How to Use Financial Models in Startup Valuation?

Valuing a startup can be difficult because of the lack of historical data for the company and the industry they are in. Often, startups will have little to no financial history. They might even be attempting to create a new industry segment altogether. This creates a lack of inputs for financial models, making startup valuation quite challenging.

However, in this article, we will guide you in overcoming these challenges to deliver a solid startup valuation through financial modeling. We will discuss the meaning of financial modeling, why startups require special financial models, and how you can build a financial model appropriate for startup valuation.

What is Financial Modeling?

Financial modeling utilises spreadsheets to forecast a business’s financial performance. In financial modeling, historical financial data is treated as the input to forecast future financial performance, which is then used to value the company.

So, the variables in financial modeling can be classified as:

Analysts will typically narrow their focus on the three main financial statements, i.e. income statement, balance sheet, and cash flow statement. However, if more relevant data is available you must not ignore it.

Financial modeling is also useful for strategic financial decision-making. It can reveal how a company should raise capital, if it can grow the business organically or should consider an acquisition, and if any factors are hindering its progress.

You can also use financial modeling to identify assets that could be upgraded or should be sold off. If you are under-utilising a business segment, this can also be recognized through financial modeling. Such insights can help optimise capital allocation.

Why Do Startups Need a Financial Model Geared to Their Needs?

The main problem with using traditional financial modeling techniques with startups is that there is not much historical data available. So, it becomes hard to estimate future financial performance.

Hence, we must rely on advanced statistical techniques and the financial performance of other related companies to value startups. It would be best if you could find data for a company like the startup you are valuing.

For instance, if you are valuing a payment processing startup, you could use Stripe’s early days as an input. You will need to adjust the values for inflation. You also need to identify how the industry segment has evolved.

Since Stripe is already an established player in the industry, the new startup may need to invest more, and spend more on marketing, and it may have to wait longer to scale up. You also need to account for macroeconomic conditions like existing interest rates, and gross domestic product (GDP) growth.

Which Financial Modeling Techniques Should You Use for Startup Valuation?

You can use the following techniques to power up your financial models for dealing with startup valuations.

Monte Carlo Simulations for startup valuation

Since a startup’s financial performance cannot be estimated with certainty, it would be better to consider multiple scenarios instead of just one. We cannot be 100% confident that things will play exactly as per our model’s scenario because of the unpredictable variables involved. This is the idea behind Monte Carlo simulations.

When you take this approach, you will be running multiple simulations by assigning different values to the unpredictable variables (a.k.a. random variables) every time. Then, the results will be averaged to estimate the target variable.

Example of Monte Carlo Simulation

Suppose we have the following information about a company.

This company is seeking an investment that will increase the user growth to 25% in 2025, and 30% in 2026, and increase the growth in transactions to 10% in 2025 and 2026. We will also assume the per transaction cost will go down to $0.001 and the fixed costs will increase to $200. This should play out like this.

Now, there are two key variables here- user growth and growth in transactions per user. We will assume that these variables follow the normal distribution. So, first, we will find their average. We will also need to assume how much these variables can change for a standard deviation of 1.

This information can be summed up as:

So, for a standard deviation of 1, user growth will increase or decrease from the mean by 7%. Similarly, the growth in transactions per user will increase or decrease from the mean by 1% for a standard deviation of 1.

Based on this information, we can simulate numerous values for user growth and growth in transactions per user. Based on these simulations, we can simulate various sets of cash flows and net profits. These net profits can then all be discounted by a reasonable cost of equity to get the numerous values of ownership.

We will simulate cash flows up till 2031.

If we make 500 simulations where the cost of equity is 15%, the results will look like this:

This graph can be summarised as:

Benefits of Monte Carlo Simulations

A major benefit of Monte Carlo simulations is that we do not need to make perfect assumptions. Instead, we can leverage our understanding of how certain variables behave. Here, we have assumed a normal distribution for both variables. However, you should use the distribution that describes the variable best.

If you want to use MS Excel for computing such data, you can use pre-built formulas for Poisson, binomial, and exponential distributions as well.

Vector Autoregression (VAR)

Vector autoregression involves applying linear regression to variables that affect each other. In VAR, every variable is an input for the other variable. When you are valuing startups, you will need to forecast sales growth which will depend on industry growth and market share trends.

Since startups may sometimes try to create an industry segment for themselves, there will not be much data to calculate industry growth. In such cases, you will need to study the relationship between industry growth and macroeconomic factors like interest rates, inflation, unemployment, and GDP. If the startup’s product is aimed at consumers rather than businesses, you must understand the trends in social and demographic factors like population growth, age distribution, and income levels.

Macroeconomic factors and social and demographic factors tend to affect each other.

You will also need to consider technological factors like advancements in supporting technology which will depend on the industry’s growth. Advancements in technology will also depend on the startup’s investments in research and development (R&D).

Example of vector autoregression

Suppose you are dealing with three variables whose value can be determined using the values of all variables from the previous period. Let us name these variables X, Y, and Z. So, the regression model for these three variables will look like this:

XT = BA + XT-1 × BXX + YT-1 × BXY + ZT-1 × BXZ
YT = BB + YT-1 × BYY + XT-1 × BYX + ZT-1 × BYZ
ZT = BC + ZT-1 × BZZ + YT-1 × BZY + XT-1 × BZX

Here, T is the current period, and T-1 is the previous period. Every variable from the previous period is matched with a unique coefficient in each equation, and each equation has a constant or intercept.

You can create a model such that:

• Macroeconomic factors affect each other
• Social and demographic factors affect macroeconomic factors and vice versa
• Technological factors depend on industry size and investments in R&D
• Industry size depends on technological factors, social and demographic factors, and macroeconomic factors
• Investments into R&D depend on the company’s sales growth
• The company’s sales growth depends on industry size growth

How to effectively utilise VARs and Monte Carlo simulations?

In this process, from the third to the sixth point, you will be making guesses about relationships you do not have data for. You can make educated guesses about the coefficients and intercepts, and then, you can apply Monte Carlo simulations. Thus, you can create a solid model that gives a fair estimate of industry growth and how it impacts a company’s sales.

Count on Eqvista for Your Valuation Needs!

You must consider looking at the financial data of similar companies if extensive financial data is not available for a startup. It is no use looking at other startups since you will again struggle for abundant data. You should look at past financial data of established players and adjust it according to the current economic and industry conditions.

Since the input variables like expected sales growth can vary unpredictably, you should rely on summaries or averages of numerous simulations instead of trying to build a perfect financial model.

If a startup is trying to create a new industry segment, you can estimate the industry growth by creating a vector autoregression model, i.e. a model where every variable is determined with a set of other variables. This will help you simulate how macroeconomic factors, social and demographic factors, and technological factors affect each other as well as industry growth and in turn, startup valuations.

If you need assistance with startup valuations, reach out to Eqvista on this page. We specialise in valuing startups from all stages for various purposes.