OPINION
LAMB, Vice Chancellor.
I. PRELIMINARY STATEMENT
In this appraisal action, filed pursuant to Section 262 of the Delaware General Corporation Law (“DGCL”), I am called upon to determine the fair value of the shares of common stock of WXL International, Inc. (‘WXL”), a Delaware corporation, as of December 16, 1997, the date on which WXL was merged with and into its parent corporation, Communications Telesystems International (“CTS”) (the “Merger”). As of that date, petitioners were the holders of 500,000 shares (or five percent) of WXL common stock, CTS being the holder of the other 95 percent of the WXL shares. It is conceded that petitioners made a timely demand for appraisal in accordance with the requirements of Section 262.
Although the parties and their experts differ widely in their conclusions as to the appraised value of the shares, the issues in contention at this point of the proceeding are quite limited in scope. The parties’ testifying experts both used the same method of analysis (the “comparable company method”) and both testified that it was the only method properly relied upon for this case. The experts also agreed that using total revenues as the multiplicand was the only appropriate iteration of that method. Moreover, the two experts each identified eight companies as “comparable,” five of which companies were used by both experts. As a result of examination at trial, there is now agreement to include a sixth company as a “comparable.”
The experts’ analyses differed materially in only a few respects: (i) whether, in deriving a multiplier for WXL from the data generated as to comparable companies, a premium should be added to reflect WXL’s better than average rate of growth over the 12-month period leading up to the merger, (ii) at what point in the analysis the information derived should be adjusted to eliminate the minority discount inherent in the comparable company method, and (iii) whether the value derived from that analysis should, in this case, be reduced to account for any observed “private company” discount.
In summary, I conclude as follows: (a) no growth premium is warranted, (b) a control premium should be added to adjust the market value of the equity derived from the comparable company method, and (iii) on the record before me, no “private company discount” is appropriate. Based on these conclusions, and for the other reasons explained in this opinion, I find that the fair value of the shares of common stock of WXL, as of December 16, 1997, was $0.6253 per share, or a total of $312,650 for the 500,000 shares subject to appraisal. I will allow interest on this sum at the legal rate, compounded quarterly, through the date judgment is entered in accordance with these findings.
II. BACKGROUND
A. WXL’s Creation
CTS is a facilities-based long-distance telephone company that operates throughout most of the United States and has
been attempting to expand internationally. Founded in 1991, it offers to its customers direct dial long-distance services as well as prepaid calling card, Internet and other communications services.
In 1994, petitioner Carl Borruso (“Bor-ruso”) approached Roger Abbott, Chairman and CEO of CTS, and presented a business plan to begin long-distance telecommunications operations in the United Kingdom, France and Germany. At the time, CTS had no European operations. Thereafter, in November 1994, CTS and Borruso entered into a Formation Agreement and a Shareholders Agreement with respect to the to-be-formed WXL. The Shareholders Agreement attached the business plan proposed by Borruso, including financial projections. The business plan provided that CTS would fund WXL for two years, estimated to include $2 million in “seed money,” with the expectation that at the end of that period, WXL would be cash flow positive. Monthly pre-tax losses were expected until the spring of 1996. On the projected balance sheets, WXL was expected to have retained earnings of nearly $2 million by September, 1996.
At WXL’s formation, Borruso received 400,000 shares, or 4%, of the company’s stock; the other petitioner, William Lee received 100,000 shares, or 1%. CTS received the remaining 9,500,000 shares, or 95% of the common equity.
B. WXL’s Operations
WXL began operations sometime in the first half of 1995. It had revenues of $0.6 million for the fiscal year (“FY”) ended September 30, 1995. Revenues increased to $5.2 million in FY’96. These revenue figures were far lower than the $7 million and $89 million revenues projected in the business plan attached to the Shareholders Agreement for FY’95 and FY’96, respectively. Revenues grew to $10.7 million in FY’97, $11.9 for the 12 months ended December 31, 1997, and were $2.8 million in the first quarter of FY’98 (ended December 31,1997).
Over this period, the mix of WXL’s revenue streams changed rapidly. Commercial revenue amounted to 72.4% of revenues in FY’95 but only 14.7% in FY’97. Carrier revenue represented 17.4% of revenues in FY’96, 68.5% in FY’97 and only 17.9% in the first quarter of FY’98. Finally, inter-company revenue grew from 6.1% of total revenue in FY’96 to 48.0% of total revenue in the first quarter of FY’98. These radical changes in revenue sources reflected quickly changing competitive pressures and opportunities. For example, when WXL initially failed to attain the projected amount of revenues from residential or commercial end-users, it sought to enter the more “niche” calling card market. This foray was successful at first but revenues shrank when competitive pressures mounted. WXL then identified “call shops” as a potential source of revenues.
WXL was, at first, successful in generating substantial revenues in this business. But WXL was quickly forced to change its credit and collection procedures when bad debt charges rose to unacceptable levels. These procedural changes caused the call shop revenue flow to dry up.
WXL’s revenues were derived principally from operations in the United Kingdom. WXL had other, less substantial, operations in France and Germany. The United Kingdom had voluntarily deregulated its market for telecommunications services in the early 1990’s, which led to greater opportunities for WXL in those markets. Deregulation of the markets in France and Germany did not occur until after the effective date of the merger. Even then, the actual process and timetable for deregulation in those and other European countries remained uncertain. The key element of
deregulation is the ability of newly competing companies to interconnect with the dominant telephone company in each country. At the time of trial, WXL had still not achieved interconnection in Germany.
Operating income (loss) before interest, taxes, depreciation, and amortization (“operating EBITDA”) was ($2.6) million in FY’95, ($2 .9) million in FY’96 and ($4.5) million in FY’97. For the first quarter of FY’98 the operating EDITDA was ($0.4) million. By contrast, the projections attached to the Shareholders Agreement called for operating EBITDA of ($1.6 million) in FY’95 (on revenues of $7 million) and of $3.6 million in FY’96 (on revenues of $39 million).
By early 1997, CTS began to consider its options with regard to WXL, including the possibility of shutting it down. Ultimately, CTS decided not to shut WXL down, reflecting the concern that doing so would adversely affect CTS’ operations in other countries in which CTS was having greater success. Also, CTS hoped that, with an infusion of needed additional capital, WXL’s operations in the UK and Europe could prove to be profitable.
C. The Merger
On December 16,1997, CTS, as the owner of 95% of WXL’s common shares (the only class of shares outstanding), was able unilaterally to effect a short form merger with WXL, pursuant to Section 253 of the DGCL. In the Merger, CTS fixed the consideration to be paid to Borruso and Lee at $.02 per share. The method by which that price was determined had none of the attributes that would justify giving it any consideration in this proceeding.
No outside financial analysis of WXL was obtained by CTS. Instead, the $0.02 per share price was developed informally by the principal officers or owners of CTS, who did not retain their original work papers. That proposed price was then approved by the CTS board. It would appear from the record that the directors of CTS concluded that there was no positive value to WXL.
Petitioners point out several flaws in the procedure followed by CTS in arriving at a fair price for the WXL shares. First, the valuation was done at the end of 1996 or the beginning of 1997 — up to a year before the merger became effective (Tr. at 252, 256). According to Mr. Abbott, CTS’ Chairman and CEO, the most recent financial numbers for WXL at his disposal were the first or second quarter of fiscal 1997
— when WXL was much smaller than on the Merger date.
Moreover, CTS used revenue multiples derived from the prices at which United States telecommunications companies were selling at the time. By using these multiples, of course, CTS’ owners failed to take into account that WXL was a European-based telecommunications company, which CTS acknowledged would expect to enjoy higher revenue, margins and growth rates than its American counterparts. Additionally, the fair value analysis performed by CTS may have improperly included a minority discount, not appropriately applied under the law of this State.
See Cavalier Oil Corp. v. Harnett,
Del.Supr., 564 A.2d 1137, 1144-45 (1989).
Petitioners also point out that CTS did not comply with the requirements of 8
Del. C.
§ 262, in that it delayed in mailing notice of the merger to petitioners until
January 12, 1998 — 17 days beyond the statutorily permitted time. In addition, the notice violated § 262(d)(2) by not including a copy of the appraisal statute.
II. DISCUSSION
The parties’ experts were in agreement about significant aspects of the method properly employed in this case to determine the fair value of Petitioners’ shares. I find no reason to deviate from the process they jointly followed. Thus, I make the following findings, based on the record before me in this matter:
1. It is proper to rely on the comparable company method of analysis in determining the fair value of WXL.
2. No other method of valuation is appropriately relied upon here.
3. Because no single comparable company’s profile and operations is more indicative of the value of WXL than any other’s, the median multiple exhibited by the comparable companies is the best indication of value for WXL.
4. The median unadjusted multiple for the basket of the five common comparable companies is 2.6x. Including the sixth comparable, about which there is no dispute, the unadjusted median multiple is 2.5x.
5.The adjusted multiple (after it is derived) should be applied to WXL’s latest pre-Merger 12-month revenues of $11,480,945 to derive the market value of invested capital (“MVIC”).
Thus, I will turn to that analysis and the parties’ points of disagreements.
A. The Comparable Companies
Petitioners’ testifying expert was Peter S. Huck, P.E., ASA. Mr. Huck is Assistant Vice President and Principal for the Milwaukee Financial Valuation Group of American Appraisal Associates, Inc. (“AAA”). He is a Professional Engineer and is certified as an Accredited Senior Appraiser by the American Society of Appraisers. Respondent CTS called Mr. Anthony P. Kern of PriceWaterhouseCoop-ers, LLP to testify on its behalf. Mr. Kern is Partner-in-Charge of that firm’s FAS Information, Communications and Entertainment Industry Group in Washington, D.C.
To perform their analyses, Huck and Kern both identified a set of eight companies judged by them to be sufficiently comparable to CTS. Five of the eight were common to both sets of comparables. The comparables used were as follows:
[[Image here]]
Respondent concedes that Esprit is appropriate to use as a comparable company, but continues to object to the inclusion
of Colt and IDT. I agree that Colt and IDT are not useful comparables and should be eliminated from the analysis. Colt is a local exchange carrier in the UK. It is not a long-distance company, but rather the company that owns the wires that connect homes or offices to the local switch. Thus, it does not compete with WXL or the other comparable companies. Its lack of comparability is also apparent from the great variation of its derived revenue multiplier (10.5x) from those of the other guideline companies utilized by Huck (ranging from 1.2x to 4.5x). I also agree that IDT should be disregarded in the analysis due to the fact that a substantial portion of its revenues is derived from its operations of an Internet network. The growth of that segment of IDT’s business is very strong and, as is commonly known, current valuations of all Internet related businesses are unusually high. While Petitioners are correct that IDT also competes in the European long-distance market and might otherwise be a good comparable, its participation in the Internet market offers too great a chance of distortion to allow its inclusion.
Petitioners continue to object to the inclusion of any of Kern’s distinct compara-bles: ACC, Startec and Telegroup. I agree with Petitioners that the inclusion of any one of these three companies presents substantial issues. ACC was, at the relevant time, subject to an announced agreement for its acquisition by another company. Thus, reliable information about its stock trading activity during the pertinent period of time could not be directly observed. Because there is otherwise a sufficiently large group of other companies to use, this is a sufficient reason to reject ACC as a comparable. Startec derives most of its revenue from non-European markets and only 2.2% from Western Europe and, thus, was not significantly involved in the same market as WXL. Tele-group has a highly concentrated share ownership, with 85% of its stock held by founders or otherwise in private hands. On the record before me, I am not able to discern whether there is a sufficiently active market for its shares from which to derive reliable information for valuation purposes. For these reasons, I will exclude Startec and Telegroup from the basket of companies used for purposes of comparison.
For the foregoing reasons, I find that the group of comparables should consist of the five companies identified by both experts plus Esprit.
B. The Derived Multiple
As stated above, the unadjusted median multiple for the five common guideline companies plus Esprit is 2.5x.
The next issue to be addressed is the downward adjustments made by the experts to that multiple.
Huck applied his adjustments directly to the median multiple (and not to individual comparable company multiples) to account for three factors: size, profitability and growth. He concluded that WXL’s small size and lack of profitability relative to the guideline companies, together justified a downward adjustment of 25% to 35%. He also determined that WXL’s superior latest-12-month growth rate justified a 15% to 20% upward adjustment to the median. On balance, he discounted the median multiple by 15% to adjust for differences between WXL and the guideline companies.
Kern compared WXL to each guideline company individually to assess the appropriate level of discounting. In doing so, he factored in revenue (size), profitability, and operating markets/product diversity. In my view, Kern’s company-specific discount methodology more reliably correlates WXL to the guideline companies. Nevertheless, I conclude that Kern’s decision to discount PGEX’s multiple by 15% on ac
count of profitability is excessive and will reduce that discount to 10%, which correlates better with his treatment of STRX.
Huck, but not Kern, adjusted the median multiple upward due to WXL’s superior FY’97 revenue growth rate. For several reasons, I conclude that Kern properly rejected such adjustment. There is nothing in the record to show that WXL’s rate of growth was sustainable. Even the FY’97 growth rate is in question, as it does not reflect the substantial bad debt booked in FY’97, since it appears on the income statement “below” the net revenue line.
Moreover, WXL’s failure to meet its business plan, its heavy indebtedness, its lack of access to capital and its unstable mix of business leave me unable to conclude that anyone would pay a premium to acquire WXL on account of its higher than median growth rate in FY’97.
The resulting analysis is summarized in the following chart:
From this, I derive a median adjusted multiple for the six guideline companies of 1.74x.
C. Application of a Control Premium
The parties sharply disagree about the proper application of a control premium. There is no dispute between them that the comparable company method produces a minority valuation of the shares subject to appraisal, as has been recognized in decisions of this court.
See, e.g., Kleinwort Benson Ltd. v. Silgan Corp.,
Del. Ch., C.A. No. 11107, Chandler, V.C., mem. op at 6, 1995 WL 376911 (June 15, 1995). Rather, the dispute is over the point in the analysis at which to apply a control premium and, to a lesser degree, the amount of that premium.
Huck added a 30% equity control premium to the median multiple used to determine MVIC for WXL. He then subtracted the value of WXL’s long term debt to derive the fair value of the equity. Kern also used a 30% equity control premium figure but applied that premium only after he had completed the steps of comparable company method and determined MVIC for WXL. Having determined WXL’s MVIC, Kern (i) subtracted the fair market value of WXL’s long term debt, and then (ii) increased the result by 30% to eliminate the minority share discount that inheres in the comparable company method. Respondent argues that Huck’s method improperly applies a control premium to WXL’s debt and, in doing so, inflates the value of the equity. I agree with Respondent on this point and will apply the eon-
trol premium only to the equity market value derived from the comparable company method.
I first observe that the difference in methodology has a significant impact on the result in this case. Using an adjusted median multiple of 1.74, an equity premium of 30% and a value of $15 .17 million for long term debt, Huck’s approach yields a fair value of the appraised shares equal to $540,000, while Kern’s approach yields a fair value of $312,650.
Huck’s result is approximately 85% higher than Kern’s.
The problem that both Huck and Kern sought to address is, as I have said, that the comparable company method of analysis produces an equity valuation that inherently reflects a minority discount, as the data used for purposes of comparison is all derived from minority trading values of the comparable companies. Because that value is not fully reflective of the intrinsic worth of the corporation on a going concern basis, this court has applied an explicit control premium in calculating the fair value of the equity in an appraisal proceeding.
See id.
at 6. It would seem to me to be particularly appropriate to do so where, as is true here, the comparable company method is the only method available to me to value the shares in question.
I am persuaded that Petitioners’ approach to the control premium is wrong. The observation that the comparable company method of analysis produces a minority equity value does not require that I change the methodology of that analysis. Rather, it requires only that I adjust the result derived from it to eliminate the implicit minority discount. Instead of adjusting the result, Huck undertook to alter, in a significant way, the methodology itself. Doing so, in my view, introduced an analytical distortion that, in this case, significantly overstates the value of the WXL equity. Neither Petitioners nor Huck has supplied authority in the valuation literature justifying this change in the accepted methodology. Because Huck’s methodology is inconsistent with the methodology followed in
Harris v. Rapid-American Corp.,
Del. Ch., C.A. No. 6462, Chandler, V.C., mem. op. at 6, 8-9, 1992 WL 69614 (April 1, 1992), and because it is otherwise hard to square with the comparable company method,
1 will not adopt it.
In
Rapidr-American,
the court valued the three wholly-owned subsidiaries of Rapid-Ameriean Corporation by multiplying the equity values of the three (in each case, derived by the comparable company method) times the indicated control premiums for each. Unlike Huck, the court in that case did not calculate the equity value of the subsidiaries by adjusting the multiple used to derive MVIC. Rather, MVIC was derived from the comparable company analysis. That amount was reduced by the value of the debt to determine the value of the equity. Only then was the control premium calculation performed. That is the method Kern used and the one I find appropriate to apply in this case.
D. Appropriate Amount of a Control Premium
Kern identified and used 30% as the appropriate control premium. Huck identified a 40% premium from his review of available information, which he reduced (as a part of his rejected methodology) from 40% to 30% to reflect the average debt-to-equity ratio in the industry.
He also recognized that, under our law, I must exclude from that amount elements of value to be realized from the Merger and opined that a small portion of the total was reflective of such amounts. He then testified that, at most, his premium calculation should be reduced to 25% to eliminate impermissible elements of post-merger value. Because I reject Huck’s method for applying the control premium, it is unnecessary to follow the first step of this process. Instead, I will reduce Huck’s 40% premium to 30% to reflect the elimination of impermissible elements of post-merger value.
See Hintmann v. Fred Weber, Inc.,
Del. Ch., C.A. No. 12839, Steele, V.C., mem. op. at 24 & n.52, 1998 WL 83052 (Feb. 17, 1998);
Kleinwort Benson Ltd. v. Silgan Corp.,
Del. Ch., C.A. No. 11107, Chandler, V.C., mem. op. at 6-10,1995 WL 376911 (June 15, 1995). Because Kern’s testimony also supports my use of the 30% figure, I determine to employ that figure.
E. Private Company Discount
In his analysis, Kern applied a final 20% discount, which is described in his report as a “Discount for Lack of Liquidity”. His report states that this discount is “based on market observations” and “reflects the fact that WXL International’s common stock is not publicly traded.” Respondent acknowledges that in
Cavalier Oil Corp. v. Harnett,
Del.Supr., 564 A.2d 1137, 1144 (1989), the Supreme Court held that it is wrong to apply a discount “at the shareholder level” for a lack of marketability of shares. Nevertheless, Respondent argues that a discount “for lack of marketability that affects all shares equally is an appropriate corporate level discount,” and claims that Kern “applied his liquidity dis
count to all the equity of WXL because WXL was not a publicly traded company.”
To the extent Respondent is arguing for the application of a “corporate level” discount to reflect the fact that all shares of WXL shares were worth less because there was no public market in which to sell them, I read
Cavalier Oil
as prohibiting such a discount. This is simply a liquidity discount applied at the “corporate level.” Even if taken “at the corporate levél” (in circumstances in which the effect on the fair value of the shares is the same as a “shareholder level” discount) such a discount is, nevertheless, based on trading characteristics of the shares themselves, not any factor intrinsic to the corporation or its assets. It is therefore prohibited.
Respondent argues, alternatively, that Kern’s discount is justified on the basis of Kern’s testimony about the existence of studies showing that privately held corporations sell at valuation multiples substantially lower than publicly held corporations. I agree with Respondent that, if the record supported this assertion,
Cavalier Oil
would not prevent taking such studies into account in determining the fair value of shares of a privately held corporation. Unquestionably, it would be appropriate to use a lower control premium than I have because the size of that premium is based substantially or entirely on data derived from sales of publicly traded companies.
The record in this matter is not adequate, however, to support the application of such a discount. The parties refer to Shannon P. Pratt, et ah,
Valuing a Business: The Analysis and Appraisal of Closely Held Companies,
355-56 (3d ed.1996). There, Dr. Pratt discusses a study published in
Mergerstat Review 199k
(Los Angeles: Houlihan Lokey Howard & Zukin, 1995) at 107, which shows that for certain years during the period 1985-1994, private companies were acquired at lower median price/earnings ratios than were public companies. I note, however, as does he, that during the last two years of the study, 1993 and 1994, the opposite effect is observed,
i.e.,
private companies sold for higher median p/e ratios than private companies. Discussing the general trend in earlier years, Dr. Pratt suggests that the relatively smaller size of the private companies may partly explain the observation. Here, of course, I have already discounted the median multiple derived from the comparable company method to account for WXL’s small size. Dr. Pratt also refers to an article discussing the issue. Z. Christopher Mercer,
Should “Marketability Discounts” Be Applied to Controlling Interests of Private Companies?,
in
Business Valuation Review,
June 1994. Mr. Mercer argues against the application of such a discount based on the available data.
Considering this evidence as a whole, I am unable to conclude that there is presently a sufficient theoretical support for the private company discount for which Respondent contends.
F. The Fair Market Value of WXL’s Long Term Debt
The parties do not agree on the correct way to account for WXL’s debt. Although .1 conclude that the difference between them is immaterial to my decision, I conclude that Respondent’s approach is more consistent with the valuation of WXL as a going concern. Kern used a figure of $15,169,057 (or $15.17 million) as the fair market value of WXL’s long term debt. This figure is taken from WXL’s 12/31/97 balance sheet, where it appears as “Intercompany Loan Payable.” Huck reduced long term debt by the amount of an asset also reflected on that balance sheet appearing as “Intercompany Loan Receivable” in the amount of $1,468,080 (or $1.47 million). Because he concluded that the elimination of this asset resulted in a working capital deficit of approximately $2.0 million, Huck subtracted that amount from his final result to derive the ultimately fair value of the equity. As I understand it, this difference in
methodology between the experts results hi a negligible difference in outcome. Using the parameters I have established in this opinion, the outcome changes from $310,500 using Kern’s approach to $308,-200, using Huck’s approach.
Because I believe it is more consistent with the valuation of WXL on a going concern basis, I will adopt Kern’s approach.
G. Interest
Section 262(i) provides, in pertinent part, that after appraising the shares:
The Court shall direct the payment of the fair value of the shares, together with interest, if any, by the surviving or resulting corporation to the stockholders entitled thereto. Interest may be simple or compound, as the Court may direct.
8
Del. C.
§ 262(i).
The statute gives the Court of Chancery discretion to award simple or compound interest.
Cede & Co. v. Technicolor, Inc.,
Del.Supr., 684 A.2d 289, 301 (1996);
Grimes v. Vitalink Communications Corp.,
Del. Ch., C.A. No. 12334, Chandler, V.C., mem. op. at 25 & n.65, 1997 WL 538676 (Aug. 26,1997).
The parties are in agreement that the legal rate of interest is the correct rate to apply in this matter. The issues are whether interest should be simple or compound and, if compound, the period of compounding. Respondent’s opposition to compound interest is only half-hearted and, in any case, is inconsistent with the trial testimony of its own expert. Mr. Kern, on cross-examination, gave the following testimony:
Q. To your understanding, all financial investments carry compound as opposed to simple interest; correct?
A. As far as I know, yes. Well, all— I’d like to restate that. “All” is pretty inclusive. I don’t know if that’s the case all over the world.
Q. In the United States you’re not aware of any financial investment that carries simple interest?
A. Generally, speaking, that’s correct.
This verity was recently recognized by Chancellor Chandler in
Onti, Inc. v. Integ-ra Bank,
Del. Ch., 751 A.2d 904, 926 (1999), wherein he said: “[i]t is simply not credible in today’s financial markets that a person sophisticated enough to perfect his or her appraisal rights would be unsophisticated enough to make an investment at simple interest.” To the extent it is necessary to look beyond Mr. Kern’s testimony, I also note that CTS did not make a good faith effort to value WXL in the merger, instead offering only $0.02 per share, or a nominal value of $10,000 for all of Petitioners’ shares. I find the appraised value of those shares to be 32 times as much and judge that compound interest on that amount is required fairly to compensate Petitioners for having to wait to be paid fair value.
The only testimony in the record about the period of compounding is Huck’s, who opined that quarterly compounding is most appropriate. Huck reached this conclusion because he thought the legal rate of interest (10%) most nearly resembles a return on a bond which, he stated, typically compounds quarterly. On cross-examination, Huck also stated that annual compounding is often used in certain methods of appraisal, such as the discounted cash flow analysis. He did not, however, abandon his view that quarterly compounding is appropriate here.
Giving weight to Huck’s testimony, I will award interest at the legal rate, compounded quarterly from the date of the merger until the date of final judgment.
IV. CONCLUSION
For all the foregoing reasons, I determine that the fair value of each share of WXL common stock, as of the date of the Merger, was $0 .6253 and, thus, will enter an order awarding Petitioners a total of $312,650 plus interest at the legal rate, compounded quarterly. The parties are directed to present an order of final judgment in conformity with this opinion.