The "scientific" part of private firm valuation is really corporate finance. As you know, corporate finance, in its theoretical and practical applications, assumes or relies on the existence of "listed" firms. This is why, to understand private firm valuation, one needs to understand corporate finance and apply the concepts, in as much practicality as possible, to private firms. For example, in corporate finance one pillar is the CAPM. Of course we can apply it to private firms that do not have stock prices......

Good Luck,

I think you'll gain more by breaking it into a number of pieces. It's important to divide private companies into two categories, mature vs high growth. Venture capital valuation will differ substantially from private equity valuation in practice. You'll notice that the science falls off a cliff in VC. Beyond those categories, you'll want to dig into traditional control premiums, leveraged buyouts, etc.

Hope that helps a bit.

Let say that you restart from scratch. For listed companies, it exists a valustion that you do not have to qualify as a researcher, which is the market value. Then you can study the relationship between this market value and some of the caracteristics of the firm. That had been discussed in the literature but is far from being settled. What is relevant in the market price of a stock?

Then you can apply the algorithm you will have derived from this study to unlisted firms to see if it works. You take unlisted firms that have been sold recently and you compare the actual price with the results of your model. Seen like that it is not a consultant job, but a scientific research.

The private companies can offer important study perspectives on the corporate governance topic. Good luck!

Dear Florin Nijazi Al

I think in private companies many topics related with this subject , such as APT theory , you can apply in listed firms in stock exchange market .

Best Regards, 

Seems to me like a excessively wide topic. there are quite a lot of specific issues that may require research. Valuation is a huge topic. I suggest to ask your supervisor to narrow the field. You should chose either specifics of valuation in a specific country and industry, or chose one of the existing valuation approaches and improve on-it.

There are a number of underlying theoretical issues you can consider.  For example, estimating cost of equity for firms that face unsystematic risk.

Since I value private companies for a living, let me wade into this. 

@ Florin: I'm sorry, but I don't understand your question. What do you mean by 'scientific'?

@Ghassan: You're on the right track. Certainly corporate finance and the (modified) CAPM matter. One has to use the modified CAPM because private firms have no 'beta'. Therefore, unsystematic risk (USR)--which modern portfolio theory (MPT) assumed away the existence of--comes to the fore. USR comes in four forms:

When MPT made the issue of USR moot because it can be eliminated through low correlations among stocks in a portfolio, it also precluded future research into USR, the development of propositions, hypothesis-testing, and the like. That's fine for MPT, but it's hell for those of us who make our livings valuing non-public entity.

When I got started in this line of work more than two decades ago, I saw that dealing with USR issues was taking 65-70% of my time on each project. And it was doing that because no framework for USR existed. I tried such tools as Porter's Value Chain, McKinsey's 7-S Framework, and Galbraith's Star Model, but either they were not intuitive or clients found them difficult to deal with. I needed something simple and intuitive so I set out to come up with a framework.It took me a dozen years, but finally, in 2005, I developed the Trilevel Framework of Unsystematic Risk. It is the linchpin of the book I wrote for Wiley in 2010: Value Maps: Valuation Tools That Unlock Business Wealth (link: http://bit.ly/ValueMaps). I backtested it, and have continued to backtest it, and it has stood up with no 'hiccups' (i.e., no exceptions).

The tri-level framework relies on five (5) disciplines that are distinctly 'non-traditional' in business valuation:

In fact, I devote a full chapter (Chaps. 2-6) to each of those disciplines. Each has 'tools' that we who value the equity in non-public companies can use. At the end of the day, it's all about identifying cause-and-effect relationships in the 'outlier performance metrics' of the company being valued. That process begins with the creation of an industry composite from public companies and culminates with one-on-one interviews with company personnel near the end of the valuation process.

@Charles: I agree w/you that VC valuation falls off a cliff where science is concerned. But it does so because VC valuation has more in common with astrology than valuation per se.

@Gaétan: There's not just one 'market value' of a private company. There are several dozen, but, for our purposes here, let's stick with these four: minority-nonmarketable value, minority-marketable value, control-marketable value, and strategic-control value. The latter comes into play in M&A when the buyer is in the same industry. For a graphic of this concept, click here: http://bit.ly/Levels-of-Value and then scroll down to the second picture.

In addition, your suggestion to "take unlisted firms that have been sold recently and compare the actual price with the results of your model" will encounter several obstacles: 

As a result of these shortcomings in 'transactional databases' for private companies, I have used this method as a primary indicator of valuation exactly one (1) time in nearly 500 valuation and advisory engagements in the last 23 years. And that one time concerned a dental practice in a high-dollar divorce: I found a database that had more than 1,000 transactions involving the sale of such practices. I could slice and dice the transactional data in order to control for size (measured by annual revenues), geographical area, and time frame. That left me with enough transactions (more than 30, as I recall) to be comfortable in using this method as a primary indicator of value. But it is the exception.

@Bradut: I respectfully disagree with your recommendation that Florin "improve" on one of the valuation approaches. What's to 'improve'?? There are three such approaches--cost, market, and income. Within each of those, there are methods. Both approaches and methods are well-established in the United States. They began with a government pronouncement in 1959: Revenue Ruling 59-60. You can find it in its entirety here: http://bit.ly/RevRul_59-60. Section 4 ('Factors to Consider') lists eight primary factors. If one reads them closely, one sees that all three valuation approaches are included.

Knowledgeable valuation professionals typically use two (2) primary methods in their work: (a) discounted cash flow (DCF) under the income approach, and (b) guideline-public-companies (GPC) under the market approach. The challenge in both methods is gauging risk, esp. USR. In DCF for a typical private company with, say, $3-10 million in annual revenue, the equity risk premium is usually north of 20%. But one has to quantify each of the four levels of USR to get there. Done properly, these two methods produce estimates of value that are within 10% of one another. If the difference is > 10%, then the professional has made at least one major mistake somewhere along the line.

Where GPC analysis is concerned, one uses public companies "in the same or a similar business." Similarity is defined very broadly, per a U.S. Tax Court estate-tax case called "Estate of Joyce C. Hall v. Commissioner (1989)"; that case is here: http://bit.ly/Estate_of_Joyce_Hall. Besides identifying public companies that are broadly similar to the company being valued, one must also compensate for differences in size-related risk between large public companies and a small private one. It would take far more time than I have available to explain the complexity of that process in this venue. 

Since RR 59-60, many books have come out about private-company valuation. The two primary reference books in the field are Valuing a Business (5th Edition): The Analysis and Appraisal of Closely Held Companies by Shannon Pratt and Alina Niculita (2008) and Financial Valuation (3rd Edition): Applications and Models by James R. Hitchner (2011). The Pratt-Niculita book is excellent; however, it has a terrible index and is also long in the tooth. The Hitchner book is also excellent and comes with a top-flight index, which, IMHO, is essential in any reference book. Its 4th edition, which is due out in April 2016, will be a notable improvement.

Anyone who would like more information about valuing private companies or who would like to see the graphics I've described above should contact me directly: [email protected].

I hope these comments are constructive, illuminating, and helpful.

See J. Morgan 'Private Equity Finance: Rise and Repercussions,' Palgrave Macmillan, 2009

If value exists and is something different than price is a philosiphical thing. If you conclude that exists then for the estimation you can use scientific tools like statistcs, maths, finance etc. Good luck!

@Jamie: The concept of 'private equity' and of 'valuing non-public companies' are two entirely different notions. The latter is the subject of this discussion. The former refers to firms that invest in non-public companies, typically by buying them, adding them to a portfolio of similar companies, combining that portfolio into a single company, and then taking them public in order to cash out. 

@Warren

My point is that the topic "Valuation of Private Companies" is to wide for a PhD study. As such the PhD researcher will not be able to cover the topic within the PhD study duration. I can imagine a great number of topics within this topic that may be subject to a PhD research.

I am fully aware of the complexity of valuation you mentioned, this is why i suggest for the topic to be narrowed down. Paraphrasing something you said, it would take more than a PhD study time availability to explain the complexity of that processes in this venue.

There is a basic fact that showcases the complexity of the issue, Accounting IS valuating all private companies, yet there is a need for a professional valuation outside-it. Given the complexity of accounting, something that need to improve on-it is surely bound to be more complex.

Valuation professionals are bound to some methods and approaches because of legal liability issues, while accountants and auditors act upon other legal liability issues. They are all professionals, they all use highly evolved methods, yet they may all generate completely different values.

Even among evaluation professionals, there is a certain degree of specialization.

How about countries that have different fiscal systems, various market systems, different levels of information disclosure, different liability issues for the professionals?

It is to much for a PhD study to cover all of this within a decent amount of time, information and resources available.

@Florin,

1. Science is positively allowed to be practical.

2. The procedures and steps you have read are designed to cover the professionals against legal responsibility. The methods and assumptions underlying the procedures are the real deal for scientific research. 

3. As a scientist, you are not bound to valuation professionals procedures and assumptions. As a scientist you are bound to use a sound scientific method in order to determine something. You should study the professional standards, and know them very well, and than find their weaknesses, and work from there on.

4. Given the size of your topic you may end up being caught in a never-ending loop of development.  Take a simple segment of corporate valuation, land valuation, and you will find a huge number of articles on this topic, and it's the most basic asset of all.

5. while professionals operate with bounded rationality, you have to operate on a scientific rationality. What is good enough for a professional, should not be good enough for a scientist. However, it is unlikely you will be able to achieve this by being focused on the whole issue of evaluation,  you need to zoom in to some specific issues. Personally, i consider "Brand evaluation" or "Land evaluation" or something like this a good focus level for a PhD study, but you can also narrow down based on country level, type of country level, social model, market type, market environment, and many other points of view that may fit your expertise. 

6. The key issues when choosing and planning a PhD is a good match between the topic, your expertise, your interests, available data, supervisor expertise and interests.

I agree, valuation of private companies is mostly not a valid scientific enquiry. However issues surrounding this valuation would greatly enhance the literature in Corporate Finance. This could be around Misvaluation of Private companies, IPOS and financing of mergers by private companies and others and a specific examination of valuation issues affected by culture industry etc. However, if you need  a profoundly scholarly subject within Corporate Finance , Valuation is not that. 

Thank you, Amit. I agree with your observation that more about private-company valuation "would greatly enhance the literature in corporate finance."

Here's the problem: Getting the data to make a meaningful contribution to that literature is very, very difficult. Unlike, say, financial data on public firms, easily accessible through EDGAR and a number of for-a-fee private vendors, the same data on private companies is hard to get. It is also likely to be of dubious quality because, at least the in the States, few non-public companies are audited. That is compounded by the fact that many--I would wager, most--owners of private companies would rather minimize their taxes than maximize their profits. So, for any meaningful analysis to take place, balance sheets and income statements must be 'normalized' and adjusted. That process will be different for every private company.

Finally, grappling with private-company valuation would surface the huge problem that resulted when modern portfolio theory assumed away the existence of unsystematic risk. In order to value a middle-market or lower-middle-market private company accurately, one has to get a handle on unsystematic risk, not just conceptually, but for each private company in a database.

I'd love to see it happen, but I cannot imagine that it ever will. I hope to goodness I'm wrong!

I agree with you on the practical side of all that. But I would like to start one step behind. Market values are supposed to reflect the value of the firm. All the theories of the market says that, including Fama and others. So, if we have examples of prices which are faithful reflects of the value of firms, it might be possible, by studying public firms to find a recurrent relationship between firms'characteristics and their prices. If not, all that is pure bull shit.  If such recurrences are found, they can lead to a model useful to valuate firms that are not listed.

If we conclude that it is impossible it must be because market prices are not reflecting the potential of firms in a systematic manner. It is quite possible.

Hmmmm. I disagree. Let me explain why.

In the first place, I believe that your statement about prices being "faithful reflects of the value of firms" assumes that markets are efficient. The empirical evidence is that markets are merely semi-efficient. Price, therefore, may not be a perfect reflection of value.

Second, I believe your assuming that there is just one (1) value of a firm. There isn't. There are millions of possibilities. What the value is depends on a number of factors. In the United States, those include the following:

I could go on, but you get the idea, Gaétan. Heck, if you consider only the possible combinations in the first four factors I listed above--6 x 4 x 5 x 51--you get 6,120 possibilities. Add in #5 and #6, and the possibilities are endless.

Finally--and I say this respectfully--I'd like to quote the late H.L. Mencken (https://en.wikipedia.org/wiki/H._L._Mencken), also known as 'The Sage of Baltimore'. He once wrote, "For every problem, there is always at least one solution which is quick, easy. . .and wrong."

It's too bad that Mencken died long before you had a chance to meet him, Gaétan. But, you can do the next-best thing: read some of his books. He wrote six volumes entitled "Prejudices." Any one of them will get your attention, but the first one. is especially good. Published in 1919, it is as relevant today as it was then. Here's a link: http://www.amazon.com/Prejudices-first-H-L-Mencken/dp/B0153DP3CS/.

Closing comment: I believe you're looking for a perfect solution to a vexing problem. Such solutions do not exist. There area good ones, though, and that is why I hope you will keep in mind that the perfect is the sworn enemy of the good.

In my opinion, the origin of this debate rests on the fact that the current theories of valuation have their bases in capital market models, especially the CAPM. As you leave this "market place", you are thrown back to subjective goals and restriktions. In the search of superindividually accepted principles, you must look for rules that may not as hygienic as the implications of the CAPM, but at least not to dirty. Regularly, you have to create heuristics, whose distance to theoretical grounds cannot be measured because this distance is caused by an accepted violation of the model in question. This problem seems unsolvable, but theorists should try to solve it, because a surrender would just give room for cheap talk of practitioners who see themselves confirmed in their opinion that valuation theory is senselees for practical purposes.

@Bradut Vasile Bolos: Labeling a PhD student a 'scientist' strikes me as a stretch. I mean no disrespect to Florin or any other PhD students gathered here. (I was once a PhD student myself.) A PhD student today, after all, can be someone who was unemployed yesterday. Again, I mean no disrespect. A PhD student is just that: a student--until she or he passes comprehensive written examinations, selects a dissertation topic, and then successfully defends the proposal underpinning that selection in front of faculty members. At that point, I think that a fine case can be made that the person is indeed a scientist. But not until.

@Mittal: I agree generally. However, I believe that a dissertation that focused on "unsystematic risk [USR]" would be a potentially great contribution to what we know about finance and valuation. MPT's assumption of diversification effectively precluded research into USR, and that has done no one any good whatsoever.

@Leonard Knoll: You wrote, in part, "Regularly, you have to create heuristics, whose distance to theoretical grounds cannot be measured because this distance is caused by an accepted violation of the model in question. This problem seems unsolvable, but theorists should try to solve it, because a surrender would just give room for cheap talk of practitioners who see themselves confirmed in their opinion that valuation theory is senselees for practical purposes."

Where did you get your idea of "practitioners who see themselves confirmed in their opinion that valuation theory is senseless for practical purposes"? Would you provide some citations, please? Thanks in advance for your help.

@ Warren D. Miller:

First, my name is Leonhard, not Leonard. Second, I am Professor at the University of Würzburg, but also a free consultant. From that, I know many practitioners of that mind.  You will not find practitioners who declare this in journal articles or book chapters, but that is exactly the way in which they think, talk and act! Needless to say, there are also practitioners who are well trained in economic theory and models, but the smaller and "more private" the firm to be valued the greater the chance that you will  find people who are far away from economic theory.

@Leonhard: I apologize for misspelling your name. Mea culpa.

What you have is anecdote. As George Stigler once said, "The plural of anecdote is data." Where's your data, sir?

@Warren D. Miller:

Once again: The practitioners I think of don´t write in journals or textbooks. Maybe I start a talley in which I can mark statements of the kind you want to have as "data".

@Warren D Miller a comment on your words: " A PhD student is just that: a student--until she or he passes comprehensive written examinations, selects a dissertation topic, and then successfully defends the proposal underpinning that selection in front of faculty members"

Yes, true, a PhD student is a student, but the objective of the whole doctoral process is to acquire and prove the ability to do scientific research. Thus the thesis research MUST be a scientific research. As  he is not a proven researcher there is a thesis coordinator that is a proven researcher that oversees the process.

@Bradut: I agree with you 100%.

@Leonhard: The vast majority of those who hold PhDs have never published a single paper in a double-blind-refereed journal. These "scientists" either have no desire to publish or submit papers that get rejected for publication. The lack of post-PhD scholarly activity is, I think, a major reason that most PhD programs in the U.S. have changed from requiring a dissertation to requiring three "publishable" papers.

Based on what I've read, this change results from criticism and from close scrutiny from both legislative budget committees and big donors about the lack of research productivity exhibited by most holders of terminal degrees. Politicians and donors are rightly concerned about the high cost of PhD programs and their return on investment in the form of new knowledge. . .which is what research in the social science is (or should be) about.

Unfortunately, the publishing system itself works against science. Writing and article involves time, effort, expenses. Most publications do not reward the author, i can say quite the opposite, as they require payment for publication. Anyway, even if the journal would pay royalties, the amount is by no means close to the costs involved in generating the research. So, for most researchers the balance costs/benefits  is negative.

@Warren, you say: "The lack of post-PhD scholarly activity is, I think, a major reason that most PhD programs in the U.S. have changed from requiring a dissertation to requiring three "publishable" papers."

A small comment on this is required. Universities hunger for ratings may be the main reason, as doctoral students are a significant source of articles that can be counted as scientific output. Add citations of "master's" papers and authorship given without work and you may see a pretty dark picture. Actually obtaining a PhD is far less rewarding than people may think.

@Bradut: I appreciate your comments about universities 'hunger for ratings'. I hadn't thought of that, but I think you're undoubtedly right.

However, I disagree with your assertion--at least as it applies to PhDs in business disciplines--that "Actually obtaining a PhD is far less rewarding than people may think." Of course, it depends on how people measure what is "rewarding." If it's low stress after getting the PhD, I don't think anyone can beat it with a stick. If a scholar is even minimally productive, tenure will follow; short of having sex with a student in public at high noon, tenure means a job for life. 

And, at least in the business disciplines, where salaries for newly minted PhDs are north of USD100,000/year at most institutions, I suspect most people would see that as not only "rewarding," but also "highly rewarding." In fact, in the U.S. today, I'd wager that most Americans would kill to have a job with that much freedom, that little stress, and that much pay, all with no risk of losing one's job.

As for "the costs [incurred] in generating research," I understand that there are costs. First, of course, is time. But that is covered by a salary, isn't it? After all, 'teaching, research, and service' are the three activities that a faculty member's paycheck pays for. As for out-of-pocket expenses, the universities with which I'm familiar reimburse those, at least in the business disciplines. Even if they didn't, though, the pay in those disciplines is more than high enough that an unreimbursed submission fee can be an itemized deduction on one's personal income-tax return, at least in the United States.

Well, i was not specifically targeting US. US is a very special place, as research funding is flowing like nowhere else.

@Bradut: Your point is well-taken. I'm sorry to be so U.S.-centric, but it's the only system that I have any in-depth knowledge of.

following Good question

According to me the valuation of a company is a more of a scientific and less of a practical topic. As private companies undertake valuation when the company intends to sale it to a private body or a public firm or going to public with IPO or venture capital, the process of valuation has to be scientific.

The normalization of company's earnings and discounted methods that are used are all based on scientific foundation. Whatever method you may use it has a scientific logical step by step process and hence the valuation of a company is scientific topic.

Dr. K N Sheth, before I respond in detail, I hope you won't mind if I ask you two questions:

1. For how many years have you been valuing companies?

2. Would you please tell me your first name so I can do my part to make our 'conversation' more collegial?

And please call me 'Warren', if you don't mind.

Many thanks--

Warren