2022

Assignment 2

ACHINTYA ACHARYA
ASHUTOSH PORWAL
PARTH MAU
RUPAL BILAIYA
Q1. Developing New Car at GF Auto

a. Complete the partially built simulation model (see the file NewCar.xlsx) and explain the logic

behind each formula in the spreadsheet model.

• Fixed Development Cost has been calculated using the three-point estimation on the data from

the first sheet as the triangle-fit was found to be the best fit for the given data.

• Unit Sales for the first year has been calculated using the three-point estimation of the data

from the question. Unit sales for the years afterwards have been calculated using the decay

(calculated from the three-point estimation of the data from the question) on the sales from the

previous year.

• Unit Contribution for the first year is given and Unit Contribution for every other year

has been decreased by 4% as per the question.

• Revenue minus variable cost is the product of Unit Sales and Unit Contribution for that

particular year.

• Depreciation has been calculated by dividing the Fixed Development Cost by 5. We did so

because we would like a uniform tax write-off across the five-year period. Dividing the total

cost by 5 would ensure parity across this period.

• Before Tax Profit has been calculated by subtracting Depreciation value from the Total Profit

(Revenue minus variable cost). This is the value that is going to be taxed. Thus, fraction of

fixed cost which we intend to use as a tax write-off is being subtracted from the profit earned

in that particular year.

• After Tax Profit has been calculated by simply multiplying the Before Tax Profit value by

(1 - Tax Rate). As the name suggests, this value is nothing but the profit that remains after
 taxes. Thus, multiplying the pre-tax profits by (1-Tax Rate) deducts the taxes levied, and leaves

us with After Tax Profits.

• Cash Flow has been calculated by adding the After-Tax Profit to the Depreciation value.

This is because the depreciation cost has already been made and is being uniformly considered

every year as a tax write-off. Thus, at the end of every year the cash-in-hand for company is

the money earned after deducting taxes and the depreciation value.

• Net Present Value of any venture is given by (Today's value of the expected future cash flows

– Today's value of invested cash). To calculate the Net Present Value of Cash Flow, we

subtract the Fixed Development Cost from calculated NPV using inbuilt Excel NPV formula

(Parameters: Discount Rate, Individual Cash Flow at the end of every year).

b. Use your model to simulate the NPV with 1,000 iterations and estimate the mean and

standard deviation of the NPV. Also, determine an interval so that you are 90% certain that

the NPV will be within this interval.

Standard Deviation of NPV: $227,994,250.84

90% CI for this sample (1000 iterations): -$218 million to $536 million

c. The simulation results could be somewhat comforting but also a cause of concern for GF. As

per your risk analysis, what are the bright side and dark side of developing this new car?

Pros –

• With a greater than 78% confidence, it can be said that GF will remain profitable.

• Best case scenario NPV that can be derived from the venture is $808 million.

Cons –

• This could be a high-risk project as there is 22% chance of GF losing money on the project.

• Although the expected value of NPV is positive ($175 million), there is a high standard

deviation of $227 million. If GF is a risk averse organization, this may not be a viable

option.
d. Financial analysts often call the 5th percentile of a distribution the value at risk, or VAR,

because it indicates nearly the worst possible outcome. From your simulation, what is

approximately GF’s VAR?

For GF, according to financial analysts, nearly the worst possible outcome would be a loss of $218

million as given by results of the simulation. The worst would be a loss of $514 million.
e. Which uncertain input has the largest (positive or negative) influence on the NPV? If GF wants

to get a more favorable NPV, what do you suggest GF to boost?

The initial fixed cost uncertainty has the largest negative impact on the NPV for GF. To get a more

favorable NPV, unit sales for 1st year should be boosted as they have the most positive impact on

the NPV. Alternatively, the unit contribution per car can be increased by a higher priced model to

increase the overall revenue generated through all the sales. Various other factors would need to

be considered in either of these cases such as increase in marketing costs and manufacturing costs

or reduction in demand due to higher pricing.

a. Complete the partially-built simulation model (see the file Retirement.xlsx), and explain the

logic behind each formula in the spreadsheet model.

For completing simulation model for this case, we first assigned weight as 1 for most recent year

2021 and then the weight for any year is damping factor of 0.98 multiplied with weight of next

year. For example, weight for year 2019 is the multiple of weight of 2020 - (0.98 * 0.98) = 0.9604.

This calculation can be seen in ProbWts column in excel. Moving on to the Probability column,

as the name suggest it is just the probability corresponding to the weights and to calculate this, we

can just divide the weight for the corresponding year with sum of all the weights.

Next, we need to decide some random weights for the 3 bonds adding up to 1, to produce largest

mean final cash. Some percentage is assigned to stock and remaining is divided between T-bonds

and T-bills. We can try 6 different sets of weights in our case as mentioned in “Alternative sets

of weights to test” in excel. To decide which weight to use out of these 6, we have RiskSimTable

created in cell A19 and then corresponding weights for all the 3 stocks is fetched based on Index

value using VLOOKUP.

Moving to the Simulation Model, the goal is to simulate 40 scenarios one for every year of Sally’s

investing. We use RiskDiscrete in column K23:K62 to fetch a random year from A23:A98 based

on the probabilities mentioned in G23:G98. Once we have the Scenario, we use VLOOKUP to

fetch the corresponding value of T-Bills, T-Bonds, Stocks, Inflation in columns L, M, N, O

respectively for all the 40 scenarios. To calculate the Ending cash, we will just multiply the values

of 3 bonds with their corresponding weights from row 19 for that row’s Beginning cash. And next
year Beginning cash would be previous year money plus additional investment of $1000 every

year.

Now we need to deflate future’s dollar to today’s dollar and therefore to calculate Deflator for that

scenario, we will divide 1 by the product of inflation factor for all the scenarios before that. And

finally, to calculate Final cash as per today’s dollar we will just get the Final cash value from cell

P62 and multiply it with Q62 to adjust the final cash according to deflator factor. We will just the

define the cell K18 as output and run simulation now.

b. Set the number of iterations to 1,000 and the number of simulations to 6 (one for each set of

investment weights you want to study). Then run @RISK as usual, and discuss the simulation

results.

Below are the summary reports for all the 6 simulations:

we look at the histogram for the same:

Here we have put 80% of the investment in stocks and we can observe a lot of skewness. In case

of investments, the 5th percentile which represents VAR 5% plays a very important role. As Sally

has invested $40,000 and with a 5% chance, she is able to make a minimum of $50,421.48. So
overall this investment might turn out to be profitable, and Sally can further try running simulations

with different weights to increase this value as much as possible.

c. Which set of investment weights has the best upside potential (use the 95th percentile)? Which

set of investment weights has the best downside (use the VAR)?

The first simulation has best upside potential with 95% percentile value close to $407,424.07. The

division for first simulation is 10%, 10%, and 80% for T-Bonds, T-Bills and Stocks respectively.

The same simulation has the best downside with $50,421.48. This shows that assigning higher

weight to stocks leads to more profit, so sally can try few more combinations of weights for all 3

and check for high 95th percentile and VAR 5% value.

d. Modify your spreadsheet model slightly for a shorter time horizon such as 15 years, and report

your simulation results and findings.

Although stock strategies are more profitable over long horizons, it might not be the case all the

time, and therefore it is better to explore results for 10 or 15 years of horizon as well. If Sally

invests for 15 years rather than 40 years, the max mean final cash is coming out of simulation 1

again that is 10%-10%-80% ratio. The histogram below shows results for the same. We can see

that 95% percentile value is $38,299and VAR 5% value is $11,526. But the total investment is of

$15,000 and therefore it is risky as there is 5% chance that Sally might be in loss by the end of 15

years. We believe that Sally should invest money for longer period as it might be less risky.