Quantifying The Private Company Discount: Multiples Approach and Acquisition Approach
Quantifying The Private Company Discount: Multiples Approach and Acquisition Approach
1
                                                                                                               JANUARY 2015
2
The Kooli study recognized the Koeplin study’s weak-                      study focused on private and public company transactions
nesses and attempted to control for these weaknesses.                     between 1995 and 2002. These transactions were controlling
In general, the Kooli study used similar procedures to                    interest transactions for U.S.-based companies.
those used in the Koeplin study with the exception of its
matching procedures (i.e., using a portfolio approach for                 The Kooli study found that transaction multiples of public
transaction matching).                                                    companies were typically greater than the transactions
                                                                          multiples of private companies. More specifically, the
The Kooli study identified 331 private company transactions               transactions multiples based on sales, earnings and cash
using the DoneDeals database and the SDC database. The                    flow were greater by 17 percent, 34 percent and 20 percent,
                                                                                                            respectively.
    Table 2 — Kooli Study: Median Discount Indications                                                     The Kooli study used re-
  Across Industry Categories Study Results for Transactions                                                gression analysis to identify
                                                                                                           contributing factors that
            Occurring Between 1995 and 2002                                                                may help to explain the
                                                       Discount         Discount         Discount          variation of PCD observa-
                                                      Indication    Indication Based    Indication
                                                    Based on Trans- on Transaction Based on Trans-         tions. The study concluded
                                                    action Multiple  Multiple Price/ action Multiple       that the PCD varied due to
                                                      Price/Sales       Earnings     Price/Cash Flow       firm characteristics and in-
 Industry Sector                                          (%)             (%)               (%)            dustry classification. For ex-
 Agriculture and Mining                                  -58.6             49.0             31.5           ample, companies that were
 Construction                                             70.2             59.0             19.1           classified as large and grow-
 Manufacturing                                            36.7           a 30.5          b 21.6            ing generally had a smaller
 Transportation and Communication                        -30.3             18.1             21.6           PCD than small companies
                                                                                                           with lower growth.
 Wholesale and Retail Trade                               60.1           a 55.7            -10.4
 Finance, Insurance, and Real Estate                     -35.3           b 29.2          a  3.8
                                                                                                           The results of the Kooli
 Services                                                 15.4             33.6          b 34.1            study as classified by in-
 a. Statistically significant at the 1 percent level.                                                      dustry sector, including the
 b. Statistically significant at the 5 percent level.
                                                                                                           identification of statistical
                                                                                                           significance, are provided
                                                                                                           in Table 2.
           Table 3 — Officer Study: Private Company Discount
           Estimate Study Results for Transactions Occurring                                               The Kooli study found that
                         Between 1979 and 2003                                                             private companies operat-
                                                                                                           ing in the wholesale and
                                                    Private Company Target    Unlisted Subsidiary
                                                     Discount/(Premium) to Company Target Discount         retail trade industry and
                                                    Public Company Target to Public Company Target         construction industry sec-
Financial Transaction Metric                                   (%)                    (%)                  tors transacted at greater
Price to Book Value of Equity               Average          -15.61                  27.47                 discounts than businesses
                                                                                                           operating in most other
Price to Book Value of Equity                Median            -15.22                   35.18              industries. In general, the
Number of Transactions (in #)                                     106                     145              results suggest that the
Price to Earnings Per Share                 Average             22.85                   28.90              PCD varies by industry.
Price to Earnings Per Share                 Median              27.82                   38.03
Number of Transactions (in #)                                     148                     136              The Kooli study also pre-
Deal Value to EBITDA                        Average             17.18                   26.91              sented a regression analy-
Deal Value to EBITDA                        Median              20.14                   35.07              sis to determine statistical
                                                                                                           significance of explana-
Number of Transactions (in #)                                     111                     107
                                                                                                           tory factors that impact
Deal Value to Sales                         Average             18.15                   29.99
                                                                                                           the PCD. The regression
Deal Value to Sales                         Median              18.72                   40.91              results indicated that the
Number of Transactions (in #)                                     308                     590              PCD tends to be smaller
Average Acquisition Discount                                    17.28                   28.31              for large (as measured by
Median Acquisition Discount                                     19.51                   35.95              assets) and growing com-
Number of Transactions (in #)                                     364                     643              panies. The study results
                                                                                                                                            3
    also suggest that there are many unexplained variables that            current dates—that is, it was conducted over the 1999–2006
    impact the PCD.                                                        time period. The De Franco study is considered to be similar
                                                                           to the Officer study. That is, it uses similar two-digit SIC code
    Officer Study (2007). The “Officer study,” published in the            matching procedures.
    Journal of Financial Economics, provides another perspec-
    tive of PCD evidence.5 One of the primary purposes of this study,      The Block study, as published in 2007, reported PCD indications
    in addition to calculating the PCD, was to determine if illiquidity    of 14 percent based on enterprise value-to-book value multiple
    of the target company influenced the size of PCDs. To determine        analyses and 24 percent based on enterprise value-to-revenue
    if illiquidity influenced PCDs, the Officer study analyzed both        multiple analyses. The De Franco study, as published in 2007,
    (1) private company acquisition pricing multiples and (2) unlisted     reported PCD indications of between 21 percent to 37 percent.
    subsidiary acquisition pricing multiples, to compare to (3) public
    company acquisition pricing multiples.                                 Another acquisition approach study not specifically addressed
                                                                           herein is the James A. DiGabriele study (“DiGabriele study”).7
    The Officer study initially identified 12,716 company acqui-           The DiGabriele study presents a statistical analysis used to
    sition bids (both successful and unsuccessful) using SDC.              investigate the impact of the Sarbanes-Oxley Act of 2002
    The search was conducted to find transactions that occurred            (SOX) on private company valuation.
    between 1979 and 2003. The study then actively eliminated
    transactions in which SDC merger and acquisition transaction           According to the DiGabriele study, transaction data sug-
    data was incomplete.                                                   gests that the PCD is greater post SOX than it was pre SOX.
                                                                           Therefore, valuations of private companies were adversely
    In order to measure the private company (and unlisted subsid-          impacted by SOX. According to the study, this impact is
    iary) acquisition discounts, the comparable industry transaction       generally due to SOX compliance costs. These costs include
    method was used. For this method, Officer formed portfolios of         increased due diligence costs that a public company typically
    publicly traded acquisition targets to compare to each unlisted        incurs after acquiring a privately held company in order to
    target, similar to procedures used in the Kooli study.                 comply with SOX.
    Portfolio selection was based on finding (1) public targets in the     The Multiples Approach
    same two-digit standard industrial classification (SIC) code as        The John K. Paglia and Maretno Harjoto study (“Pa-
    the unlisted target, (2) deal value excluding assumed liabilities      glia study”) attempted to determine if a PCD can be
    within 20 percent of the unlisted target deal value and (3) acquisi-   quantified based on a multiples approach analysis. 8
    tions that were announced within a three-calendar-year window          According to the Paglia study, acquisition multiples studies
    centered on the announcement date of the unlisted acquisition.         have weaknesses. The most significant weakness is lack of
                                                                           good matches between private company transactions and public
    The results of the Officer study, including number of observa-         company transactions. In certain acquisition multiples studies,
    tions per financial metric, are presented in Table 3.                  the sample sizes were less than 100 in total count. In certain
                                                                           other studies, the matching criteria employed cast a relatively
    Based on the Officer study results, unlisted targets—private           wide net based on industry classification to establish matches—
    companies and unlisted subsidiaries—are acquired at approxi-           e.g., relying on two-digit SIC code matching. Another noted
    mately 15-percent to 30-percent lower transaction multiples            weakness is that it is unknown if any of the transactions used
    relative to comparable publicly traded acquisition targets.            for comparison incorporate strategic value.
    According to Officer, and based on other evidence provided             The Paglia study attempted to address weaknesses of the ac-
    in the study, the study results support the hypothesis that            quisition multiples approach by (1) identifying a larger group
    acquisition prices are sensitive to the liquidity needs of the         of comparable transactions, and (2) identifying better private
    target company owners. As such, the study concluded that               company and public company matches using a multiples ap-
    selling parties are willing to sell assets at a discount because       proach instead of the acquisition approach. This study com-
    of liquidity needs. The greater the liquidity needs, the greater       pares the value multiples derived by (1) public market pricing
    the discount indications.                                              of publicly traded stocks and (2) private company acquisition
                                                                           transaction pricing.
    Other Acquisition Studies. Other acquisition approach studies
    not extensively discussed herein include (1) Block and (2) De          Paglia Study (2010). The Paglia study’s quantification of the
    Franco et al.6 I only mention these studies in passing because         PCD is subject to the presumption that publicly traded market
    these studies are considered to be similar to other PCD studies.       prices approximate controlling interest values. This condition
                                                                           is based on the premise presented by Eric Nath.9 If true, then
    According to the Paglia study (which is discussed below), the          the merger and acquisition (M&A) transaction values of pri-
    Block study is an extension of the Koeplin study using more            vate companies represent a similar level of value to publicly
4
traded company values since each value indication is based         Based on the matching criteria, the Paglia study identified 674
on control-level pricing indications.                              matched pairs based on annual net sales and 635 matched pairs
                                                                   based on EBITDA.
In order to quantify the PCD, or DLOM as Paglia refers to
it, the Paglia study relied on the following four analytical       Third, market value of invested capital (MVIC) pricing mul-
procedures.                                                        tiples to (1) net sales and (2) EBITDA pricing multiples were
                                                                   calculated for the matched pairs.
First, screening criteria were developed to identify privately
held company M&A transactions. The Paglia study used the           Fourth, the study compared the matched pairs based on MVIC-
following methodology and screening criteria.                      to-sales and MVIC-to-EBITDA pricing multiple indications. The
                                                                   differences between the matched pairs yielded DLOM estimates.
1.	 Privately held company M&A transactions, as provided in        In general, the Paglia study found that all measures of market
    the Pratt’s Stats database, occurring between 1993 to 2008     multiples—including MVIC/Sales, MVIC/Gross Profit, etc., for
2.	 M&A company targets with annual net revenues of at             private companies were significantly less than the same multiples
    least $10 million                                              for publicly traded companies. That finding is generally consistent
3.	 M&A company targets located in the United States               with the acquisition approach studies. In contrast, the study found
4.	 Companies classified as utilities, financial services and      that mean and median profitability measures—i.e., return on eq-
    other service related companies were excluded                  uity, net profit margins, etc.—for private companies were generally
                                                                   equal to or greater than the matched publicly traded businesses.
Second, matching criteria were developed to identify publicly
traded guideline companies to match to the privately held          The following two equations were used to calculate the private
companies involved in M&A transactions. The Paglia study           company DLOM estimates:10
identified publicly traded companies listed on the AMEX,
NYSE and NASDAQ and matched them to privately held                 1.	 DLOMSALE (%) = [1– (MVIC/Sale for private firm) /
companies based on a two-step procedure.                               (MVIC/Sale for public firm)] x 100
                                                                   2.	 DLOMEBITDA (%) = [1– (MVIC/EBITDA for private
1.	 First, matching was performed based on industry clas-              firm) / (MVIC/EBITDA for public firm)] x100
    sification, as represented by six-digit North American
    Industrial Classification System (NAICS) code matching.
                                                Based on DLOMSALE calculations, private company transaction
2.	 Second, financial fundamentals of net sales and EBITDA
                                                multiples were 67-percent lower, on average, than the similar
    were used to identify matches.              publicly traded companies, and 73-percent lower than similar
                                                                                   public companies based
   Table 4 —       Paglia Study: Private Company Discounts Across                  on median transaction
                                                                                   multiple indications.
   Industry Sectors Study Results for Transactions Occurring
                       Between 1993 -2008                                                                 Based on DLOMEBIT-
                                2 Digit                                                                   DA calculations, private
                                                      PCD DLOM              PCD DLOM
                                NAICS      Number of Based on    Number of Based on MVIC/                 company transaction
                                 code      Businesses MVIC/Sales Businesses   EBITDA                      multiples were 66-per-
NAICS Code - Industry Sector      (#)         (#)        (%)        (#)          (%)                      cent lower, on average,
Mining                             18          18        70.40        22         67.00                    than the similar publicly
Construction                       22          22        58.97        35         52.37                    traded companies, and
Manufacturing                   31-33         257        71.79       245         76.46                    72-percent lower than
Wholesale Trade                    42          46        65.73        47         64.28
                                                                                                          similar public companies
                                                                                                          based on median transac-
Retail Trade                    44-45          58        66.65        44         57.43
                                                                                                          tion multiple indications.
Transportation                  48-49          17        51.81        25         37.42
Information                        51          92        88.91        65         83.65                    The Paglia study pre-
Professional Services              54          84        81.20        51         88.70                    sented two-digit NAICS
Staff Support & Waste              56          34        74.10        19         71.43                    industry category sector
Management                                                                                                PCD indications. This
Healthcare                          62          26         43.79            35             80.22          information is presented
Art & Entertainment                 71           4         59.12             4             58.81          in Table 4.
Accommodation                       72          15         62.82            18             76.01
& Food Service                                                                                            As presented in Table 4,
Total Number of Companies                     673                          610                            companies in informa-
                                                                                                                                         5
    tion and professional services sectors had the largest PCD             These transactions provided evidence of PCDs of 17 percent
    indications. In contrast, companies in the transportation sector       using revenue-based transaction multiple comparisons, 34
    had the lowest PCD indications.                                        percent using earnings-based multiple comparisons and 20
                                                                           percent using cash flow-based multiple comparisons.
    In addition to the matched-pairs analysis, the Paglia study ex-
    amined factors that influence the DLOM. More specifically, the         As published in 2007, the Officer study identified various
    study investigated the influence of size, profitability, financial     private company transactions that occurred between 1979
    distress, purchase and purchaser characteristics, market liquid-       and 2003. These transactions provided evidence of PCDs of
    ity, market volatility, time period and industry affiliation on        15 percent to 30 percent. The Officer study also presented
    observed PCD. In order to study these influential factors, the         evidence suggesting that the PCD is sensitive to the liquid-
    Paglia study developed the following hypotheses:                       ity needs of the target private company owners. That is, the
                                                                           greater the need for liquidity, the larger the PCD.
    1.	 Larger firms have lower discounts.
    2.	 Private firms with positive profits have lower discounts.          The Kooli study and the Officer study were different than the
    3.	 Private firms that are bought by strategic buyers have lower       Koeplin study primarily due to company matching (private
        discounts compared to those that are bought by financial           to public) procedures. That is, the Kooli study and the Of-
        buyers.                                                            ficer study used a portfolio matching approach in order to
    4.	 Firms exhibiting greater risk of financial distress have higher    match private companies to a portfolio of public companies.
        discounts than those with lower levels of financial risk.
    5.	 Discounts are larger due to decreased liquidity of public          According to these studies, this matching approach was per-
        markets.                                                           formed to lessen the potential noise that is often created by re-
    6.	 Discounts are larger when public markets are more volatile.        lying on only one statistical point of reference. In other words,
                                                                           by relying on only one public company as a reference point,
    A multivariate regression analysis was used to test the Paglia study   certain differences between the public and private companies
    hypotheses. In general, the regression results support several of      can result in unintended analysis indications.
    the Paglia hypotheses. The results indicated that private firms with
    (1) a larger book value of assets, (2) positive net income and (3)     As published in 2010, the Paglia study identified 674
    lower probability of financial distress (that is, firms with higher    matched pairs based on sales revenue and 635 matched pairs
    Altman’s Z scores) had significantly lower PCD indications.            based on EBITDA between 1993 and 2008. These transac-
                                                                           tions provided evidence of PCDs of 66 percent to 73 percent.
    In contrast, the regression results indicated that (1) the buyer       The Paglia study used multivariate regression analysis to test
    type (publicly traded company buyer or private company buy-            certain hypotheses related to the level of PCD. The study
    er), (2) the transaction type (asset purchase or stock purchase)       found that larger and profitable private firms generally had
    and (3) the organization type (C corporation or pass-through           lower PCD indications.
    entity) do not influence PCD indications. Furthermore, the
    regression results (1) did not support the hypothesis that greater     Collectively, these studies provide evidence that private com-
    discounts are observed when market volatility increases, and           panies often sell at lower multiples than their public counter-
    (2) indicated only mild support for greater discounts when             parts. These lower multiples are likely influenced by the lack
    market liquidity decreases.                                            of liquidity/marketability of private company ownership as
                                                                           compared to public company ownership.
    Summary and Conclusion
    All studies discussed herein provided evidence of PCDs.                Therefore, when valuing a private company, by reference to
    These studies identified PCD evidence using the acquisition            an otherwise-similar but public company, a DLOM should
    approach and the multiples approach. Of the listed studies,            typically be considered when the public company multiples
    only the Paglia study employed a multiples approach to                 are not otherwise adjusted. In general, study research suggests
    estimate the PCD.                                                      that transaction multiples are influenced by subject company
                                                                           size and profitability.
    As published in 2000, the Koeplin study identified 84 domestic
    company matched-pair transactions and 108 foreign company              In addition to citing PCD evidence as a factor used to support
    matched-pair transactions that occurred between 1984 and               DLOM decision making, another practical use of the PCD
    1998. These transactions provided evidence of PCDs of 20               evidence, and more specifically the Paglia study data, is in
    percent to 30 percent for domestic company transactions and            the context of a market-based valuation approach—primarily
    40 percent to 50 percent for foreign company transactions.             the guideline publicly traded company method.
    As published in 2003, the Kooli study identified 331 private           A valuation analyst might consider citing PCD data as
    company transactions that occurred between 1995 and 2002.              means to support the selection of a guideline pricing
6
multiple to apply to a subject private company financial                               amination of the Private Company Discount: The Acquisition Approach.”
fundamental. In other words, if guideline publicly traded                              The Journal of Private Equity, Summer 2003.
companies are trading at an average of 10 times EBITDA, an                         4
                                                                                     	 Brav, A., C. Geczy, and P. Gompers. “Is the Abnormal Return Following
analyst might consider citing the Paglia study as a reference                          Equity Issuance Anomalous?” Journal of Financial Economics, 56
to support a lower-than-average market-based valuation                                 (2000), pp. 209–249.
analysis conclusion. ◆                                                             5
                                                                                     	 Micah S. Officer, “The price of corporate liquidity: Acquisition discounts
                                                                                       for unlisted targets,” Journal of Financial Economics, 83, (2007),
Kevin M. Zanni, ASA, CBA, CVA, CFE, is a manager in the                                571–598.
Chicago office of Willamette Management Associates, a valu-                        6
                                                                                     	 Stanley Block, “The Liquidity Discount in Valuing Privately Owned
ation consulting, economic analysis, and financial advisory                            Companies.” Journal of Applied Finance, Fall Winter 2007.
services firm, www.willamette.com. He may be reached by                            	 Gus De Franco, Ilanit Gavious, Justin Jin, and Gordon D. Richard-
email at kmzanni@willamette.com or (773) 399-4333. An                                  son, “Do Private Company Targets that Hire Big4 Auditors Receive
earlier article written on this topic by Kevin Zanni, “Private                         Higher Proceeds?” University of Toronto Working Paper, December
Company Discount Studies,” appeared in the July/August 2014                            19, 2008.
issue of The Value Examiner.                                                       7
                                                                                     	 James A. DiGabriele, “The Sarbanes-Oxley Act and the private company
                                                                                       discount: An empirical investigation,” Critical Perspectives on Ac-
     Endnotes                                                                          counting, Volume 19 Issue 8, December 2008.
1	
    John Koeplin, Atulya Sarin, and Alan C. Shapiro, “The Private Company
                                                                                   8
                                                                                    	 John K. Paglia and Maretno Harjoto, “The Discount for Lack of Marketability
    Discount,” Journal of Applied Corporate Finance, Volume 12 Num-                    in Privately Owned Companies: A Multiples Approach,” Journal of Busi-
    ber 4, Winter 2000.                                                                ness Valuation and Economic Loss Analysis 5, n. 1 (2010): Article 5.
2
  	 Enterprise value = number of targeted shares multiplied by offering price
                                                                                   9
                                                                                     	 Nath, Eric W., “Control Premiums and Minority Interest Discounts in
    plus the book values of (1) short-term debt, (2) straight debt, (3) convert-       Private Companies.” Business Valuation Review, June 1990, 39–46.
    ible debt, and (4) preferred stock less marketable securities.
                                                                                   10
                                                                                      	The Paglia study excluded outlier DLOM indications. That is, the study only
3
  	 Maher Kooli, Mohamed Kortas, and Jean-Francois L’Her, “A New Ex-                   relied on DLOM indications that fell between zero percent and 100 percent.