<< . .

. 30
( : 66)



. . >>

per-share value as a 100% ownership interest, this does not affect the fact
that the block in question is still a minority block of stock having no
attributes of control over the company. He contends it would be illogical
to subtract a DLOC from a block that has no attributes of control.9 Rather,
the oftentimes extreme price differentials between public and private mi-
nority interests must be explained by other ownership attributes besides
control, including but not limited to differentials in relative liquidity, rel-
ative level of information availability, and relative information reliability.
Bolotsky claims”reasonably, in my opinion”that there are many
public ¬rms whose perceived 100% ownership value will be more than
their minority value, but not enough more to make a tender offer worth-
while. In addition, Bolotsky™s theoretical framework is the only one that
can readily accommodate several market features that appear anomalous
when relying on the linear levels of value framework, such as 100% own-
ership pricing at levels considerably below IPO pricing for many com-
panies in today™s markets.

Jankowske
Wayne Jankowske™s article (Jankowske 1991) corrects certain key errors
in the articles by Nath and Bolotsky. He says one does not have to accept
Nath™s assertion that the marketable minority value is a control value to
accept the proposition that DLOC can differ between public and private


9. This is a very logical statement and appears to be self-evident. Nevertheless, I will disagree with
this later in the chapter.




PART 3 Adjusting for Control and Marketability
204
¬rms. He says differences in legal and contractual protection, agency
costs, relative incentives, and differential economic bene¬ts can account
for differences in the public versus private DLOC.
Prevailing wisdom™s assertion is that since public market prices are
minority prices, we can use public guideline company prices to value
private minority shares, with only DLOM necessary. Jankowske says that
for that to be true, it implies that the economic disadvantages of lack of
control associated with public minority shares is equal to that of private
minority shares, which is unrealistic.
Conceptually, the magnitude of DLOC in guideline public prices
makes no difference, whether it is 30%, or, as Nath contended in his ¬rst
two articles, 0%. The difference between the public and subject company™s
DLOC must be recognized to avoid an overvaluation.
He developed the following formula to value a private minority in-
terest:10
FMVMM
Additional DLOC (DLOCSC DLOCGC)
1 DLOCGC
where
FMVMM the marketable minority fair market value
DLOCGC discount for lack of control in the public guideline companies
DLOCSC discount for lack of control in the subject company
He gave the following example: FMVMM, the marketable minority interest
value is $900; DLOCGC, the discount for lack of control implicit in the
public minority stock is 10%; and DLOCSC, the discount for lack of control
appropriate to the subject company, is 40%. His calculation of incremental
discount is:
$900
(40% 10%) $1,000 30% $300
1 10%
He disagrees with Bolotsky that the guideline ¬rms must have iden-
tical shareholder attributes.
In his second article on the topic (Jankowske 1995), he stressed that
it is the economic bene¬ts to which we must look as a justi¬cation for
control premiums, not the powers that come with control. He cites the
following economic bene¬ts of control:
— Company level.
— Performance improvements.
— Synergy.
— Shareholder Level.
— Wealth transfer opportunities”the Machiavellian ability to
transfer wealth from the minority shareholders.
— Protection of investment”the ¬‚ip side of the above point is
that control protects the shareholder from being exploited. This


10. I have changed his notation.




CHAPTER 7 Adjusting for Levels of Control and Marketability 205
motivation is important because it relates to ambiguity
avoidance in the academic literature reviewed in this chapter.
— Liquidity”control enhances liquidity in privately held
businesses.
Of the company level advantages, the extent to which performance
improvements on a standalone basis account for control premiums prop-
erly belongs in our calculations of fair market value. That portion ac-
counted for by synergy is investment value and should not be included
in fair market value.

Roach
George Roach (1998) summarized percentage acquisition premiums from
a database of business sales. The premiums were measured as (PAcq/P5Day)
1, where the numerator is the acquisition price and the denominator
is the minority trading price ¬ve days before the announcement of the
acquisition. He also provided the premium based on the price 30 days
prior. We excerpt from his Exhibit IV to our Table 7-2.
There is no pattern to the ¬rst three premiums listed in Table 7-2.
The 50 SIC code difference level between buyers and sellers has the
highest premium, which is counterintuitive. Roach found similar patterns
in the results for median premiums. Additionally, while the 30-day pre-
miums were higher than the 5-day premiums, the patterns were similar.
Under the assumption that acquisitions of ¬rms in the same or al-
most the same SIC code are more likely to be strategic acquisitions than
¬rms acquired in very different SIC codes, Roach™s analysis appears to be
strong evidence that premiums paid for strategic buys are no larger than
premiums paid for ¬nancial buys. This leads to the conclusion that ac-
quisition premiums are control premiums, not premiums for synergy.
This conclusion supports Mercer (1990) and Bolotsky in their opinion
that the similarity of premiums for acquiring minority positions and con-
trol positions can be explained as acquiring ˜˜creeping control,™™ because
it appears to rule out synergy as an explanation. This is in contrast to
Nath™s position and Mercer (1998 and 1999).

Mercer (1998) and (1999)
Mercer (1998) represents a signi¬cant change in thinking since his 1990
article. He now believes that the majority of premiums for mergers and

T A B L E 7-2

Acquisition Premiums by SIC Code


SIC Code Differences Between Buyer & Seller 5-Day Avg Prem

0 37.2%
1“9 41.0%
10“48 37.1%
50+ 50.8%

Source: George P. Roach, ˜˜Control Premiums and Strategic Mergers,™™ Business Valuation Review (June 1998), Table IV, p. 47.




PART 3 Adjusting for Control and Marketability
206
acquisitions recorded in Houlihan Lokey Howard & Zukin™s Mergerstat
represent strategic premiums for synergies, which do not qualify for fair
market value. He has modi¬ed the traditional levels of value chart in the
top of Figure 7-1 to the one in the middle. It is the same as the one above,
with the addition of a strategic value above the control value.
For sake of discussion, let™s look at an end-of-year Gordon model
formula to calculate value.

CFt 1(1 b)
PV(Cash Flows)
r g

where r is the discount rate, g is the constant growth rate, and b is the
retention ratio for the company, i.e., it retains b (with 0 b 1), leaving
(1 b) times the next year™s forecast cash ¬‚ow as forecast dividends.
Mercer™s position (Mercer 1999) is that the discount rate is the same
at the marketable minority level as it is in all levels of value above that
(though not the same as the private minority level, which almost always
carries a higher discount rate). The main difference in the valuation comes
from the numerator, not the denominator. Control buyers, whether ¬nan-
cial or strategic, upwardly adjust forecast cash ¬‚ows. He details the types
of adjustments as follows:
1. Normalizing adjustments: these adjust private company earnings
to well-run public company equivalent. Mercer classi¬es two
types of normalizing adjustments. Type 1 is to eliminate
nonrecurring items and adjust for non-operating assets. Type 2
is to adjust insider compensation to an arm™s length level,
including eliminating discretionary expenses that would not
exist in a public company.
2. Control adjustments: Mercer lists two types of control
adjustments. Type 1 are for what Jankowske calls performance
improvements and apply to both ¬nancial buyers and strategic
buyers. Mercer says these are adjustments for improving the
(existing) earnings stream, i.e., running the company more
ef¬ciently. This could also include the volume discounts coming
from the buying ef¬ciencies achievable by being owned by a
larger company that is a ¬nancial buyer. Type 2 are for changing
the earnings stream, i.e., running the company differently, and
apply only to strategic buyers. These include consolidating G&A
expenses, eliminating duplicate operations, selling more product,
and negotiating power with suppliers, distributors, or customers
that is above and beyond that which can be achieved by a
¬nancial buyer.
Mercer is in very good company in this position. Consider (Pratt
1998, p. 134): ˜˜The exploitation of minority shareholders is far less prev-
alent in public companies than in private companies, at least in larger
public companies. If company cash ¬‚ows are already maximized and the
returns are already distributed pro rata to all shareholders, then there
may be no difference between a control value and a minority value.™™




CHAPTER 7 Adjusting for Levels of Control and Marketability 207
Other similar opinions can be found in Ibbotson (1999), Zukin (1998), and
Vander Linden (1998).
Actually, Mercer is not the originator of this position on control pre-
miums, but he may be the person who has written the most about it. The
original statement of this position came from Glass and McCarter (1995).

Summary of Professional Research on Control Premiums
Now we will summarize the past 10 years of professional research on
control premiums. The primary research supporting the traditional con-
trol premiums of 35% to 40 % is Roach™s. The other extreme”a zero
control premium”is represented by Nath and Mercer. While the original
gap between Nath™s and Mercer™s positions on control premiums was
large, the current gap is much smaller and often may not even exist. The
logic of how they arrived at their opinions is different, but they would
probably come to similar calculations of control premiums in the majority
of circumstances.
Nath does not ever use control premiums. Mercer, following Glass
and McCarter, now agrees that after taking into consideration increases
in forecast cash ¬‚ows from performance improvements and arm™s length
salary adjustments for the control shareholder that are appropriate for a
¬nancial buyer, there are no control premiums. In the absence of infor-
mation to do so, he says control premiums should be very small”no
more than 10%, with the implication that it could still be as little as zero.
That is essentially Nath™s position, that the public minority value is a
control value and therefore we should not apply control premiums to a
discounted cash ¬‚ow valuation. Mercer rarely assigns control premiums
unless there are identi¬able increases in forecast cash ¬‚ows at the control
level. In that case, he would simply increase the cash ¬‚ow forecast and
the control level value would increase vis-a-vis the marketable minority
`
interest level. Nath would do the same thing, except he does not like the
term marketable minority level of value (or its synonym, as if freely traded).
Their terminology differs far more than their results, at least with regard
to control premiums.
Regarding the discount for lack of marketability (DLOM), neither
Mercer nor Nath believes in taking DLOM for a control interest”a po-
sition with which I disagree in the DLOM section of this chapter. Nath™s
reason for this position is that the M&A and business brokerage markets
are very active, and that activity negates any tendency to a DLOM. He
makes an analogy of the real estate market to the market for companies.
Rather than apply DLOM, real estate appraisers put an ˜˜expected mar-
keting time™™ on their values. Mercer™s main reason for opposing DLOM
for control interests is that they control cash ¬‚ows until a sale.
Nath always begins with a controlling owner™s value, which in his
view is the greater of the values obtained by the M&A markets, the public
markets (reduced by the restricted stock discount that one would expe-
rience in going public), and liquidation. It is extremely important to note
that to the extent that the M&A market values contains synergies and
whenever the M&A valuation dominates, Nath™s fair market values will
contain synergies.
His opinion is that that is what buyers will pay, and therefore that
is fair market value. There is a question as to whether that is investment

PART 3 Adjusting for Control and Marketability
208
value rather than fair market value. I agree that it probably is not in-
vestment value, as it is not value to a particular buyer. It is value to an
entire class of buyers. If all buyers in the M&A market are strategic”
which is certainly not completely true, but may be largely true is some
industries”then that is what buyers would pay. With this ¬ne distinction,
it is very important to make sure that if one follows Nath™s method, one
must be careful that the subject company ¬ts with the assumptions un-
derlying Nath™s logic.
I do not believe that most small ¬rms and many midsized ¬rms are
serious candidates for the M&A market. They are business broker mate-
rial, and such buyers would rarely ever be synergistic. Therefore, it is
imperative to be realistic about the market in which the subject company
is likely to sell.
For a private minority interest, Nath takes both the discount for lack
of marketability and lack of control. In conversation, he has revealed his
own dissatisfaction with the lack of relevant information for calculating
DLOC, since there is still nothing to use other than the traditional ¬‚ip
side of the control premiums that he personally demolished as being valid
premiums to add to a ˜˜minority value.™™
Mercer comes to what probably amounts to a very similar result
through a different path. He does not calculate a discount for lack of
control for private minority interests. Instead, he uses his quantitative
marketability discount model (QMDM) (which we cover in more detail
later in this chapter) to subsume any DLOC, which he feels is automati-
cally included in his DLOM. I agree with Mercer that the QMDM includes
the impact of DLOC because, in the QMDM, one must forecast the spe-
ci¬c cash ¬‚ows to the minority shareholder and discount them to present
value. Thus, by using the QMDM, Mercer does not need a DLOC. Mer-
cer™s position is internally consistent.


Prior Research”Academic
Now that we have summarized the professional literature, we will sum-
marize the results of various academic studies relevant to our topic. The
primary orientation of academic research in ¬nance is on publicly traded
stocks. It is generally not directly concerned with the issues of the valu-
ation profession, which is focused on valuing private ¬rms. Often a
slightly interesting side point in an academic article is a golden nugget
for the valuation profession”if not a diamond.
There are two types of evidence of the value of control. The ¬rst is
the value of complete control. The second deals with the value of voting
rights. Voting rights do not represent control, but they do represent some
degree of in¬‚uence or partial control.11
The academic research falls into the following categories:
1. The article by Schwert focuses primarily on analyzing returns in
mergers and acquisitions during two periods: the runup period,


11. Mergerstat Review does track premiums for acquisitions of minority interests, which make up a
third category of evidence. I am not aware of any academic literature dealing with this
issue.


CHAPTER 7 Adjusting for Levels of Control and Marketability 209
which is the time before announcement of the merger, and the
markup period, which is the time period after the
announcement. This is signi¬cant in the context of this book
primarily as providing empirical evidence that is relevant in my
economic components model for the discount for lack of
marketability (DLOM). It could easily belong to the DLOM part
of this chapter, but I include it here with the rest of the
academic articles.
2. The articles by Lease, McConnell, and Mikkelson; Megginson;
Houlihan Lokey Howard & Zukin (HLHZ); and the section on
international voting rights premia all deal with the value of
voting rights and provide insight on the value of control that
¬ts in the de¬nition of fair market value.12
3. The articles by Bradley, Desai, Kim and Maquieira, Megginson,
and Nail are about the value of complete control. In particular,
their focus is on measuring the synergies in acquisitions, which
is a critical piece of evidence to understand in sorting through
the apparently con¬‚icting results and opinions in the
professional literature.
4. The article by Menyah and Paudyal is an analysis of bid“ask
spreads and is primarily related to DLOM, not control. It could
also have been included in section on marketability.

Schwert (1996)
Since business appraisers calculate control premiums and discounts for
lack of control from merger and acquisition (M&A) data of publicly
traded ¬rms, it is important to understand what variables drive control
premiums in order to be able to properly apply them to privately held
¬rms. Schwert™s article has some important ¬ndings.
Schwert™s main purpose is to examine the relationship between run-
ups and markups in M&A pricing. (The runup period is that period of
time before the announcement of a merger in which the target ¬rm™s price
is increasing.) Schwert ¬nds that cumulative abnormal returns (CARs)13
begin rising around 42 days before an acquisition. Thus, he de¬nes the
runup period from day 42 to day 1, with day 0 being the announce-
ment of the merger. The markup period is from day 1 to day 126 or
delisting, whichever is ¬rst. The sum of the runup and markup period is
the entire relevant timeline of an acquisition, and the sum of their CARs
is the total acquisition premium.
Schwert ¬nds that CARs during the runup period for successful ac-
quisitions between 1975“1991 average 25%, with CARs for unsuccessful
acquisitions, i.e., where the bidder ultimately fails to take over the target,
averaging 19%.14 After the announcement date, CARs for successful ac-


12. The HLHZ article is professional rather than academic, but its topic ¬ts in better in our
discussion of academic research.
13. These are the cumulative error terms for actual returns minus market returns calculated by
CAPM.
14. One thousand, eight hundred and fourteen transactions in total, which are later reduced to
1,523 in his main sample.


PART 3 Adjusting for Control and Marketability
210
quisitions increase to 37%, while for unsuccessful acquisitions they de-
crease to zero.
He discusses two opposite bidding strategies, the substitution hy-
pothesis and the markup pricing hypothesis.
The substitution hypothesis states that each dollar of preannounce-
ment runup reduces the post-bid markup dollar for dollar. The assump-
tions behind this hypothesis are that both the bidder and the target have
private information that is not re¬‚ected in the market price of the stock
and that no other bidder has valuable private information. Therefore,
both the bidder and the target will ignore price movements that occur
prior to and during the negotiations in setting the ¬nal deal price.
The markup pricing hypothesis is that each dollar of preannounce-
ment runup has no impact on the post-bid markup. Thus, the prean-
nouncement runup increases the ultimate acquisition premium dollar for
dollar. The assumption behind this hypothesis is that both the bidder and
the target are uncertain about whether movements in the market price of
the target™s shares re¬‚ect valuable private information of other traders.
Therefore, runups in the stock price could cause both the bidder and the
target to revise their valuations of the target™s stock. Schwert used the
example that if they suspect that another bidder may be acquiring shares,
both the bidder and the target will probably revise their valuations of the
target™s stock upward.
The markup hypothesis re¬‚ects rational behavior of bidders and tar-
gets when they have incomplete information. A different explanation of
the markup hypothesis is that of Roll (1986), who postulates that bidders
are interested in taking over targets regardless of cost (the hubris hy-
pothesis). This would re¬‚ect irrational behavior. Using regression analy-
sis, Schwert ¬nds strongly in favor of the markup hypothesis, while re-
jecting Roll™s hubris explanation as well as the substitution hypothesis.
Had the substitution hypothesis been the winner, this would have
implied that the acquisition premiums that occur in the market would
require major adjustments for calculating fair market value. It would have
meant that the post-bid markups are based on private information to a
particular buyer and seller, who ignore the effects of the pre-bid runup
because they both believe that no other bidder has valuable private in-
formation. This would then be investment value, not fair market value.
With the markup hypothesis being the winner, at least we do not have
that complication.
For professional appraisers, the most important ¬nding in Schwert™s
paper is the impact of competitive bidding, i.e., when there is more than
one bidder for a target, on the cumulative abnormal returns on the tar-
get™s stock. Approximately 20% of the takeovers were competitive (312
out of 1,523), with 80% (1,211 out of 1,523) noncompetitive. Table 4 in the
article shows that the presence of competitive bidding increases the pre-
mium paid by 12.2%.15 This is signi¬cant evidence of the impact of com-
petition that will have an important role to play in calculating D2, the



15. That is, it adds an absolute 12.2% to the premium. It does not increase the premium by 12.2%.
For example, if the average premium with only one bidder is 30%, with two or more

<< . .

. 30
( : 66)



. . >>