Fairness Opinions: Fix them or Flush them!

My post on the Tesla/SCTY deal about the ineptitude and laziness that Lazard and Evercore brought to the valuation process did not win me any friends in the banking M&A world. Not surprisingly, it drew some pushback, not so much from bankers, but from journalists and lawyers, taking me to task for not understanding the context for these valuations. As Matt Levine notes  in his Bloomberg column , where he cites my post, "a fairness opinion is not a real valuation, not a pure effort to estimate the value of a company from first principles and independent research" ( Trust me. No one is setting the bar that high. I was looking for biased efforts using flawed principles and haphazard research and these valuation could not even pass that standard )  and that "they (Lazard and Evercore) are just bankers; their expertise is in pitching and sourcing and negotiating and executing deals -- and in plugging in discount rates into preset spreadsheets -- not in knowing the future". ( Bingo! So why are they doing these fairness opinions and charging millions of dollars for doing something that they not good at doing? And there is a difference between knowing the future, which no one does, and estimating the future, which is the essence of valuation. ) If Matt is right, the problems run deeper than the bankers in this deal, raising questions about what the purpose of a   "fairness opinion" is and whether we it has outlived its usefulness (assuming that it was useful at some point).

What is a fairness opinion? I am not a lawyer and I don't play intend to play one here, but it is perhaps best to revert back to the legal definition of the term. In an excellent article on the topic , Steven Davidoff defines a fairness opinion as an "opinion provided by an outsider that a transaction meets a threshold level of fairness from a financial perspective". Implicit in this definition are the assumptions that the outsider is qualified to pass this judgment and that there is some reasonable standard for fairness .  In corporate control transactions (acquisition, leveraged buyout etc.), as practiced today, the fairness opinion is delivered (orally) to the board at the time of the transaction, and that presentation is usually followed by a written letter that summarizes the transaction term,and the appraiser's assumptions and attests that the price paid is "fair from a financial point of view". That certainly sounds like something we should all favor, especially in deals that have obvious conflicts of interest, such as management-led leveraged buyouts or transactions like the Tesla/Solar City deal, where the interests of Elon Musk and the rest of Tesla 'sstockholders may diverge.

Note that while fairness opinions have become part and parcel of most corporate control transactions, they are not required either by regulation or law. As with so much of business law, especially relating to acquisitions, the basis for fairness opinions and their surge in usage can be traced back to Delaware Court judgments. In Smith vs Van Gorkom , a 1985 case, the court ruled against the board of directors of Trans Union Corporation, who voted for a leveraged buyout, and specifically took them to task for the absence of a fairness opinion from an independent appraiser. In effect, the case carved out a safe harbor for the companies by noting that “ the liability could have been avoided had the directors elicited a fairness opinion from anyone in a position to know the firm’s value ”.  I am sure that the judges who wrote these words did so with the best of intentions, expecting fairness opinions to become the bulwark against self-dealing in mergers and acquisitions. In the decades since, through a combination of bad banking practices, the nature of the legal process and confusion about the word "fairness", fairness opinions, in my view, have not just lost their power to protect those that they were intended to but have become a shield used by managers and boards of directors against serious questions being raised about deals. 

There are appraisers who take their mission seriously and evaluate the fairness of transactions in their opinions but the Tesla/Solar City valuations reflect not only how far we have strayed from the original idea of fairness but also how much bankers have lowered the bar on what constitutes acceptable practice.  Consider the process that Lazard and Evercore used by  to arrive at their fairness opinions in the Tesla/Solar City deal, and if Matt is right, they are not alone:

Given that the fairness opinion, as practiced now, is more travesty than protection and an expensive one at that, the first option is to remove it from the acquisition valuation process. That will put the onus back on judges to decide whether shareholder interests are being protected in transactions. Given how difficult it is to change established legal practice, I don't think that this will happen. The second is to keep the fairness opinion and give it teeth. This will require two ingredients to work, judges that are willing to put fairness opinions to the test and punishment for this who consistently violate those fairness principles.

Many judges have allowed bankers to browbeat them into accepting the unacceptable in valuation, using the argument that what they are doing is standard practice and somehow professional valuation.  As someone who wanders across multiple valuation terrain, I am convinced that the valuation practices in fairness opinions are not just beyond the pale, they are unprofessional. To those judges, who would argue that they don't have the training or the tools to detect bad practices, I will make my pro bono contribution in the form of a questionnaire with flags (ranging from red for danger to green for acceptable) that may help them separate the good valuations from the bad ones.

Question	Green	Red
Who is paying you to do this valuation and how much? Is any of the payment contingent on the deal happening? ( FINRA rule 2290 mandates disclosure on these )	Payment reflects reasonable payment for valuation services rendered and none of the payment is contingent on outcome	Payment is disproportionately large , relative to valuation services provided, and/or a large portion of it is contingent on deal occurring .
Where are you getting the cash flows that you are using in this valuation?	Appraiser estimates revenues, operating margins and cash flows, with input from management on investment and growth plans.	Cash flows supplied by management / board of company.
Are the cash flows internally consistent?	1.      Currency : Cash flows & discount rate are in same currency , with same inflation assumptions. 2.      Claim holders : Cash flows are to equity (firm) and discount rate is cost of equity (capital) . 3.      Operations : Reinvestment, growth and risk assumptions matched up.	No internal consistency tests run and/or DCF littered with inconsistencies , in currency and/or assumptions. -        High growth + Low reinvestment -        Low growth + High reinvestment -        High inflation in cash flows + Low inflation in discount rate
What discount rate are you using in your valuation?	A cost of equity (capital) that starts with a sector average and is within the bounds of what is reasonable for the sector and the market.	A cost of equity (capital) that falls outside the normal range for a sector , with no credible explanation for difference.
How are you applying closure in your valuation?	A terminal value that is estimated with a perpetual growth rate < growth rate of the economy and reinvestment & risk to match.	A terminal value based upon a perpetual growth rate > economy or a multiple (of earnings or revenues) that is not consistent with a healthy, mature firm.
What valuation garnishes have you applied?	None.	A large dose of premiums (control, synergy etc.) pushing up value or a mess of discounts (illiquidity, small size etc.) pushing down value.
What does your final judgment in value look like?	A distribution of values , with a base case value and distributional statistics.	A range of values so large that any price can be justified.

If this sounds like too much work, there are four changes that courts can incorporate into the practice of fairness opinions that will make an immediate difference:

1. Deal makers should not be deal analysts : It should go without saying that a deal making banker cannot be trusted to opine on the fairness of the deal, but the reason that I am saying it is that it does happen. I would go further and argue that deal makers should get entirely out of the fairness opinion business, since the banker who is asked to opine on the fairness of someone else's deal today will have to worry about his or her future deals being opined on by others.
2. No deal-contingent fees : If bias is the biggest enemy of good valuation, there is no simpler way to introduce bias into fairness opinions than to tie appraisal fees to whether the deal goes through. I cannot think of a single good reason for this practice and lots of bad consequences. It should be banished.
3. Valuing and Pricing : I think that appraisers should spend more time on pricing and less on valuation, since their focus is on whether the "price is fair" rather than on whether the transaction makes sense. That will require that appraisers be forced to justify their use of multiples (both in terms of the specific multiple used as well as the value for that multiple) and comparable firms. If appraisers decide to go the valuation route, they should take ownership of the cash flows, use reasonable discount rates and not muddy up the waters with arbitrary premiums and discounts. And please, no more terminal values estimated from EBITDA multiples!
4. Distributions, not range s: In my experience, using a range of value for a publicly traded stock to determine whether a price is fair is useless. It is analogous to asking, "Is it possible that this price is fair?, a question not worth asking. Instead, the question that should be asked and answered is "Is it plausible that this price is a fair one?"  To answer this question, the appraiser has to replace the range of values with a distribution, where rather than treat all possible prices as equally likely, the appraiser specifies a probability distribution. To illustrate, I valued Apple in May 2016 and derived a distribution of its values :