Monter Carlo Simulation: Examples

“…we note that best practices have developed for the use of Monte Carlo simulation and
the use of such models (with the assistance of an outside valuation specialist) is now
commonplace.” PwC guide Stock-based compensation, 9/2023, P. 8-25

“It seems that only the large firms use it (i.e., Monte Carlo simulation- insert added),
probably due to the complexity of their assignments.” Gary Trugman Understanding
Business Valuation 6th Edition, P. 277

Executive Summary
A Monte Carlo simulation uses samples of input data together with a known mathematical model to
predict probabilistic outcomes. The model we use in our publication is geometrical Brownian motion or
GBM. GBM is widely used to model many financial variables such as changes in stock prices, company’s
EBITDA, revenue, etc.

GBM is a probabilistic model in the sense that it includes an element of uncertainty or randomness in its
prediction. Specifically, included in GBM formula are randomly generated numbers following normal
distribution with a mean of 0 and a standard deviation of 1.

Generally, Monte Carlo simulation is used in cases involving path dependency, i.e., situations where the
result in a subsequent period depends on the results in a prior period.

When you use a probabilistic model to simulate an outcome, you will get different results for each trial.
Runing enough simulations to produce different outcomes will mimic real-life results. Using probabilistic
models contrasts conventional forecasting methods which produce more deterministic results. In our
case we ran 10,000 trials in Microsoft Excel to determine the expected sough-after value.

Our illustration of Monte Carlo simulation includes three examples. As part of first example, we
determine fair value of a contingent consideration payable in connection with acquisition of a business.
Second example presents an employee bonus plan with terms similar to contingent consideration from
first example. Last example has to do with determination of a bonus liability, dependent on changes in
company’s stock price. All our examples cover Monte Carlo simulation to estimate fair values for
financial reporting purposes.

As part of our analysis, we explain key concepts used in GBM valuation, including risk-neutral valuation,
use of randomly generated number following normal distribution pattern. Our publication also includes
detailed step by step application of GBM formula.

1|P age
Background
A Monte Carlo simulation uses repeated random sampling to obtain numerical results of a sought-after
variable subject to estimation. A typical Monte Carlo model spans multiple periods with features that
create path dependency, i.e., a situation where the result in a subsequent period is calculated knowing
the results in a prior period. Simulated results are calculated for each period and are discounted to
present value at the rate reflecting risk associated with the relevant metric.

Other cases where Monte Carlo models apply involve estimated matric, which is a function of multiple
correlated financial outcomes.

A Monte Carlo simulation model assumes that the underlying variable (e.g., entity’s stock price, EBITDA,
etc.) follows a Geometric Brownian Motion stochastic process.

Monte Carlo modeling applies to determination of fair value of stock compensation as well as other
payoffs such as contingent consideration related to acquisition of a business and employee bonus plan.
For stock-based compensation, a large number of sample paths are simulated, using generator of
random numbers. After that, a fair value of the award is determined for each sample path based on the
payout value of the award discounted to the grant date. Fair value used for financial reporting purposes
is the simple average of individual fair values of each sample.

Lattice modelling is another valuation approach applied in relation to stock-based compensation. A

simple lattice model employs a binomial tree to show different paths that the price of an underlying
stock might take over option term. The binominal share price tree consists of branches representing
possible future share price paths. Simple binominal trees have the so call combining or overlapping
paths where a decrease in the share price proceeding an increase gets to the same price as if the share
price experienced an increase and then a decrease.

Unlike lattice simulation, Monte Carlo models generally present a set of “straight” paths as opposed to a
tree of possible up or down price movements.

For many practitioners it is easier to learn Monte Carlo method through examples. In the publication we
provide a detailed example covering a contingent consideration and two other supporting examples. All
our examples cover Monte Carlo simulation to estimate fair value for financial reporting purposes.

Valuation Example: Contingent Consideration

The example in our publication is based on the example of contingent consideration payable as part of
business combination provided by The Appraisal Foundation in Valuation In Financial Reporting (VFR)
Valuation Advisory 4- Valuation of Contingent Consideration 1.

Example A: Contingent Consideration

According to the terms of acquisition agreement, Company A, the acquiror is required to pay 20% of the
excess of Company B, an acquiree’s annual EBITDA above $ 1,000 K for the first year and 20% of the
excess of the acquiree’s annual EBITDA above $ 1,400 K for the second year. First year payment cap is $
200 K. Second year payment cap is $ 250 K. According to payment terms, if first year payment cap is not
reached, then any shortfall, i.e., the difference between actual payment and the payment cap will be
added to the second-year cap.

1
Example 9.10: Multi-year, Path Dependent VFR Valuation Advisory# 4- Valuation of Contingent Consideration

2|P age
The payments are due three months after the end of each earnout period.
Key terms described above can be documented as follows:

Percentage payout: 20%

Payout cap- year 1 $ 200 K
Payout cap- year 2 $ 250 K
Payout period- period 1 1.25
Payout period- period 2 2.25
The following assumptions apply:

Forecasted EBITDA- year 1: $ 1,000 K

Forecasted EBITDA- year 2 $ 1,400 K
Company A credit risk 2%
Risk free rate 3%
EBITDA discount rate 15%
EBITDA volatility 40%

Forecasted EBITDAs for both years represent an expected value, i.e., a probability-weighted result across
possible outcomes. The fact that expected EBITDAs are the same as contractual thresholds mean that
the thresholds were determined using the same expected prospective financial information. However,
the thresholds do not have to equal expected EBITDAs.

Company’s A credit risk was determined using estimated credit spread, i.e., rate of return above the
risk-free rate required by investors to compensate for counterparty’s credit risk. EBITDA discount rate
reflects market participant risk view of expected EBITDA. Conceptually, the rate reflects systematic or
non-diversifiable risk as the expected value already incorporates other risk, i.e., idiosyncratic or
unsystematic risk.

Valuation Advisory: Valuation of Contingent Consideration presents detailed guidance for estimating
relevant discount rate and volatility that would apply to EBITDA or another metric with non-diversifiable
(i.e., market or systematic) risk 2.

Geometric Brownian Motion

We will use geometric Brownian motion (GBM) as the model to simulate EBITDA behavior. The motion
represents a certain probabilistic pattern frequently used as part of Monte Carlo simulation. As part of
the motion, changes in EBITDA in a small period of time are normally distributed and the changes in two
nonoverlapping periods are independent. More specifically, changes (referred to as a “drift”) have a
mean of zero and a variance rate of 1 per unit of time (e.g. a year).

Normal distributions with a mean of 0 and a variance of 1 are a special case of a common, continuous
probability distribution with a symmetrical, bell shape.

2
Sections 5.2.2 Discount Rate and Market Risk Considerations, 5.2.3 Methods for Estimating the Required Matric Risk Premium,
5.2.4 Estimating Volatility VFR Valuation Advisory# 4- Valuation of Contingent Consideration.

3|P age
Geometric Brownian motion is a type of stochastic process, which in broad terms, is defined as any
process describing the evolution in time of a random phenomenon. Note that geometric Brownian
motion is also used in the Black-Scholes-Merton model.

Use of GBM to model future changes in EBITDA means that future EBITDA values have a lognormal
distribution 3. A variable has a lognormal distribution if the natural logarithm of the variable is normally
distributed.

Before we get into further details of GBM, let us clarify certain key concepts of the model.
Risk-neutral Valuation
Expected changes in EBITDA, i.e., ∆EBITDA/EBITDA or a drift is represented by a risk-free rate. Use of a
risk-free rate as a return or a discount factor in valuations is an attribute of the so-called risk-neutral
valuation.
Expected value approach to discounting has two methods. Method 1 adjusts the expected cash flow of
an asset for systematic (that is, market) risk by subtracting a risk premium to produce a risk-adjusted
expected cash flows. Such adjusted expected cash flow is referred to as a certainty equivalent cash flow.
Since the expected cash flows were reduced by the systematic risk factor, the discount rate is the risk-
free rate.
Method 2 adjusts for systematic risk by including this risk in the discount factor. The two methods are
described ASC 820-10-55-13 through 20 with the example of both methods provided in ASC 820-10-55-
19 4.
Method 1 serves as a foundation for the risk-neutral valuation, i.e., the valuation performed by
removing the systematic risk and applying the risk-free rate as a discount factor. This valuation approach
is illustrated by using risk-free rate as an expected return.
Another way to describe a risk-neutral valuation is by using the assumption that the investors are risk-
neutral. Risk-neutral investors are concerned about expected returns without regard to the distribution
of individual return amounts and their probabilities. In other words, the assumption means that
investors do not increase the expected return they require to compensate for the increased risk.
According to John Hull and Sankarshan Basu 5:
The world we live in is , of course, not a risk-neutral world. The higher the risks investors take, the higher
the expected returns they require. However, it turns out that assuming a risk-neutral world gives us the
right option price for the world we live, as well as for a risk-neutral world. Almost miraculously, it
finesses the problem that we know hardly anything about the risk aversion of the buyers and sellers of
options.

Since risk-neutral investors are not concerned about risk, the discount rate relevant to them is risk-free
rate.

3
Section 14.7 The Lognormal Property, Chapter 14 Wiener Processes and Ito’s Lemma, John C. Hull, Sankarshan Basu, Options,
Futures, and Other Derivatives.
4
See also pages 122-123 of AICPA Accounting and Valuation Guide Assets Acquired to Be Used in Research and Development
Activities.
5Chapter 13.2 Risk-Neutral Valuation, John C. Hull, Sankarshan Basu, Options, Futures, and Other Derivatives. Also see section

4.6 Risk-Neutral Valuation VFR Valuation Advisory# 4- Valuation of Contingent Consideration.

4|P age
Standard Normal Random Distribution
One of the variables in the GBM formula uses generator of random numbers. More specifically, the
formula uses a normal distribution with a mean of zero and a standard deviation of 1. Numbers
following this distribution can be generated using the following excel formula: NORM.INV(RAND(),0,1).
Function RAND () produces a random number greater than or equal to 0 and less than 1.

Reference to “INV” in function NORM.INV(RAND(),0,1) refers to inverse cumulative distribution. The

reference to “inverse” means that for a given cumulative probability “Y”, varying between 0 and 1, the
formula produces variable value reflected on the horizontal line “X” as shown on a standard cumulative
normal distribution graph. We are providing such a graph below:

Graph 1

Norm Distribution: cumulative

1.200
Y- CUMULATIVE PROBABILITY

1.000

0.800

0.600

0.400

0.200
X- VALUE OF THE VARIABLE OR DATA
-
-6 -4 -2 0 2 4 6
(0.200)

In a bell-shaped normal distribution probability is reflected along axis Y, a dependent variable while
individual observations are reflected along axis X, an independent variable. Producing individual
observations as if they were dependent variables explains the reference to “inverse”.

In practice, formula NORM.INV(RAND(),0,1) produces data varying, in practically all cases, between -4
and +4. This data can be shown along the line Y on the graph below:

Graph 2

5|P age
 Norm Distribution: cumulative
5

Y- VALUE OF THE VARIABLE OF DATA

4
3
2
1
0
(0.200) - 0.200 0.400 0.600 0.800 1.000 1.200
-1
X- CUMULATIVE PROBABILITY
-2
-3
-4
-5

Graph 2 switches axis X and X, as compared to Graph 1.

Pattern of variables following standard normal distribution is used to describe Wiener process.

Wiener Process
A variable E follows a process referred to as Wiener process if a) the change ∆𝐸𝐸 during a small period of
time ∆𝑡𝑡 is determined as ∆𝐸𝐸 = 𝜖𝜖√∆𝑡𝑡 and b) changes in z should be independent of each other.

∆𝑡𝑡 is a time step for which a change in z is calculated. 𝜖𝜖 has a standard normal distribution with a mean
of 0 and standard deviation of 1.0, determined using NORM.INV excel formula. If the formula uses RAND
component as described above, the independent requirement is met.

You will see that GBM model uses 𝜖𝜖√∆𝑡𝑡 or the Wiener process.

Brownian (or Wiener) process can be defined as a stochastic process where the change in a variable
during each short period of time of ∆𝑡𝑡 has a normal distribution with a mean zero and a variance of ∆𝑡𝑡.

The discrete-time version of GBM model is as follows:

∆𝐸𝐸
= 𝜇𝜇∆𝑡𝑡 + 𝛿𝛿𝛿𝛿√∆𝑡𝑡 [1]
𝐸𝐸

Where ∆𝐸𝐸 is a change in EBITDA E, 𝛿𝛿 is volatility, 𝜇𝜇 is expected change in EBITDA or a drift, ∆𝑡𝑡 is a time
step, 𝜖𝜖 has a standard normal distribution with a mean of 0 and standard deviation of 1.0.

Note that a drift is represented by a risk-free rate, consistent with prior explanations.

6|P age
A more accurate solution which is based on simulation of LN(E) rather than E. The respective GBM
model is as follows 6:

𝛿𝛿2
𝐸𝐸t = 𝐸𝐸o𝑒𝑒 (𝜇𝜇− 2 ) 𝑡𝑡+𝛿𝛿𝛿𝛿√𝑡𝑡 [2]
Time step used for period 1 is assumed to be 0.5, which is consistent with the mid-period convention.
Time step for period 2 is 1, i.e., the length between mid-period of the first period or 0.50 and the mid-
period of the second period or 1.5, which, again is consistent with the mid-period conversion.
Application of GBM model
Appendix A, Application of GBM Model- One Trial illustrates use of GBM model for one trial, i.e., one set
of random normal distribution numbers for period 1 and period 2.
We already described how first three categories (i.e., Expected EBITDA, Payment Period and Mid period)
shown in the Appendix were determined. We also described how time steps were determined for period
1 and period 2.
Let us focus on other calculations for each period.
Period 1 calculations
Present value (PV) of expected EBITDA is calculated as follows:

PV of EBITDA year 1 = Expected EBITDA ∗ e−0.5 ∗ EBITDA Discount Rate

Use of e, i.e., the base of normal logarithm is part of a continuous discounting method, i.e., the
discounting performed when the number of compounded time periods approaches infinity. Use of 0.5
reflects the mid-period assumption.
Formula [2] was applied to calculate Geometric Brownian Motion (GBM) as follows:
𝛿𝛿 2
𝑒𝑒 (𝜇𝜇− 2 ) 𝑡𝑡+𝛿𝛿𝛿𝛿√𝑡𝑡
GBM was then multiplied by PV of EBITDA for period 1 to determine risk-neutral random EBITDA.
After Risk neutral random EBITDA is determined, payout for year 1 is calculated as follows:
Earnout payment = Percentage pay-out * (MAX(Risk neutral random EBITDA year 1 - Min EBITDA
threshold year 1,0)- MAX(Risk neutral random EBITDA year 1 - Max EBITDA threshold year 1,0)).
Essentially, the above formula calculates the actual payout as a positive number not exceeding the cap.
According to the terms of the example, percentage pay-out is 20%. Minimum EBITDA threshold for year
1 is $ 1,000 K. Payout cap for year 1 is $ 200 K. Therefore, maximum EBITDA threshold for year 1 is $
2,000 K or $ 1,000 K + $ 200 K/0.2.
Earnout payment is discounted from the expected payment period to present as follows:
′
PV of Payout = FV Payout ∗ e−1.25 ∗�risk free rate+company s credit risk�
Coefficient 1.25 reflects the payment period being 1 year, 3 months after the present.

6
See formula 21.17 in Chapter 21.6 Monte Carlo Simulation, John C. Hull, Sankarshan Basu, Options, Futures, and
Other Derivatives.

7|P age
Use of the discount factor without regard to the systematic risk associated with EBITDA is consistent
with risk-neutral valuation approach.
Period 0 EBITDA- Additional Considerations Note that risk neutral random EBITDA was calculated for
period 1 using present value of EBITDA for the same period. However, same results can be obtained
by using EBITDA for period 0, i.e., the current period. In this case, drift 𝜇𝜇, as used in formula [2] is
calculated using EBITDA in period 1 as compared to period 0, i.e., LN(EBITDA1/EBITDA0) as well as
EBITDA risk premium, i.e., EBITDA discount rate less risk-free rate. Specifically, instead of 𝜇𝜇*t used in
formula [2] , the following calculations are used: LN(EBITDA1/EBITDA0) – (EBITDA discount rate - Risk
free rate)*Time Step.
Let us try to explain the relations between the two approaches. When no EBITDA0 is used, effectively,
EBITDA1 is used as a replacement for it. In this case, LN(EBITDA1/EBITDA1) equals 0. There is no need
to apply EBITDA discount rate because EBITDA was already discounted to time 0 to calculate present
value of EBITDA. Therefore, when using EBITDA1 only, formula LN(EBITDA1/EBITDA0) – (EBITDA
discount rate - Risk free rate)*Time Step is equivalent to Risk free rate*Time Step or 𝜇𝜇*t.
One can see that using EBITDA for period 0 produces the same risk-neutral EBITDA as using present
value of EBITDA for period 1 in the example of Monte Carlo simulation provided in Valuation
Advisory# 4- Valuation of Contingent Consideration 7.
Period 2 calculations
Year 2 calculations are similar to those performed for year 1.
Present value (PV) of expected EBITDA is calculated as follows:

PV of EBITDA year 2 = Expected EBITDA ∗ e−1.5 ∗ EBITDA Discount Rate

Same Geometric Brownian Motion (GBM) formula applies except for time-step in year 2 equals 1 and
year 2 has its own random standard normal distribution numbers.
Risk-neutral random EBITDA for year 2 is calculated as follows:
GBM for year 1 * GBM for year 2 * PV of EBITDA year 2 [3]
The same result can be archived as shown in formula [3] by using PV of EBITDA for year 1:
Expected EBITDA year 2 δ2
LN( )−(EBITDA Risk Rate− μ)t− 2 t+δϵ√t
Risk neutral EBITDA year 1 ∗ e Expected EBITDA year 1

These calculations are consistent with those used to determine risk-neutral random EBITDA for year 1
with the use of EBITDA for year 0.
Earnout calculations for year 2 are performed taking into account year 1 shortfall that can be carried
forward to second year. First year short fall increases second year payment, provided EBITDA in the
second year can accommodate the increase.
Year 1 shortfall is calculated as follows:
Catch-up adjustment = Payout Cap in year 1 – Earnout payment in year 1
Maximum EBITDA threshold for year 2 is then calculated as follows:

7
See Example 9.10 on pages 102-104 Valuation Advisory# 4- Valuation of Contingent Consideration. Specifically,
the same results are shown in lines [11] and [12] on page 104.

8|P age
Adjusted maximum EBITDA = Min EBITDA threshold year 2 + (Catch-up adjustment + Payout cap for year
2) / Percentage pay-out.
Min EBITDA threshold for year 2 is $ 1,400 K. Payout cap for year 2 is $ 200 K. Percentage pay-out is
20%.
The earnout for the second year is then calculated as follows:
Earnout payment w/ catch-up= Percentage pay-out * (MAX(Risk neutral random EBITDA year 2 - Min
EBITDA threshold year 2,0)- MAX(Risk neutral random EBITDA year 2 - Adjusted maximum EBITDA,0)).
Similar to year 1 formula, year 2 calculates the actual payout as a positive number not exceeding the
maximum threshold. The maximum threshold is determined considering year 1 shortfall.
Earnout payment is discounted from the expected payment period to present as follows:
′
PV of Payout = FV of Payout ∗ e−2.25 ∗�risk free rate+company s credit risk�
Coefficient 2.25 reflects the payment period being 2 years, 3 months following the valuation date.
Estimated fair value of payout is determined as a sum of the PV of payout in year 1 and year 2.
One trial covering period 1 and period 2 in Appendix A resulted in estimated present value of the payout
to be $ 80.59 K.
Multiple Trials
The above calculations for period 1 and period 2 are repeated significant number of times, e.g., 10,000
times. Total estimated fair value of payout is determined as an average of all individual calculations.
For the above example, total estimated fair value of payout is ap. $ 40.6 K. You will note that the exact
amount will vary slightly as it depends on the generator of random numbers.
Appendix B: Application of GBM Model- 10,000 Trials presents an extract of applying GBM model
10,000 times. Note that first trial in Appendix B is the same as the trial shown in Appendix A.
Appendix C: Random Normal Distribution (0,1) presents an extract from random normal distribution
numbers generated to support 10,000 trials. Note that random numbers for first trial in Appendix C are
the same as per the trial in Appendix A.
Using Microsoft Excel
Note that 10,000 trials can be performed in Microsoft Excel. However, some valuation specialists use
Python programming language or other standard software packages.
Practically, we found it convenient to generate random normal distribution numbers in a separate excel
sheet. Normal distribution numbers were then linked to core calculations of the payout for year 1 and
year 2 using Microsoft Office INDEX formula. Core calculations also link all key terms and assumptions as
shown in Appendix A.

When using Microsoft Excel function NORM.INV(RAND(),0,1), one should bear in mind that making any
changes in the workbook and saving it, results in regenerating cumulative normal distribution numbers,
which changes the result of valuation. When simulating random numbers for a large number of
observations (say, 10,000), the changes in valuation are insignificant. Changes are likely to be more
substantial as the number of observations decreases. To fixate final valuation results, a valuation
specialist may want to use COPY and PASTE SPECIAL/VALUES function in Microsoft Excel in relation to
random numbers.

9|P age
While random numbers are not fixated, a valuation specialist can validate the statement that final
valuation results do not change significantly for various combinations of random normal numbers.

Other Examples
Let us provide two more examples of where GBM model and Monter Carlo simulation can be used. As
before, we focus on valuation examples relevant to financial reporting.

Example B: Employee Bonus

General terms of the example below are the same as in first example except for the payment is due to
company’s employees determined in accordance with company’s bonus plan.

According to the terms of the bonus plan, Company A, is required to pay employees 20% of the excess of
Company A, annual EBITDA above $ 1,000 K for the first year and 20% of the excess of annual EBITDA
above $ 1,400 K for the second year. First year payment cap is $ 200 K. Second year payment cap is $
250 K.

If the payment cap in the first year is not reached, then any shortfall, i.e., the difference between actual
payment and the first-year cap will be added to the second-year cap.

The payments are due three months after the end of each bonus year.
The above bonus plan is considered a profit-sharing arrangement as compared to share-based
compensation arrangement. This is because a grantee typically is only able to participate in the
company’s profit measure, i.e., EBITDA while providing services to the entity, and a residual interest
(equity) is not retained upon termination 8.
Profit-sharing plans are accounted for under ASC 710-10 or ASC 450. According to 710-10-30-1,
compensation amounts “equal to the then present value of all of the future benefits expected to be
paid”. The entire bonus amount is accrued over the period of the employee’s service (e.g. a year) in a
systematic and rational manner (ASC 710-10-25-9).
ASC 450 measurement principle refers to a better estimate than any other amount within the
range, which, if exists, should be recognized as a liability and expense. When no amount within the
range is a better estimate than any other amount, the minimum amount in the range shall be
accrued (ASC 450-20-30-1). None of the two standards, i.e., ASC 710-10 and ASC 450- 20 require the
use of fair value concept under ASC 820.
In practice, bonus liability is often times determined using “expected” value, which, can be different
from the “fair value” as defined in ASC 820 or fair-value-based measurement per ASC 718.
Although accounting requirements applicable to profit-sharing plans do not prohibit the use of
Monte Carlo simulation, they do not prohibit the use of Monte Carlo either. If Monte Carlo
simulation was used to determine the amount of bonus liability, total amount of expected liability
at the valuation date would be $ 40.6 K, i.e., the same as determined in Example A. This liability and
compensation expense would be recorded as the bonus is earned over the term of 2 years,
generally, on a ratable basis.

8
Section 2.6 Profits Interests, Deloitte’s Roadmap to Accounting for Share-Based Payment Awards provide a good,
comprehensive analysis of how to differentiate profit sharing plans from share-based compensation plans.

10 | P a g e
Monte Carlo simulations would have to be re-run during the 2-year term to determine revised
bonus amount. Updated simulations would use shorter time steps, i.e., amounts t1 and t2 per
Appendix A, reflecting passage of time. Frequency of updated simulations will depend on
company’s financial close, which is generally performed on a monthly or quarterly basis.
A simpler approach used to determine employee bonus will satisfy GAAP reporting requirements as
well. Specifically, a reporting entity may use “expected” EBITDA amounts for year 1 and year 2 to
estimate the bonus amount. In this case, using information provided in Example B, estimated bonus
would be zero. This is because bonus thresholds were established at the level of expected EBITDA.
Bonus calculations would be updated as part of company’s financial close. Specifically, increase in
updated expected EBITDA for year 1 or year 2 will increase the amount of bonus. In this case,
calculations similar to those shown in the bottom of Appendix A, i.e., below line-item “Risk-neutral
Random EBITDA” will be performed.
Regardless of which approach is followed, a reporting entity will have to determine if bonus will be
recognized separately for each year using respective year’s estimate or estimated combined bonus
(i.e., year 1 plus year 2) will be recognized over the period of two years. Allocation methodology as
it applies to individual years will likely impact amount of bonus expense recorded in each reporting
period, i.e.., individual year, a quarter or month.
Note that if the portion of the arrangement is to be settled in equity or if the settlement amount is
based, at least in part, on the price of company’s shares, generally, the arrangement is in the scope
of ASC 718. ASC 718 requires the use of fair-value-based measurement, which in many cases will be
similar or ASC 820 fair value measurement.
Example C: Share-Based Compensation
Our last example covers a stock compensation plan, i.e., a compensation plan linked to changes in
stock prices.
A public entity offers a bonus plan to certain of its employees. At the beginning of a two-year
period, a target cash bonus, being a specific dollar amount is established for each employee. Each
employee will receive 50% of a target bonus at the end of two years if total return on public entity’s
stock price is between 10 and 15% and 100% of a target bonus if the return exceeds 15%. No bonus
is received if return on the stock is under 10%.
Because the bonus is settled only in cash, reporting entity’s obligation under the bonus is classified
as a liability (ASC 718-10-25-11(b)). Since the liability is based, in part, on the price of reporting
entity’s shares, the bonus plan is within the scope of and is accounted for in accordance with ASC
718. Under ASC 718-30-35-3, the reporting entity “shall measure a liability award under a share-
based payment arrangement based on the award’s fair value remeasured at each reporting date
until the date of settlement.”
Future changes in stock can be modeled using either discrete or logarithmic GBM formula. Note
that Monte Carlo simulations are performed in a risk-neutral world. Therefore, for a non-dividend
paying stock, expected stock return equals risk free rate. Volatility can be estimated using historical
information for the public company or other appropriate techniques.

11 | P a g e
Application of Monte Carlo simulation to estimate fair value of stock compensation would follow
steps as described below 9:

1) Sample a random path for S, i.e., stock price in a risk-neutral word for 2 years;
2) Calculate the payoff from each path;
3) Repeat steps 1) and 2) to get many (e.g. 10,000) sample values of the pay-off;
4) Calculate the mean of the sample payoffs to get an estimate of the expected payroll;
5) Discount the expected payoff at the risk-free rate plus reporting entity’s credit risk;

Note that estimated future stock price (S1) is determined using the stock price as it exists on the
valuation date (S0).
Calculation of payoff as part of step 2) will be based on projected stock increase as compared to
contractual thresholds of 10 and 15% and respective bonus targets set in specific dollar values.
Monte Carlo simulation will have to be updated at least on a quarterly basis, i.e., as part of public
entity’s quarterly close until the end of second year. Each simulation will have updated inputs,
including risk-free rate, entity’s credit risk, estimated volatility, remaining time steps, company’s
stock price S0.
Options, Futures, and Other Derivatives by John Hull and Sankarshan Basu provides an example of
projecting future stock prices using discrete GBM formula10.

9
The process is described on page 496, Chapter 21.6 Monte Carlo Simulation, John C. Hull, Sankarshan Basu,
Options, Futures, and Other Derivatives.
10
page 330-331, Chapter 14.3 The Process for a Stock Price, John C. Hull, Sankarshan Basu, Options, Futures, and
Other Derivatives.

12 | P a g e
FINACCO CONTACTS:

Nick Larchenko, Managing Partner

646.713.4764 / [email protected]

Denis Podshivalenko, Director- Accounting, Valuation Advisory

[email protected]

© 2023 FinAcco. All Rights Reserved.

_____________________________________________________________________________

Important Note: This publica�on is for informa�onal use only. Nothing in this publica�on is
intended to cons�tute legal, accoun�ng or valua�on advice. Consult accoun�ng or valua�on
professional concerning specific situa�on or ques�on.

13 | P a g e
Appendix A: Applica�on of GBM Model- One Trial
Appendix B: Applica�on of GBM Model- 10,000 Trials
Appendix D: Random Normal Distribu�on (0,1)