The authors discuss basic methods of statistical surveillance used to assess risks associated with construction investment projects. The methods considered in this article include a sensitivity analysis (based on the assessments of initial parameters of a project and its final characteristics (usually IRR or NPV)), a method of scenarios (based on assessment of the risk of the project failure considered as the total of all probabilities of a negative NPV), a method of discount rate adjustments (based on the adjustment of the discount rate by adding a risk premium calculated by an expert or an in-house specialist), a decision tree method (whereby a project is reduced to development of a hierarchical scheme of all possible actions; results of an investment project have a treeshaped structure), a Monte Carlo method (meaning a special case of simulation). The authors describe their main advantages, disadvantages and problems that accompany their practical application. The authors also describe methods based on the fuzzy logic theory and suggest formalization of inputs and outputs as a fuzzy interval. General milestones include fuzzification, development of fuzzy rules, de-fuzzification, and methods of de-fuzzification (including a method of maximum centre, a method of maximum value, and a centroid method). The authors make their conclusion in respect of the method that may be regarded as the most suitable for investment projects in the construction industry.
Рассмотрены основные статистические методы количественной оценки рисков строительно-инвестиционного проекта, а также методы, основывающиеся на теории нечеткой логики.