The Concept
04-2-1
Frequency
Cumulative Frequency
28.83
The procedure:
generate a random number between 0 and 1 (calculator, table,
program, ...
read the quantile associated to that random number
For Example:
Random Number
0.7807
0.1562
0.6587
0.8934
Simulated Number
28.83
...
04-2-2
Geostatistics
Expert knowledge
Past prices/costs
Incorporate uncertainty in the different parameters
Incorporate the probability of certain
unknowns/decisions/options/scenarios
The Bootstrap
04-2-3
Simple Examples
0.4
Number of Data 17
mean 1225.28
std. dev. 672.14
coef. of var 0.55
Frequency
0.2
mean 1223.46
std. dev. 156.98
coef. of var 0.13
3360.87
1305.14
1145.50
758.81
485.00
maximum
upper quartile
median
lower quartile
minimum
0.3
Frequency
maximum
upper quartile
median
lower quartile
minimum
1814.71
1320.91
1214.98
1111.62
836.53
0.2
0.1
0.1
0.0
0.0
0.
2000.
3000.
4000.
0.
1000.
2000.
3000.
1000.
9.0
0.06
= 0.54
7.0
0.04
5.0
0.02
3.0
0.00
0.
5000.
10000.
15000.
20000.
25000.
0.25
0.35
0.45
0.55
04-2-4
0.65
0.75
4000.
NPV example
Consider uncertainty in:
Interest rate (normal distribution with m=10% std=1%
Mineral commodity price (flat real $) m=50$/unit std=5$/unit
Number of years of production (similar to uncertainty in the
reserves available) m=10yrs, std=2yrs
The following are known with 100% certainty:
100M capital costs
1M units/year
All operating costs are factored into the mineral
commodity price (not realistic, but the focus of this
example is on MCS not cost estimation).
Reclamation costs will be 20M/year for 2 years after mining
WHAT IS THE NPV (DISTRIBUTION) OF THIS MINE???
Lets do 1 (or 2) realizations by hand
Then 100 realizations with Excel (easy with computers)
04-2-5
Set one input variable to the HIGEST value (or maybe p90)
Set all other variables to the average or p50
Alternatively, could do n simulations with this one
variable set to the p90 and then average the results
Obtain the high value
MCS by Hand
Determine distributions
Draw values from the distributions
Apply transfer function
Draw random number, lets use 0.1313 for an example
Get a chart (make sure you know what kind of table it is
there are different ones)
Lookup the value 0.1313 in the body of the table and get 1.12
Convert this to a non-normal distribution by Multiplying the
std and adding mean: -1.12*1.0+10=8.88
check in Excel if you want: =NORMINV(0.1313,10,1)
04-2-6
04-2-7