Science-based forecasting is an important tool in modern management. It is used both for strategic planning of the development of individual enterprises, and for the development of long-term socio-economic programs at the state level. The structure and stages of this process are closely related to the methodology and the adopted model.
Definition
Prediction is a system of theoretically sound ideas about the possible future states of an object and its development directions. This concept is similar to the term hypothesis, but, unlike the latter, it is based on quantitative indicators and has greater reliability. A common feature of these two concepts is that they explore a non-existent object or process.
Applied forecasting techniques were actively developed in the 70s. XX century., And the boom of their use abroad continues in our days. This is mainly due to a new direction in research - global issues, the main task of which is to solve world resource, demographic and environmental problems.
Forecasting is a science that has a close relationship with statistics and its analytical methods. During the analysis, the achievements of mathematics, natural sciences and other sciences are widely used.
Forecasting and planning complement each other in various variations. In most cases, a forecast is developed before the plan is created. He can also follow the plan - to determine the possible consequences. In large-scale studies (at the state or region level), the forecast can act as the plan itself.
Goals
The main task of forecasting is to identify effective ways to manage socio-economic processes in society or the economic and technical development of an enterprise.
The methodological foundations for achieving such goals are as follows:
- analysis of economic and technological development trends;
- foresight of various options;
- a comparison of prevailing trends and goals;
- assessment of the possible consequences of economic decisions.
Forecast Methods
Prediction is carried out according to a certain methodology, which is understood as a system of indicators and approaches to the object under study, the logic of research. Other parameters also depend on the method chosen - how many stages of forecasting will be carried out and what will be their content.
Among the huge number of forecasting methods, the following main groups can be distinguished:
1. Individual expert assessments:
- Interviews - information is obtained during the conversation (formalized and non-formalized, preparatory and independent, directed and non-directed).
- Questionnaire survey (individual, group, mass, full-time and correspondence questioning).
- Development of a forecast scenario (used in the areas of management).
- The analytical method is the construction of a tree of goals (for evaluating hierarchical or structural processes).
2. Collective expert assessments based on obtaining an agreed opinion in a group of experts:
- meetings;
- "Round tables";
- Delphi
- "brainstorm";
- method of "trial".
3. Formalized techniques based on the use of mathematical methods of assessment:
- extrapolation;
- mathematical modeling;
- morphological method and others.
4. Integrated techniques that combine several of the above:
- “Double tree” (used for basic research and R&D);
- forecast graph;
- "Pattern" and others.
A correctly selected forecasting method significantly affects its errors. For example, in strategic planning, the method of extrapolation is not used (foresight outside the experimental data or the distribution of properties from one subject area to another).
Stages
The sequence of stages of forecasting in the General case is the work according to the following scheme:
- Training.
- Analysis of internal and external conditions in retrospect.
- Development of options for the development of events along an alternative path.
- Expertise.
- Selection of a suitable model.
- Her score.
- Analysis of the quality of the examination (a priori and a posteriori).
- Implementation of forecasting developments, their control and adjustment (if necessary).
Below is a description of the main stages of forecasting and their characteristics.
Preparatory stage
At the first stage, the following issues are resolved:
- Pre-forecast orientation (formulation of the object of study, problem statement, definition of goals and objectives, primary modeling, design of working hypotheses).
- Information and organizational training.
- The statement of the forecast task.
- Preparation of computer support.
At the stage of forecasting, the performers who must conduct the forecast are also determined. This group may consist of competent employees responsible for organizational work and information support, and also includes an expert commission.
The following points are documented:
- forecasting decision;
- composition of working commissions;
- work schedule;
- analytical review of the problem being studied;
- contracts or other agreements with specialists involved in forecasting.
Analysis
At the second, analytical stage of forecasting, the following types of work are carried out:
- the study of information about the object in retrospect;
- separation of qualitative and quantitative indicators;
- analysis of internal conditions (in relation to the enterprise it can be: its organizational structure, technologies, personnel, industrial culture and other qualitative parameters);
- study and assessment of external conditions (interaction with business partners, suppliers, competitors and consumers, the general state of the economy and society).
In the process of analysis, they diagnose the current state of the object and determine the trends of its further development, identify the main problems and contradictions.
Alternatives
The stage of identifying other, most likely options for the development of an object is one of the key stages of forecasting. The accuracy of the forecast and, accordingly, the effectiveness of decisions made on its basis depend on the correctness of their determination.
At this stage, the following work is carried out:
- development of a list of alternative development options;
- the exclusion of those processes that in a given period have a probability of realization below a threshold value;
- detailed study of each additional option.
Expertise
Based on the available information and previous analysis, an expert study of the object, process or situation is carried out. The result of this forecasting stage is a reasonable conclusion and determination of scenarios in which development will be most likely.
Examination can be carried out by various methods:
- interviewing;
- questioning;
- one-time or multi-round survey of experts;
- anonymous or open exchange of information and other ways.
Model selection
The forecasting model is a simplified description of the studied object or process, which allows you to obtain the necessary information about its future state, directions for achieving such a state, and about the relationships of individual elements of the system. She is chosen based on the research method.
In economic science there are several types of such models:
- functional, describing the operation of the main components;
- models characterized by methods of economic physics (determination of mathematical relationships between various variables of the production process);
- expert (special formulas for processing expert assessments);
- economic, based on determining the dependencies between the economic indicators of the forecasted system;
- procedural (describing managerial interactions and their order).
There are also other model classifications:
- According to the aspects reflected in them - production and social.
- Models designed to describe income, consumption, demographic processes.
- Economic models of various levels (long-term for predicting economic development, intersectoral, industry, production).
In forecast models, the following forms of describing phenomena are distinguished:
- textual;
- graphic (extrapolation methods);
- network (graphs);
- building block diagrams;
- matrix (tables);
- analytical (formulas).
The model is formed using such methods as:
- phenomenological (direct study and observation of occurring phenomena);
- deductive (allocation of parts from the general model);
- inductive (generalization from particular phenomena).
After selecting a model, a forecast is made for certain periods. The results obtained are compared with currently known information.
Quality control
The stage of verification of the forecast, or verification of its reliability, is carried out on the basis of previous experience (a posteriori) or independently of it (a priori). Quality assessment is done using the following criteria: accuracy (spread of forecast trajectories), reliability (probability of the chosen option), reliability (measure of process uncertainty). To assess the deviation of forecast criteria from their actual values, such a concept as forecast errors is used.
The controlling process also compares the results with other models, develops recommendations for managing the facility or process, if such an impact can affect the development of events.
There are 2 methods for quality assessment:
- Differential, in which clear criteria are used (determining the clarity of setting a forecast, the timeliness of phased work, the professional level of performers, the reliability of information sources).
- Integral (generalized estimate).
Key factors
The prediction accuracy is influenced by the following main factors:
- competence of an expert group;
- quality of prepared information;
- accuracy of measuring economic data;
- level of methods and procedures used in forecasting;
- correct model selection;
- consistency of methodological approaches between different specialists.
Often, large errors also arise due to the fact that the features of the conditions in which this model is applied are not taken into account.
Implementation
The last stage of forecasting is the implementation of the forecast and monitoring the progress of its implementation. If critical deviations are identified that can significantly affect the further development of events, the forecast is adjusted.
The level of adoption of corrective decisions may vary. If they are insignificant, then the adjustment is carried out by the analytical group, which is responsible for the development of the forecast. In some cases, experts are involved in this work.