Skip to main content
Loading Explore guides…
Startup Tech Valuation Methods · Acedra
Home
Explore
Business
Startup Tech Valuation Methods
Sign in to save
Chat
If YouTube asks you to sign in, open the video on YouTube.
Watch on YouTube
Reload player
Transcript
76 segments
Copy
Hide video
0:00
Today I'm going to talk about the valuation of startup technology companies. Before getting into business valuation, I was in both venture capital and investment banking before finally settling into a business valuation. My practice involves emerging technology companies, whether it's software, data analytics, artificial intelligence. I do a little bit of work in the medical device and pharma space as well, and I routinely teach classes to graduate students business people, as
0:28
well as fellow business appraisers. We're going to accomplish four things today. One is income based valuation techniques.
0:37
Ineffective, in credibly determining the value of a startup. It's often a methodology that is used, but it is not necessarily the best. In fact, I think it often leads to poor valuation outcomes. So instead, I recommend using guideline public company data and transaction data, which we're going to discuss today. I'm going to show you how you can use that data to determine the value of a startup in a way that's much more meaningful. And credible and reliable than the conventional techniques.
1:09
And I'm going to expose you to some other unorthodox valuation approaches as well. It's important to understand that valuing startups is hard. Startups, relatively speaking, are fairly new phenomenon. The reason why valuing startups is hard is because they're young. They have no operating history. Most companies, if they're mature have a basis of value that lies in their historical operating performance that can be extrapolated over a period of time in the future.
1:40
But with startups, we don't have that.
1:43
They're hard to observe. . The data can be hard to obtain, at least in the most conventional cases. Startups live in a world of uncertainty level in a world of uncertain markets, uncertain customers. Uncertain competition, uncertain regulation, you name it, it's uncertain. Two startups can be of a similar stage of development. They can be in a similar space, have even a similar technology, but due to other complicating factors, their values can present extremely differently.
2:12
And then finally, there's a random element. I think it is helpful to view startup valuation as binary in nature. Investors and founders don't really know if their startup is going to be successful for several years after the company is created and investments are made. And that makes conventional valuations very difficult and so we have to think about valuation differently. I mentioned, the income approach is most frequently used to value startups. That's the tried and true methodology.
2:48
What the income approach means is that you start with a series of projections cashflow that a company hopes to generate over a period of time. Now, generally that. That projected cash flow has a sharp upward trajectory. And some people call that a hockey stick path, which is why we have the image of the hockey stick here. And generally speaking, those forecasts or projections are aspirational in nature. It is mostly the product of the founder's imagination as to what
3:25
the future financial performance of the startup is going to look like. That inherently makes the valuation of a startup using the income approach.
3:35
At best, highly subjective and certainly subject to manipulation in an adverse case.
3:42
Now, another reason why the income approach is suspect is within the income approach, we apply something called a discount rate, and the discount rate refers to the return that is acquired by investors in order to accept the risk of investing in that particular asset. Now, we all know that startups are high risk, high promised return assets and therefore the risk can be very high. The required return can be very high. I have performed valuations of companies that had required
4:15
rates of return of over 100%. So great was the risk associated with them. We can approximate a good chunk of the discount rate, relying on empirical data, the discount rate is a function of five elements. The first element is the risk free rate, which is the rate of return that we would expect on a risk free investment. And as of the time of recording this particular presentation, the risk free rate is somewhere around four and a half percent, but it changes on a daily basis. Okay.
4:54
Now, we also incorporate what's called an equity risk premium, and the equity risk premium refers to the additional return that is required to convince investors to move out of risk free securities and into the broad stock market, say the S and P 500, and that rate right now is around 6%. So the expected rate of return for investing in the broad stock market in the United States is about 10 and a half % now, we also incorporate a size premium.
5:25
Many practitioners, including myself, believe that the market charges a risk premium or requires a higher rate of return for smaller companies because smaller companies have fewer resources available to them. They're likely to be not as diversified in their markets, in their product line and revenue streams. And therefore, we believe that we can observe a risk premium or additional required rate of return for companies that are small.
5:53
And the data right now indicates that risk premium is somewhere around 7%. So to invest in small companies in a broad area index the expected rate of return is now somewhere around 17. 5%.
6:10
Further, there's evidence that certain industries are considered more or less risky than others, and that's going to differ depending on what the market is doing at that particular point in time. Now, for this particular company, which I am presenting in a somewhat neutral format I'm saying that the industry specific risk premium for this particular industry is 4%. It doesn't necessarily have to be that number. It could be zero. I've seen it as high as 10 percent in some cases.
6:42
It even could be negative where there's an industry specific risk discount. If that industry is in particular favor by the financial market and all those four building blocks can be tied to specific empirically verifiable data. You'll notice that the big part of this buildup is. Associated with what's called the company specific risk premium, which refers to all the risks that are not captured in the prior four buckets of risk.
7:14
In other words, everything that's not captured in the 21 and a half percent required return up until this point. Unlike those other four buckets, there is no widely accepted empirical method for determining or even estimating the company specific risk premium for a company. That is the product of the analyst's informed professional judgment. Now, given what I see in the marketplace for early stage companies. Unless it's an artificial intelligence company, the required rate of
7:45
additional return for a company specific risk premium is about 30%. For startup technology companies, the discount rate is going to be somewhere around 50 percent or required return of 50 percent annually. What you may notice about that 30 percent is that because that company specific risk premium is the product of my informed professional judgment, and if you compare the 30% that is subjective to the 21.5%, that is verifiable, it means
8:19
over half of the input in the discount rate that drives the discounted cash flow model is determined subjectively. So you have subjective projections, which are now subjected to a subjectively determined discount rate, and you probably see where I'm heading that this makes the income approach, generally speaking, highly suspect as a valuation method when it comes to early stage technology companies. I prefer the market approach. The market approach is often ignored by analysts because the
8:54
data is not easily obtainable. I'm not sure that's true.
8:59
There's micro cap public company data that can be used. There's venture capital data. And there are other less known methodologies that are available. If you have any doubts regarding the viability of the market as an indicator of value for startups, consider that. In 2023, the angel investment market in the US alone was 18. 6 billion, consisting of at least 54, 735 angel deals. And there are over 420, 000 angel investors in the United States that we know of. Okay.
9:33
There are many others that are not reported and they're not necessarily reporting all their deals. So I think these are underestimates. In the database for angel and early stage investments called Crunch Base, there are over 1, 300 angel networks that are active. The average pre money valuation back in 2022 for angel deals was three and a half million dollars with a median amount invested of just under 340, 000.
10:00
So it's an active market. And with that much activity, there ought to be something observable that we can use. And of course, angel investments differ by sector. Angel investors, like all investors, have their particular preferences and comfort zones. And as you can see, a good chunk of investments goes into health care. Good chunk goes into software with lesser amounts going into industrial energy. And of course, biotech, it's a fair amount of investment as well.
10:29
The reason many analysts reject the application of the market approach in valuing startups is because you generally don't have market multiples. With established companies, we are used to seeing multiples of sales or revenue, profits, EBITDA, something of that nature on which to base evaluation. But startups usually don't have profit. They often don't have revenue. So in people's minds, it's easy to disregard and discard the market approach as a valuation technique.
11:01
However, Think if we approach market valuation differently, we can still obtain a very useful result. And that is we should be calibrating value to the total invested capital of the investment. And we can specify or ensure comparability of our company to others in the market. By matching up their industry, by matching up their size, their state of development, by matching up how much investment they've created and whether or not they're
11:30
generating revenue or profit, the mere existence of revenue, the mere existence of profit differentiates many startups. Now, public companies are often overlooked in valuing startups. And that's unfortunate because there are many publicly traded companies that are in effect.
11:49
publicly traded startups, and you can find them on bulletin board companies, pink sheet companies, the aim, the Tel Aviv stock exchange, the Toronto stock exchange, euronext, Nasdaq, and other exchanges as well. And there are literally thousands of startups that are being publicly traded. We just don't necessarily hear about them. Now the reason I like to use guideline public companies is first and foremost because the data is available.
12:18
Public company data is also relevant because under US GAP and international financial reporting standards, there is an explicit preference expressed for the use of market based methodologies, especially publicly traded companies.
12:35
Public company data is timely because public company data is updated to the minute. You don't have to worry about data being two years old or three years old or five years old like you do a transaction data.
12:47
And frankly, it's easy to understand. Guideline public company data has to be presented in a fairly uniform way for a given stock exchange. And it can be verified because that data is often, Subjected to a third party independent financial audit, which makes that data fairly transparent. So here's an example of how I would use guideline public company data in valuing a startup.
13:13
So say we have a subject company that's in the mobile app space. It's headquartered in Austin, Texas. Management has an average track record, total addressable market of a billion dollars, has a hundred thousand dollars of annual revenue, has a loss of two and a quarter million dollars of EBITDA. That is common and has a hundred thousand dollars of cash in the bank, needs to raise at least one more round of capital. And the technology is considered average for its stage of development.
13:42
We can actually go to the market and find guideline public company data. And these are actual companies that were publicly traded. You can see most of them don't have any revenue. None of them have positive EBITDA yet. They're able to trade in the market. And one of them, Roadships Holdings, the last time I looked, had a market cap of over 65 million, despite having no, no revenue and no EBITDA. That looks like a startup company to me.
14:08
So the market capitalizations represent the total equity valuation of the company as dictated by the market.
14:17
Now, here's an extra step. We need to separate the cash or subtract cash from the market capitalization because most startup valuations are expressed in terms of pre money value, meaning the valuation of the company prior to receiving investment.
14:34
So if we subtract cash in the balance sheet, which generally there is going to have to be disclosed as a public company, we can then determine the implied pre money value of the company. And you can see that we can do that for each of the companies presented here. And then the next step is to summarize that data. So we have a range of pre money value between a quarter of a million dollars and 65. 5 million dollars. Mean of 11 million median of about 5 million.
15:01
And I think the median is much more relevant than the mean, because it's less likely to be skewed by a super high valuation, such as 65 million. And I think it's helpful to put this information as a chart for reference. So we can see that our subject company maybe has a little bit more cash, a little bit more revenue than the peer group, but it's much less profitable. It's hemorrhaging a lot more cash.
15:26
So we make a subjective adjustment. Semi subjective, I should say, because we can do a little bit of analysis. We know that the revenue is above the median, but within one standard deviation. So it's above the median, but not terribly. So on the other hand, the EBITDA is below the median and outside of one standard deviation. So statistically speaking, our EBITDA loss is far greater than the public company a nd it has cash that's above the median, but also within one standard deviation.
15:56
So in my judgment, I think that the two pluses within a standard deviation, more or less offset the negative that's outside one standard deviation, I'm going to round it down to the, so I'm going to just accept the median pre money value. I'm going to round it down to the nearest million dollars because valuation is not so precise that we can say for certain that a company is really worth 5. 15 million dollars versus five million dollars and that's certainly
16:21
enough precision for the overwhelming majority of valuation cases. Now we also have A transaction method, and there are good reasons to use the transaction method. The data is not quite as available. My go to source for transaction data is CrunchBase Pro. It's 6 or 7K year, but it contains, I think the best data set for startups. It is as relevant as guideline public company data, because again, Gap and IFRS express an explicit preference for market based data.
17:00
I think it does a better job of replicating the market than the guideline public company data, because few angel investors rely on cash flow models, and they're also not looking at guideline public data necessarily, but they certainly are looking at investments into similar companies. And in my world, I like it because clients love it. Because when I perform a valuation based on transactions, clients love it because they gain actionable intelligence about their company and their competitors,
17:28
because not only do they see valuations of their competitors, not only do they gain insight into the valuation of their company, but they also see. How much money their competitors or at least some of their competitors have raised and they get to see who else is investing so they know who a target list might be for potential investors when they go on to raise their next round. So here's an example, our subject company has the same characteristics,
17:56
mobile app space, Silicon Valley, et cetera, et cetera.
18:00
If you want to read through this, just pause the video
18:05
and we can find transaction data. We can find a number of companies that are in the space. We can see that they sold between, it looks like 2 percent and 20 percent of the equity in their company. Most of these companies do have revenue. The transaction data set gives us a range of revenue and pre money valuation with revenue ranging between zero and 5 million per year and pre money evaluations between looks like 1. 5 and 9. 6 million.
18:36
Now again, as before, we're going to rely on the median on the median of the data set because it is less subject or less. Influenced by a significant outlier data.
18:49
The median revenue for the data set was half a million dollars. And that of our subject company was a hundred thousand dollars. There is a substantial difference indeed, the median revenue for the companies that showed up in the venture capital data set is five times that of our subject company. But if we go back to our summary data set here, the standard deviation is 2. 3 million, which suggests that the difference is not necessarily statistically significant.
19:17
However even though we're within one standard deviation of the median with our, with revenue, my experience with investors tells me that they would consider statistically or not, they would consider a difference of 100, 000 versus 500, 000 to be a material difference. And therefore, I'm going to adjust the subject company valuation from the median of the data set of 3 million down to 2 million. And so our two valuation methodologies give us a value of 5 million by using
19:48
public company data and 2 million by using data from venture capital transactions. These two are close enough that I think it's okay to simply average the two. Which of course results in the valuation of three and a half million dollars, which is pretty close to the valuation data aggregate that I referenced at the start of this webinar. Now, for technical reasons, we may need to be mindful of the level and basis of value, mainly to make sure that if we're using other valuation
20:18
methodologies, such as the discounted cashflow method, or even the net asset value method that we're comparing apples to apples and methodologies, such as the discounted cash flow method, assume a value on a controlling level. However, the valuations that are produced by these market methodologies, both the guideline public company method and the venture capital transaction method on a non controlling interest, because with the public company method, it's
20:47
assumed that these are quoted on a per share basis, and with the venture capital method, most investments in venture capital are on a minority basis. Venture capitalists, generally speaking, don't want to take control of a company. So if we do run into a situation where we need to compare the valuations by The market methodologies versus other methodologies. We need to be aware of this differential. So either the other methodologies need to be discounted for lack
21:19
of control, or we need to gross up our market based methodologies to put you at a control premium. So that again, we have an apples to apples comparison. And we may need to make this adjustment also because we want to make an adjustment for for different capital structures. Now, if we were to add a control premium, it might look something like this.
21:41
We just made a determination that the valuation based on the market method is three and a half million dollars which is a pre money valuation on a minority basis. So if we were going to make it or transform this valuation to a controlling basis, we might add something like a 30 percent control premium. Which if you run through the math means that on a controlling basis, the same company would have a valuation of 4. 55 million and you can, there aren't really great ways to estimate the
22:12
value of the, of a control premium. Generally speaking we consider the changes in value by companies that are being acquired wholesale on the public markets. It's not the greatest methodology but they haven't really thought of anything better. So generally speaking, I tend to see control premium of 30 percent in the market. This entire valuation analysis assumes that our company has an average amount of debt. But how can we be sure if our company has some amount of debt?
22:44
How do we know that's average compared to the market? And so we need to do is we need to adjust from an equity valuation, which is what these values are pointing to which is invested capital value based on the typical capital structure in the market and then subtract the actual debt. Of our subject company.
23:03
So for example, if our equity value on a controlling basis is now 4. 55 million, and we have research that shows the average capital structure in the industry is 85 percent equity and 15 percent debt. And for the sake of argument, our subject company is a million dollars in debt. So we're going to gross up to an invested capital value. So we can divide 4. 55 million by 85%, which means 5. 3 million on a on an invested capital basis, but our company
23:37
has a million dollars of debt, so we're going to subtract 1. 1 million dollars from 5. 3 million dollars. And so our adjusted equity value is 4. 35 million dollars, as opposed to 4. 55 million dollars I want to present some alternative valuation methods just so that you're aware that they exist. A methodology, which is not officially considered valid, but is used quite a bit and to some degree, not endorsed, but is at least taught
24:08
by the angel capital association is the risk factor summation method. And with the risk factor summation method a lso called a scorecard method, you create a scorecard of the most important characteristics of a given startup company, such as management stage of development, regulatory risk, production risk, sales, risk, capital risk, competition, risk, technology, risk, litigation, risk, and scalability. And you assign a score, positive or negative.
24:40
As to whether or not you think the subject company is above average, below average, about average, way above, way below. And then you assign a dollar value to the score. In this case, I've assigned arbitrarily 100, 000. This is about having a process rather than an empirically sound method. And here's an example. If we Take a subject company and we're assuming that each point, if you will, is worth a hundred thousand dollars, then if management is above average, well
25:13
above average, we're going to sign two points to it as it's pretty much the same stage of business as the peer groups. So no change there. Maybe it's exposed to higher regulatory risks than normal, say it's in the medical area or the data area, then we're going to subtract a point for that. Maybe there's no production risk or at least average production risk relative to the peer group. There's a little bit more sales risk because the company doesn't
25:38
have a lot of revenue, so it hasn't really validated the market yet. On the other hand, maybe the company's funding risk is lower than average. So we're going to give it a plus two score. And maybe that's a function of the management team being strong. Maybe they've had. A prior exit, which makes it much easier to raise money. Of course we might add a point for competition risk, meaning that the risk of competition is less than normal, perhaps because we have intellectual property
26:04
protection or first mover advantage.
26:08
We're assigning technology risk and litigation risk zero. So they're about we think they're about average there and then scalability of plus two. We think that this technology is fairly scalable, which we would expect from a mobile application. And so if our baseline value is two and a half million dollars. We have a score of plus five and a hundred thousand dollars per point We add five hundred thousand dollars to the base score of two and a half
26:35
million Which means that this risk factor summation or scorecard method yields a value of three million dollars
26:44
You can also use option theory to determine the value of a startup. And this can get very technical. So I'm not going to go through a real options model, but if you want to learn about real options modeling, if you'd like me to do a video on that, let me know in the comments below, if we get enough demand I'll be happy to make a video , but basically real options modeling considers multiple scenarios. So in, in the top scenario, we see that, an investment Initially has, and
27:13
this is a very simplified case by the way, but an investment of 1 million may trigger a bailout meaning the company's going to fail, so you might get a hundred thousand dollars back, or it may make enough progress that a second investment is going to be made of a million dollars and so forth until if the investment succeeds sufficiently, there'll be some sort of exit. And this is in contrast to a discounted capsule model where it assumes there's a static set of projections.
27:40
And the only mechanism by which to capture the uncertainty of achieving those projections is the discount rate. Again, pretty technical. I'm not trying to equip you to replicate this on your own, but rather just give you a sense that this does exist. There's another methodology out there called the venture capital method and the venture capital method basically says this if , you're going to invest X number of dollars now and then five years from now, some point in the future, five years, 10
28:10
years from now, when the company exits, you're going to get a big bucket of money later. Okay. What is the value of a 20 percent share of a company that might sell for a hundred million dollars in five years? So if we have an exit, that's going to take place in five years, we'll put a hundred million dollars in a numerator. And then in the denominator, we'll have one plus our required rate of return, 50%, which is similar to the rate of return that we discussed earlier.
28:38
And we're going to raise that to the fifth power, and that represents the compounding of the required 50 percent rate of return over a five year period. And if you work through that math, it's 13 point 2 million post money. So if 1 million is invested. Then the pre-money value, we'd subtract $1 million from the post-money value. So the pre-money value is $12.2 million, and a 20% share of the company is worth $2.44 million.
29:09
So in theory, that means it's a, that means it's a good investment because we're getting $2.44 million of value. In exchange for a million dollars of investment, but of course there's a lot of noise there because for a high risk startup over five years, literally anything can happen. And statistically speaking, the most probable outcome is an unsuccessful investment. That's why the returns are so high. So I think we've covered all of the objectives that we needed to today.
29:41
So I want to just wrap up with a few closing thoughts. Don't give up prematurely on the market approach, particularly public companies, if you're faced with a need to perform a valuation of a startup company. It's also important to understand that the early stage valuation thesis is not purely a financial one. Angel investors make investments. Partially for financial reasons, but also because they see it as an opportunity to give back to to pay things forward after their success.
30:11
Sometimes they do it as a, as almost a very expensive hobby. Sometimes they do it because it gives them access to a network of other high net worth individuals. The reality is, in, in this kind of environment an academic environment can make sure that I come up with sufficiently comparable companies for our for our analysis, but the reality is that you may not be able to identify a sufficient number of public companies and or venture capital transactions and enough data
30:42
associated with them in order to draw reliable and credible value conclusion, in which case you may need to go back and rely upon the discounted cash flow method or some other methodology.
30:52
Finally, I think you can use these methodologies at least to assist in intellectual property valuations because early stage companies, are effectively valued on the basis of their intellectual property. If a company has a particularly strong management team, then I think that's part of the equation as well. But if you have a startup tech company and chances are the value thesis is technology based, that is going to be for the most part an expression of the value of the technology.
31:19
I don't know that anybody else is necessarily thinking that way, but I do think that the logic holds.
31:26
And so with that, I would like to thank you very much for your attention.
Here to help you learn
Sign in to keep a private conversation with your course tutor.
Ask me anything about the material…
Sign in to chat