"If you have absolute control, you obviously have no adjustment to value," said Rob Grossman. "But what about 95%? Is that worth as much?" Grossman doesn't quantify this reduction of value, but, in fact, as he points out, "no one buys 95%, and there are legal and 'nuisance' rights held in that remaining shares that increase risk." Rob and Rod Burkert are leading a session on control premia and discounts here in Dallas at the NACVA Annual CTI Conference.
While "nuisance discounts" have been disallowed if they've ever been used at at, the fall along the "broad spectrum of control." If you look at the BVR/Factset Control Premium Study, you can still be dealing with subjectivity, "where no one has a lack of control, but no one has control either, given swing vote implications, or supra majority rules in your state," said Grossman. "Eventually we're down to the minority position, which most of us see when we're asked to fractionalize a business."
Burkert inserts the standard chart of degrees of control and value in his reports, and places the subject interest along the continuum. "This way the reader can see where full control is lacking, even when the subject interest is not a clear minority position."
Control position do not necessarily yield a control premium, either. "If there's nothing that can be done that would yield a higher benefit stream, there should not be a control value premium," Burkert stresses. This has been a touchy issue--whether or not public company data produces a minority position. "You have to be able to prove that new ownership could produce a better benefit stream going forward," agrees Grossman. "While the stocks in my personal portfolio could ALL be managed better, for the most part we assume public company values are created by the best possible management. So, no control premium is appropriate."
"The question will always be if you can justify a 25% haircut to achieve a minority value, or conversely a 33% premium to move from minority to control," Burkert says. "We're all familiar with the methodology, but can we support the judgement call?"
There are 3 sources for finding control premie: the first is Factset Mergerstat Review, the black book that comes out annually from BVR. Only about 15% of the NAVCA members in the audience use Mergerstat Review. Burkert advises those who use this source to remember that the data there are weighted to the industries where there's a lot of M&A--pharma and computers, for example. "Courts and others are requiring more and more specific data so there's some cause for concern," said Grossman--though questions from the audience suggested that not every one feels the focus on the heavily transacted industries has much of an effect on the overall average.
"Improvements in technical practice standards have reduced the gaps in discounts from the old days where I'd come in at 18% and the other side would come in at 35%," said Grossman. "We're just not that far apart any longer as the professional methods and benchmarks have improved."
"We're talking about Mergerstat Review first, but particularly in valuations for estate purposes that are reviewed closely, you may want more than one source," said Burkert. "It's not such a big deal when your client is looking for guidance on an acquisition."
Another issue with Mergerstat Review as a single source is that the methodology excludes the 18% of deals where there are negative control premium, "but this is market evidence that you're ignoring," says Grossman. "This might be a deal that happens in an unattractive industry having nothing to do with control," Burkert suggests. So you have some reason for debate or concern before you start, the two NACVA speakers agreed. "Here's what I do when I use Mergerstat Review: I construct tables of the previous years of total premie, premia by industry, and premie by size," suggested Burkert. "I then look at trends, up or down, and form a judgment about what my three tables say."
The second method is the BVR/FactsetMergerstat Control Premium Study. "The major difference is that the CPS focuses on individual transactions and only included transactions that have actually closed," Burkert correctly said. "There's also more information about the underlying transactions, too, putting them into five categories by type of buyer or transaction. This makes it easier, at least, to back out the extra value for synergistic value or whatever. "Without some modification of the median, many appraisers will assume that you've automatically overstated the control premium," says Grossman. "There's not other way to ultimately get to the implied minority discount."
Neither of the sources from BVR can easily be used for fair market value for tax-purpose valuations, so may appraisers now try to solve by modifying the numerator--the control or non-control cash flows.
One attendee expressed concern that you could "double count" risk in both the discount rate, AND in the minority interest discount. "There has to be some sort of nexus between the two calculations," the attendee said. "You've got 18% as a cost of capital, and a 45% minority discount. Where do the two meet or overlap?" Grossman commented that Robert Comment, from Johns Hopkins, has written about the overlap, at least, between size premia and marketability discounts.
"We've been fortunate to see the prejudices of the IRS in the practice job aid that was 'leaked' last year," says Burkert. "So, it gives us a picture and critique of what the Service thinks is out there, and what they value." It's not that it's cutting edge; it's because it "helps us see how the IRS thinks. Most importantly, we know that the IRS is no longer accepting the averages of the averages to get to 35%. Things have really really changed, pushed by activist courts." The job aid mentions 33 factors, and some appraisers have begun to address all of the relevant factors in their reports.
As a result, "the longest part of my tax valuations, probably 25-33% of my reports, are dedicated to supporting my discounts," says Burkert. "We got support for this from the BVR Tax Valuation Summit hosted by Mel Abraham and Judge David Laro at the University of San Diego School of Law," says Grossman. "Howard Lewis stood up first and said simply that 'discounts need to be supported by market evidence'."
"These 33 factors may be more helpful than Mandalbaum in some ways," suggested Burkert, at a minimum because some of the Mandalbaum factors are likely to have been considered in other parts of the valuation. "You have to do the best you can to not double count, and that's our job."
"It would be great if we had perfect non-marketable proxies available for every valuation," says Grossman. "Shannon Pratt says in the beginning of Valuing a Business that 'discounts don't really exist; we just have bad data that we need to adjust.' Sadly, though we rarely have the great proxies so we need to pay attention to adjustments to the income approach such as those suggested by the IRS job aid."
The IRS raises some clear red flags about marketability discounts in the Job Aid:
- Use of pre-IPO data alone
- Use of averages
- Lack of quantitative support
- Reliance solely on court decisions
Grossman says "a careful read through the job aid will make us all better appraisers."