This document describes using Monte Carlo simulation to analyze the investment risk of a mining project in China. It introduces Monte Carlo simulation and how it can handle multiple uncertain risk factors through random sampling. The document uses net present value as the risk assessment index and establishes a simulation model for an iron ore mining investment based on expected distributions of factors like mine life, investment, production, price, and spending. It samples those factors 1000 times to analyze the investment risk and NPV.
This document describes using Monte Carlo simulation to analyze the investment risk of a mining project in China. It introduces Monte Carlo simulation and how it can handle multiple uncertain risk factors through random sampling. The document uses net present value as the risk assessment index and establishes a simulation model for an iron ore mining investment based on expected distributions of factors like mine life, investment, production, price, and spending. It samples those factors 1000 times to analyze the investment risk and NPV.
This document describes using Monte Carlo simulation to analyze the investment risk of a mining project in China. It introduces Monte Carlo simulation and how it can handle multiple uncertain risk factors through random sampling. The document uses net present value as the risk assessment index and establishes a simulation model for an iron ore mining investment based on expected distributions of factors like mine life, investment, production, price, and spending. It samples those factors 1000 times to analyze the investment risk and NPV.
This document describes using Monte Carlo simulation to analyze the investment risk of a mining project in China. It introduces Monte Carlo simulation and how it can handle multiple uncertain risk factors through random sampling. The document uses net present value as the risk assessment index and establishes a simulation model for an iron ore mining investment based on expected distributions of factors like mine life, investment, production, price, and spending. It samples those factors 1000 times to analyze the investment risk and NPV.
2011 International Conference on Management of e-Commerce and e-Government
Mining Investment Risk Analysis Based on Monte Carlo Simulation
Jiang Wei Zhang Jian Li Jianglan
School of Information and Safety School of Information and Safety School of Information and Safety Engineering Engineering Engineering Zhongnan University of Economics Zhongnan University of Economics Zhongnan University of Economics and Law and Law and Law Wuhan, China Wuhan, China Wuhan, China [email protected][email protected] Wuhan, [email protected]
Abstract—When investing mining projects, we often cannot of the engineering investment risk analysis [4], some related make an accurate judgment of the investing project’s risk project cases have proved it [5-7]. degree because there are too many influential risk factors, with different kinds of uncertainties at the same time. To solve this II. MONTE CARLO METHOD problem, Monte Carlo simulation method is an effective way. It can be accomplished by computer software with enough Monte Carlo simulation method takes the probability times of sampling process. The analysis of the output can help and mathematical statistics theory as a foundation. It judge the degree of risk. In this paper, the Monte Carlo simulates the changes of factors affecting the project risks simulation was performed in computer and applied to an iron through several random samplings, and then calculating ore mine investment. Base on the simulating results, this how these changes influence the evaluation index. The basic project’s investing risk was analyzed. We illustrated the way is to use a mathematical model to indicate the target meanings of the simulation results and also provided experimenting variables [8-9]. suggestions to the investors. y f ( x1 , x2 , x3 , , xn ) ˄1˅ Keywords-Mining investment; Monte Carlo; Risk Among them, xi (i 1, 2, , n) is independent random identification; Risk analysis variables; Y is these variables’ function, also the evaluation I. INTRODUCTION index in the effect of risk analysis. Like any other project, mining investment has risks. The This model includes the key risk variables that influence exploration and development of ore deposit is a complex the target function. Each risk variable has its own project, we have to care about the risk when making an probability distribution. By using random numbers, we can investment decision, this is because the characteristics of ore get sampling values based on the distribution of each deposits in decision-making is completely unable to find out, variable, and then calculate the function by putting these such as market conditions, and operating costs, mining samples into the model. After repeating the above process policy and so on, these various environmental parameters several times, you can get a risk analysis result. can be perfectly clear only afterwards. Mining investment The Monte Carlo simulation method puts the uncertainty risks mainly come from three parts, the uncertainty of natural question associated with a probability model, gives us a conditions, social environment and basic materials [1]. For a statistical estimate as an approximate solution of the original mining investment, because of the long period of problem base on amount of random sampling tests. construction and investment recovery, and the existing of massive uncertain factors, whether this investment spending III. CHOOSE A RISK ASSESSMENT INDEX AND BUILD THE can get an expected return is uncertain. In addition, such MODULE investments always have a huge amount of spending, it’s hard to quit and avoid the loss once the infrastructure is done. There are many kinds of index to judge the economic Therefore, the risk analysis is not dispensable in a mining effect of a project, such as internal rate of return (IRR), feasibility study, we should give enough attention to it [2]. financial net present value (FNPV) and return on investment In the mining investment risk analyzing, many factors period (Pt), etc. The Net Present Value (NPV) turns the Net involved, and every factor exists according to their own and cash flow in every year to the Present Value with a discount different probability distribution form, so it’s a complex rate ( ic ), and then adds them together. It is the most popular work. The Monte Carlo simulation’s advantage is to solve index in economical evaluating and commercial analyzing. problems with many risk factors. If every factor has a The expression is like the following formula (2). certain distribution in its value range, or you can combine n them in any form, the simulation can give you an analyzing NPV (i ) (CI t COt )(1 i ) t ˄2˅ result [3]. At the same time, by building calculation model t 0 and using the computer simulation, Monte Carlo simulation technology can greatly improve the accuracy and efficiency
DOI 10.1109/ICMeCG.2011.64 In the formula above, CI t is the cash inflow of period T, year afterwards, CI t (t = 1, 2,..., n) = PRODUCThPRICE, COt is the cash outflow of period T, n is the life of project. COt is annual spending, i means the discount rate. We use the yield on the benchmark project to discount cash flow. If the net present value is zero, it means this project IV. SIMULATION OF MONTE CARLO has reached the required standard yields level; If the net present value is greater than zero, it means except the A. Implementation benchmark rate of return, this project can get a higher There are different ways to achieve Monte Carlo income level; If the net present value is less than zero, it simulation. Microsoft Office Excel is a popular software to means the yield rate has not reached required standards achieve. It provides such function like RAND ( ) and yields. Therefore, when evaluating the risk of an investment, NORMINV ( ) to realize sampling simulation of the the index can be judged according to the following method, uniform distribution and normal distribution. In order to get if NPV(i) ı0, this scheme can be adopted economically, if enough random sampling data and a more accurate NPV (i) < 0, this plan is not beneficial. simulation result, we can set a big number of simulation In this paper we take a domestic iron ore mining project times. In this simulation we execute 1000 times. investment as the object of study, and establish the model According to the data of similar investment projects and information from experts in this industry, the distribution of using the net present value method. CI 0 means the fixed each risk factor can be estimated as the following chart, investment at the beginning, CI t is the cash flow for every including life of mining, investment, annual production, iron ore price and annual spending, etc. TABLE I. DISTRIBUTION OF EACH FACTOR
Among the risk factors above, each of them needs a random number as a parameter, except the discount rate (10%). Because every factor is independent from each other, Ă Ă Ă Ă Ă so we need to build 5 groups of random numbers (table II). “Life of mining˄year˅” obeys even distribution 15-25. It Each group produces 1000 random numbers. Function is realized by using function RAND ( )*(25-15)+15. RAND ( ) can produce a random real between 0 and 1. Because the life of mining must be integral namely type INT, so we need to turn this number into an integer by using TABLE II. SETTING RANDOMS INT (RAND () * (25-15) + 15). "Investment", "annual output", "price" and "annual 5DQGRP 5DQGRP 5DQGRP 5DQGRP 5DQGRP spending" all obey normal distribution. It is realized by using function NORMINV (RAND (), μ, σ). Random numbers from 0 to 1 can be created by RAND (), μ is the expectation, σ is the standard deviation.
The "Net present value" NPV=B10-PV (10%, C10, D10 * E10) + PV (10%, C10, F10). PV () is a financial function, it returns a current total value of a series of investments in the future.
Each variable executes 1000 times, so that we can get 1000 NPV values. The simulation results are shown in table III.
TABLE IV. STATISTICAL DESCRIPTION TO NPV RESULTS TABLE V. FREQUENCY AND ACCUMULATIVE FREQUENCY
Statistical Description to NPV Results Interval Frequency Accumulative %
Mean 61377.9256 -30000 0 0.00% Standard error 926.958715 -20000 1 0.10% Median 60904.2387 -10000 5 0.60% Mode #N/A 0 12 1.80% Standard deviation 29313.0084 10000 23 4.10% Variance 859252459 20000 37 7.80% Kurtosis 0.20251893 30000 63 14.10% Skewness 0.16318209 40000 86 22.70% Range 195992.503 50000 117 34.40% Minimum -22394.865 60000 146 49.00% Maximum 173597.637 70000 150 64.00% Summation 61377925.6 80000 114 75.40% Times of observation 1000 90000 90 84.40% Max(1) 173597.637 100000 57 90.10% Min(1) -22394.865 110000 47 94.80%
Confidence(95.0%) 1819.00948 120000 26 97.40%
130000 11 98.50% We can get some statistics results based on the 140000 6 99.10% simulation data. Table IV shows the statistical description to NPV results. According to the maximum and minimum 150000 8 99.90% value in the simulation, the whole range can be divided into 160000 0 99.90% 22 sections. Table V listed the frequency and accumulative 170000 0 99.90% frequency that the NPV sampling results appear in each 180000 1 100.00% section. Others 0 100.00%
74 Figure VI is a histogram generated according to table V, it experimental results. shows intuitively the distribution of 1000 times’ TABLE VI. Histogram
2W The results above explains when the observation number decision makers should also consider the risk when is 1 000, NPV approximately obey a normal distribution. choosing a suitable distribution law for all the risk factors. The main value is 61377.9 and the standard deviation is 29313. In the sampling the minimum value is 22394.9 (ten REFERENCES thousand Yuan) and the maximum is 173597.6 (ten [1] Deng Peidi, “Mining investment risk analysis,” Journal of Anhui thousand Yuan). "95% confidence interval" is from 1819.0 Vocational College of Metallurgy and Technology, 2004, 14 (4), pp. 112-114. to 171778.6, it means the probability that the NPV values in [2] Yang Peng, Cai Sijing, Advanced hard-rock mining study, Beijing: between this section is 95%. In addition, we can see from Metallurgy Industry Press, 2010, 55-59. the data that the probability of “NPVı0” is up to 98.2%. [3] Wu Xuan, “The project investment risk assessment model based on monte carlo simulation method,” Science and Technology Ecnony This investment has a strong ability to resist risks. Market, 2007(5), pp. 26-29. [4] CHOIHH,CHOHN,SEOJW, “Risk assessment methodology for V. CONCLUSIONS underground construction projects,” Journal of Construction Through an actual case, an investment risk analyzing is Engineering and Management,2004,130(2), pp. 258-272. [5] Feng Wentao, “Random decision model based on Excel, ” China achieved by using Monte Carlo simulation. The analysis Management Informationization, 2007,10(1),pp. 56-58. result shows this method can clearly reveal the expected [6] Xie Lan, “Investment project risk simulation analysis based on earnings of an investment project, it can provide reasonable Excel,” China Management Informationization, 2007,10(1),pp. 58- references for investors. Especially when the risk factors are 61. [7] Xu Li, Ling Huqian, Zhu Tongbin, “Monte Carlo Simulation on too many and it is difficult for investors to judge the risk, Economic Risk of Flue Gas Desulfurization BOT Project,” this method can effectively solve this problem. But at the Technology Economics, 2008,27(5),pp. 52-56. same time we should see shortcomings of this method, the [8] Wang Jiayuan, Liu Chunle, Risk management of construction project, accuracy of analysis results when forecasting. In some Beijing, China's water conservancy and hydropower press, intellectual property rights press, 2004:109-110. projects, the distribution of factors is uncertain. We cannot [9] Shao Ruiqing, The international shipping investment decision and simply use a distribution law type to describe it. So the method, Shanghai, Shanghai SDX Joint Publishing Company, 2006:11.
