belong to the same class or to different ones. After that, the alternatives are offered for evaluation sequentially, and it turns out whether each of them can be attributed to one of the resulting classes, or whether a new class should be formed for it. The procedure is repeated until all alternatives are evaluated (classified).
Method of ranking alternatives. The essence of this method is to arrange the compared alternatives according to the degree of preference for a certain feature. If there are relatively few such alternatives (no more than 20), then the expert is presented with the entire set of alternative options, and he indicates the most preferable among them. Then the expert determines the most preferred option from the remaining ones. And so on, until all evaluated alternatives are ranked. If there are more options to be evaluated, then first they are divided into groups ordered by preference using the expert classification method, and then ranked within each group.
Method of expert curves. This method is used when the purpose of the expertise is to develop forecasts based on an analysis of the dynamics of changes in indicators characterizing the object of expertise, for example, GDP growth rates, inflation, unemployment, etc. The use of this method involves the construction, based on the available statistical data, of a graph that displays the trend in the dynamics of the analyzed indicator. Then, with the help of extrapolation, this trend continues into the future, which makes it possible to assess the value of this indicator in the forecast period. However, the use of extrapolation can lead to significant errors, because does not take into account the possibility of qualitative changes leading to a change in trend. Therefore, further, based on the generalized opinion of experts, the points are determined on the graph at which a change in the trend of the estimated indicator should be expected, which makes it possible to visualize possible alternative scenarios for the development of the situation.
Two types of scales correspond to the two groups of methods for obtaining expert assessments considered above:
scales of quantitative measurements;
scales of qualitative measurements.
quantitative scales
Absolute scale. If it is required to determine the number of seats in the audience or the number of workers required to perform a certain operation, then the number resulting from such measurements is uniquely determined and is unique. This is the feature of the absolute scale.
Relationship scale. Assumes the ability to make measurements in various systems of units. So, for example, if we evaluate the mass of an object in kilograms, we get one numerical value, in pounds - another, and so on. However, no matter what system of measurements is used, the ratio of the masses of any two objects will be the same and does not change when moving from one system to another. The same is observed when measuring the length of objects. These are typical examples of a ratio scale in which the ratios of numerical scores remain the same.
Interval scale. In the scale of intervals, when passing from one system of measurements to another, the ratios of the numerical values of the evaluated alternatives are not preserved, but the ratio of the differences in the numerical assessments is preserved. An example of measurements in the interval scale is the determination of temperature, which, as you know, can be measured on different scales: Celsius, Fahrenheit, etc. Each time we will get different numbers, because. different scales use different reference points and different measurement units. But these different numerical estimates correspond to the same temperature of the object.
difference scale. If only the reference point changes during the transition from one numerical system of measurements to another, such scales are called scales of differences. An example of quantitative estimates in the scale of differences is the systems of chronology, which differ in the starting point (from the birth of Christ, from the creation of the world, or from the resettlement of the prophet Mohammed from Mecca to Medina).
Qualitative measurements are much less rigorous, they are subjective and other types of scales are used for them, the main ones are the following.
Qualitative scales
Nominal scale (name scale). The essence of the measurements of the evaluated alternatives in this scale is their division into classes according to a certain attribute. All alternatives that fall into the same class must correspond to the same number. Examples of measurements in nominal scales can be the division of students of one course into groups, the division of products according to quality levels (grades), etc.
Ordinal scale. This scale is used for ranking, i.e. determining the order of preference for alternatives, the severity of a particular property, for example, importance, urgency, etc.
Quite often, during the examination, it becomes necessary to measure indicators, the assessments of which are obviously subjective. So, for example, the knowledge and experience of specialists are used to assess the degree of risk in the implementation of an investment project, the competitiveness of a certain type of product, the level of competence of a candidate for a vacant position, etc. In these cases, a special type of ordinal scales is used - verbal-numerical scales. Their peculiarity lies precisely in the fact that they allow you to measure the degree of intensity of any property that has a subjective character. The composition of the verbal-numerical scale includes a meaningful (verbal) description of the selected gradations and their corresponding (gradations) numerical values.
The Harrington scale, widely used in practice, shown in Table. 6.1.
Table 6.1.
Harrington scale
The numerical values given in the table were obtained on the basis of a statistical analysis of a large amount of data, due to which the Harrington scale has a universal application and can be used in appropriate modifications (for example, in the form of a score scale) to assess various qualitative indicators.
Summing up the study of expert methods, it should be noted that their significance lies in the fact that they enhance the element of collegiality in the process of making complex decisions and, using intuition and collective generation of ideas, allow finding new, original solutions to problems that cannot be reached with the help of only logical reasoning.
At the same time, it is quite difficult to establish clear boundaries between all the considered methods of developing and justifying decisions (methods of modeling and expert methods). the solution of the complex problems of modern management also requires the complex use of various logical, statistical, mathematical and heuristic techniques. Therefore, it is not just one, but the prevailing group of methods that forms this or that method. The areas of application of decision-making methods depend mainly on the nature of the problems being solved and the conditions for decision-making, which is reflected in Fig. 6.1.
conclusions
The essence of expert decision-making methods is to receive answers from specialists to questions posed before them, to process the information received using special logical and mathematical procedures and to transform it into a form convenient for choosing the most preferred solution alternative.
Among the methods of expert assessments (which differ in the procedure for organizing the examination), the main ones are the commission method, the brainstorming method, the Delphi method and the scenario development method.
In the process of forming an expert commission, the problem of assessing the qualities of an expert arises - the need to take into account his professional knowledge, experience and work results in previous examinations.
Methods for assessing the qualities of an expert are divided into three groups: a priori (not taking into account the effectiveness of his participation in previous examinations); a posteriori (based on the real results of the participation of a specialist in the examination) and test (involving a special test for a potential expert).
Expert information can be both quantitative and qualitative. To obtain quantitative estimates, the main methods are the direct quantification method and the midpoint method. To obtain qualitative estimates, expert classification, ranking of alternatives and the method of expert curves are used. These two groups of methods for obtaining expert assessments also correspond to two types of scales: scales of quantitative measurements and scales of qualitative measurements.
Questions for self-examination
What does the concept of a priori expert evaluation methods include?
In what cases is it customary to use a posteriori methods of estimation?
What are the conditions for applying test methods for assessing the qualities of an expert?
What is the difference between quantitative and qualitative expert assessments?
What are the criteria for constructing qualitative scales?
Quantitative assessments, such as the business and organizational qualities of an employee, are usually made with the help of expert assessments. At the same time, to characterize a candidate for a position, 6-7 criteria are first established (taking into account the specifics of production and working conditions). For instance:
1. the ability to organize and plan work;
2. professional competence;
3. consciousness of responsibility for the work performed;
4. contact and sociability;
5. ability to innovate;
6. diligence and efficiency.
For each of these criteria, based on a study of the activities of candidates for the position, an appropriate assessment is given on a chosen, for example, five-point scale (excellent - 5; good - 4; satisfactory - 3; not satisfactory - 2; bad - 1).
Criteria scores are usually arranged in ascending numerical order. For example, when assessing according to the criterion "the ability to organize and plan work":
"1" - clearly unorganized worker and leader;
"2" - does not know how to organize and plan his work and the work of his subordinates;
"3" - knows how to organize the labor process, but does not always successfully plan work;
"4" - knows how to organize and plan well his work and the work of his subordinates;
"5" - is able to create and maintain a clear order in work based on effective planning.
In terms of their importance in the overall assessment of a candidate for a particular position, certain qualities always have a different share, which is established by expert means. For example, certain values can be accepted for the six criteria above.
To determine the overall assessment of the business and organizational qualities of a candidate for a managerial position, a special evaluation sheet is drawn up.
Naturally, the higher the overall score for each group of qualities, the more worthy the candidate is to fill a position in the management apparatus. The highest possible score is 5 and the lowest is 1.
The education, work experience and age of an employee must be taken into account when assessing business qualities. The fact is that education is one of the main qualitative characteristics in determining the level of qualification of an employee, work experience is a quantitative measure of experience, and age is interconnected with work experience.
Data on the candidate's education, his length of service and age are taken into account by calculating the coefficient of professional prospects according to the formula:
K \u003d Oy.arr. (1 + C/4 + B/18), where
Oh. arr. - assessment of the level of education, which is usually accepted (is
0.15 for persons with incomplete secondary education;
0.60 - for persons with secondary education;
0.75 - for persons with secondary technical and incomplete higher education;
1.00 - for persons with higher education in their specialty;
C - work experience in the specialty. In accordance with the recommendations of the Research Institute of Labor, it is divided by 4 (due to the fact that, as established, experience has 4 times less effect on labor productivity than education);
B - age. In accordance with the recommendations of the Research Institute of Labor, it is divided by 18 (it has been established that the influence of age on labor productivity is 18 times less than the influence of education). At the same time, 55 years is taken as the upper age limit for men, and 50 for women.
At the stage of determining the overall assessment of a candidate for a position, the value of the coefficient of professional prospects is added to the overall assessment of business and organizational qualities. It is quite obvious that preference is given to the candidate who ultimately receives the highest rating, and besides, it is also necessary to take into account the ratings obtained in the same way for groups of personal qualities.
Testing by practice is undeniably critical to the evaluation of any managerial worker. In this regard, it is recommended to carefully and patiently test and recognize among them real organizers who combine interest in the matter with the ability to establish joint work of a team of people. To do this, when working with a reserve of personnel for promotion, it is widely practiced to involve them in the temporary performance of the duties of a leader during his absence, internships and other forms and methods of practical verification. However, even in such cases, the results of the work require objective assessments.
A number of Russian organizations have accumulated positive experience in assessing management personnel. The essence of one of the most proven and sufficiently effective methods is to use the qualifications, experience and intuition of the most competent specialists in their field of work - experts. The most important conditions for the application of this method are to ensure the anonymity of the assessments given by experts and the validity of the selection of the composition of expert commissions.
If anonymity is achieved through special questioning or testing, then the validity of the selection of experts consists in their careful preliminary assessment, as well as in the methodically competent and purposeful formation of the quantitative and qualitative composition. For example, the main requirements for an expert are his competence in production management, morality, deep knowledge and recognized ability to solve special problems in accordance with certain functions.
One of the requirements is to be well versed in one of the related special areas of activity, for example, for a technologist - in economics, for an economist - in technology, for a line manager - in legal issues, etc. The traditional requirement for the formation of a group of experts is their ability to conduct examination taking into account various aspects of the activity of the assessed employee.
The main danger in the selection of the composition of experts is to overestimate the share of any subgroup of experts who are obviously close in their opinion. This danger lies not only in the one-sidedness of the assessment, but also in the fact that when processing expert opinions, all other assessments will lose their significance. For the same reason, it is undesirable to include specialists in the expert group whose authority differs sharply from the average authority of the members of the commission.
Expert groups are used in two ways - individually or as a group. An individual, more common and effective way is that each expert gives an anonymous and independent assessment of the others, and then these assessments are summarized. The group method is based on the joint work of experts and obtaining a summary assessment from the entire group as a whole. With this method, the coordination of individual assessments is combined with the sequential acquaintance of each expert with the assessments of the others. It is possible to comply with the requirement of anonymity with this method, but with "paper technology" difficulties already arise.
Group examination includes the following mandatory steps:
Development of a program (selection of a form of group judgment with a list of options or assessments), recommendations, decisions; development of principles and methods for obtaining a group judgment;
Choice of survey technique or ways to identify individual opinions;
Formation of an expert group;
Conducting a survey of experts;
Processing of results (obtaining a group judgment);
Analysis of results.
The result of the experts' work is reflected in the document, the variants of which can be recommendations, summarizing assessments, decisions. The most public version of the design of the examination is the decision. The most common are recommendations, but indicating certain signs of a future solution.
The selection, coordination and approval of expert commissions is usually carried out by the head of the personnel department and the head of the organization (organization). The Head of Human Resources introduces the evaluation methodology to the experts with the help of a scientific consultant who, for the first time, practically supervises the entire work. At the organization level, the composition of expert commissions (a commission for evaluating the heads of the management apparatus, a commission for evaluating line managers of production units, a commission for evaluating specialists of the management apparatus) usually includes 3-5, but not more than 7 people. At the same time, both the person being assessed and his supervisor should be among the experts.
An assessment questionnaire for a manager or a specialist, an expert card for a certified person, a form for the results of a survey of experts for a certified person, as well as a form for calculating the final grade for a certified person - these are mandatory documents of almost any method, regardless of the method of calculation (manually or using a computer).
The main assessment tool is a questionnaire, which is a specially designed questionnaire that includes a list of certain qualities and answer options. At the same time, the list of qualities that a specialist should possess depends on the scope of his activity and position. Therefore, the correct selection of qualities, their list for each specific position, as well as scoring is recommended to be established by a specific expert commission in the organization. This is one of the most important conditions for an objective assessment of an employee.
After discussing the results, each expert commission approves not only the list of qualities to be included in the questionnaire, but also determines the quantitative assessment of each quality in points. For example, ten qualities are recommended in the questionnaire, and all of them are evaluated on a ten-point grading system (see Table 1).
Table 1.
These qualities are located in the questionnaire in a ranked row - from more to less significant. In accordance with this, the first assessment is given to the certified person. It characterizes the degree of importance of quality. The second assessment is an assessment of the degree of manifestation of the same qualities of the person being certified according to four answer options. At the same time, the expert must establish how often the determined quality manifests itself in the certified person according to the following point scale:
1. if always - 1.5 points;
2. if in most cases - 1 point;
3. if sometimes - 0.5 points;
4. if almost never - 0 points.
At the first assessment, a "+" sign is placed in the column against the quality and its assessment in points. In the second assessment, a "+" sign is placed in the column corresponding to the expert's opinion.
Processing of the results of evaluation of the person being certified is carried out by filling in a special form - "Form of the results of the survey of experts for the person being certified". For this purpose, the signs "+" from the questionnaires supplied by the experts are transferred with the sign "V" to this form. In another form ("Form for calculating the average mark of the certified person") for each quality, the average score is calculated using the formula:
Where Zav - the average score for assessing the quality of the degree of its functional significance;
Z10 ... 31 - assessment by the expert of the degree of significance of quality on a ten-point scale;
N10 ... n1 - the number of experts who assigned the quality an assessment according to the degree of significance (in points);
At the same time, in the same form, for each of the qualities, the average score of the degree of manifestation of this quality in the certified person is calculated according to the formula:
Where Psr - the average assessment of the degree of manifestation of qualities in the certified person;
P1.5 ... P0.5 - the degree of manifestation of qualities in the person being certified according to experts;
N1.5 ... n0.5 - the number of experts who attributed the quality to the degree of its manifestation in the person being certified to one or another assessment score;
N is the total number of experts.
Then, for each quality, the weighted average score is determined by the formula:
K \u003d Zsr * Psr, where K is the quality.
The standard is calculated as follows: the sum of points for all 10 qualities (10 + 9 + 8 + 7 + 6 + 5 + 4 + 3 + 2 + 1 \u003d 57) is multiplied by 1.5 (the coefficient of the degree of quality manifestation, if it always appears ), or by 1.0 (if it appears in most cases), or by 0.5 (if the quality is sometimes present). As a result, we get the following results:
1) 57 x 1.5 = 85.5;
2) 57 x 1.0 = 57.0;
3) 57 X 0.5 = 28.5.
Thus, in the first case, if the employee's score is more than 85 points, they formulate a conclusion about his increase or inclusion in the reserve for promotion. In the second case, if the score is from 57 to 85 points, a conclusion is made about the suitability of the position held. In the third case, if the score is below 28 points, the conclusion is recorded that the employee does not correspond to the position held. At the same time, it is also very valuable that a detailed analysis of the assessments of quality experts can help determine for each employee specific areas for improving his performance.