Case study: Roche's Acquisition of Genentech

Mads Solberg Thomas, Luca Marmori, Rok Fer_jan

la) Why is Roche seeking tobuy the 44% of Genentech it does not own?

Roche is seeking to buy 44% of shares in addition to 56% that already owns. Main purpose of
purchasing rest of shares is to merge two companies and exploit the synergies.

Roche owned a majority stake in Genentech from 1990 and since then partnership was
considered as one of the most successful in pharmaceutical industry.

Every year less and less innovative products entered the market, so pharmaceutical
companies were forced to deliver growth through mergers and acquisitions of other
pharmaceutical companies. The old model, where Roche had a majority stake in Genentech
but left the company to operate on their owill, did not work anymore for Roche, so they
decided to do a horizontal acquisition and exploit Genentech potential. Acquisition would
bring synergies, reduce risk and grow.The t\vo companies would not be competitive anymore
between each other but would increase competitive strength towards other companies. All
make sense if the synergies are lower than premium paid.

1b) From Roche's point of view what are adva11tages of owni11g 100% of Ge11e11tech?

Owning 100% of Genentech would create the Largest biotechnology company in the world.
(NY Times:http://www.nytimes.com/2009/03/ 13/business/worldbusiness/13drugs.html).

Roche wou ld benefit from many of newly established synergies in total value of $Sb. The
amount is a consequence of manufacturing, development and M&D cost reduction.
Government and administration costs would reduce as well.

Complete ownership would enable Roche to have a complete access to all technology, R&D
projects and findings held by Genentech. Currently, the conflict of interest with Genentech's
minority shareholders prevent them the access to desired information.

Moreover I 00% ownership wou ld enable Roche to extract $9,Sb currently held by
Genentech in form of cash and cash equiva lents. This cash could also be used to party
repay the debt made by accusation.

Finally there's an opportunity to create an affiliate contract extension allowing Roche to

distribute Genentech best selling drugs and reducing or eliminating the risk of losing that
right (sell Genentech'sproducts out of the US's market).

le) What are the risks?

 Run of human capital and destroying the company culture
One of the main risks of seeking to buy the rest of the Genentech's shares is the
Squeezing out consequences (hostile takeover). By the fact that most of the minority
shareholders are Genentech employees, there is a high risk of destroying the old culture
and business practices in Genentech.This company gives importance to the family
environment in which they used to work, and that is something they wouldn't like to
change in their work place. Is very risky to squeeze out minority SH that are working in
that firm, it might be not ethical or in a certain way not according to rules of business; so
there is the bad chance that they change job or place to work, losing the most important
part of Genentech: the Human capital.

Paying higher premium than the real value of synergies

The value of synergies can be easily mispriced, or the desire to acquire the company is too
big, so the bidder pays higher premium than the value of synergies. Consequently the deal
would be considered as bad and price of Roche's stock would drop.

Debt obligation risk

One of risks that Roche is facing is also a debt obligation risk. For the deal to be successful
Roche would have to borrow around $30b of debt which could be more expensive due to
financial crisis and consequently more difficult to get and repay.

Loss of selling rights for non US market

During their 100% ownership of Genentech, Roche has signed affiliation agreement which
gives them the right to sell Genentech on non US market. As contract expires in 2015, and
Roche's doesn't have majority influence in board, they bear the risk oflosing the right.
After 2015 Genentech would be free to sell or auction the right which could be very
expensive for Roche.

Risk of bad test results

Furthermore if the acquisition would be successful before April 2009, Roche would have to
bear a risk that Avast in, Genentech's cancer drug facing test period, would not be
successfully tested and will be banned from sale, reducing Genentech 's revenues.

Risk of complementary drug entering the market

A lot of start-ups in biotechnological industry are working to develop new cancer drugs
which represent a risk of sales drop of Avastin and other best selling drugs which would
reduce Genentech growth prospects.
 2) As s minority shareholder of Genentech, what responsibilities does Roche have to
minority shareholders?

According to the affil iation agreement signed in 1999 between Roche and Genentech the
obligations to minority shareholders were clearly stated:

In case of friendly bid, the board approval would be sufficient and Roche would be able to
buy all shares for the same price.

I n case of hostile bid in which Roche would get 90% or more of the Genentech total shares
and hold it for more than two months, he was obliged to for squeeze-out the existing
shareholders and merge the company. According to Delaware law this wouldn't be necessary
but only optional measure.

As we explained before, minority shareholders are part of Genentech (workers) and it could
be very risky to not have a good relation with them. By the fact, Roche should try to explain
them all positive effect of the merger and how would synergies benefit to employees. They
should clear all Lhe fears of the employees, like losing job, losing power lo make decisions,
3)

Using a WACC of 9% and calculating synergies that are dependent on the merger we get the
following results:

2009 2010 2011 2012 2013 and

thereafter
FCF $ $ $ $ 475.58 $
138.05 362.36 436.47 488.81
Terminal value $
5,431.22
Discounted $ Shares 1052
3,529.92
NPV $ Shares to buy 463
l,423.28
Total value $ Value per $10.7
4,953.2 share
Discounti ng the Cash1flows with the WACC and treating the 2013 cash a flow as a perpetuity (not
using a long-term growth of 2%), we get a total value of 4.953 and 10.7 per share.

4)
Using the DCF technique on the Long-Range Plan (LRP) from 2009 to 2018, with a WACC of 9%
and long-term growth of 2% and annual growth of 7%, we get the following results:

Terminal Value (discounted) 42.785

NPV of Cash flow (2009-20 18) 29.890
Enterprise Value 72.676
To find the equity value of the firm we need to subtract debt and add back cash (which is one of the
incentives behind the acquisition) and securities.

Enterorise Value 72.676

Commercial Paper -500
Long Term Debt -2329
+9000
7? Cash And Securitisation 78.847
E uit Value

Dividing the equity value of the number of outstanding shares, we get value per share as a stand-alone
company.

Equity Value 78.847

Shares Outstanding 1.052
Value per share 74,9
Since the value per share is heavily affected by the annual growth and long-term growth, a matrix
using different levels of these two variables would give us a better range of possible values.
 A.growt 1% 1,5% 2% 2,5% 3%
h/LT growth
5% 61 63 66 69 72
6% 65 67 70 73 77
7% 69 72 75 78 82
8% 73 76 79 83 87
As seen above, using long-term growth from 1.5%-2.5% and annual growth from 6%-8% gives us a
plausible range from S67 to $83 per share. This seem to be in range with Greenhill's own calculations.

A source of concern for the cash flow is the WACC of 9% initially used in our calculations. The
WACC has a substantial effect on value, changing it to 8% or I0% gives ranging from 87 to 65. Let's
have a look at the fundamentals behind the WACC

Market risk premium 7.1%

Beta (as of Julv 2008) 0.26
Beta of Equitv 0.3066
Treasury yield (as of July 2008,10 years*) 4.01%
Cost of capital 6.18%

Commercial debt 500 2.08%

Long term debt (AAA) 2329 5.67%
Total debt 2829 5.51%
Short and long term debt 2.829 15.21%
Shareholders' equitv 15.761 84.8%
WACC = 0.055 1 *( l -0.35)*0.152 1+ 0.848*0.0618 = 5.744%
As seen above, the WACC is quite different from the initial 9% used in the case. We can then make the
assumption that the Greenhi Uteam has used an industry Beta based on comparable firms.

Company Amgen Gilead Celgene Genzyme Biogen Cephalon

Beta 0.454 0.894 0.766 0.658 0.767 0.384
Average 0.653833
Median 0.712
Adjusted Beta 0.80992 "the blume effect"
Usmg the median Beta (the average 1s prone to outliers) and adjust mg 11 for "the Blume effect" we get an
industry Beta of 0.80992. Unlevered Beta is about 0.68. Implementi ng these variables into the cost of capital
we get a new WACC of around 8%.
 The rather conserva tive outlook of a 9% WACC makes sense from Roche's point of view. This is a
risky investment where future earn ings depend on the pipeline production of Genentech. Increasing
the WACC rather than decreasing takes this into account as stand-alone valuation. Here we would
recommend that future likelihood of successful drugs should be treated as real-options (put- option)
to assess the riskiness in a better way.

Before we start exploring the multiples, we should take a closer look at the characteristics of other
firms in order to assess whether they are comparable or not.

Companies in the same industry: all of the firms in exhibit 13 are Biotech companies as
Genentech.

Of similar size:in terms of capitalization, only Amgen (A) (57.396) and Gilead (G) (46.073)
come cl ose to Genentech 's (88.546). The rest of the compan ies have around '/.i or less
capitalization than Genentech. These firms may be less liquid and have a higher default risk
and should be left out the analysis.

Similar expected growth and risk:

Company (Genentech) L.T growth Beta

Am2en 10.5% (19.2) 0.454 (0.26)
Gilead Sciences 24.7% ( 19.2) 0.894 (0.26)
Average 17.6% 0.674
"A" has a lower forecasted growth and lower Beta than "G"and the opposite 1s true for "G''. Despite
these differences, both companies should be taken into account as comparable's,
although "G" seems to be the compa ny that resembles Genentech the most.

P/E multipl es

Trailing Forward
EPS EPS
EBIT 5241 5638
Tax 1834.35 1973.3
Earnings 3406.65 3664.7
Shares 1052 1052
EPS 3.23826046 3.48355513
Amgen 12.6 40.80208 12.1 42.15102
Gilead 25.4 82.25182 22.1 76.98657
Average 19 61.52695 17.1 59.56879
As seen above, "G"'s forward PIE (better estimate of value) gives a pnce per share of 76.98 which is in
the same range as the DCF in question 4. Adding back debt and subtracting cash and securitization
gives an enterprise val ue of about 74.818. "A"'s low &rrowth prospects lowers it PIE multiple and
makes it a poor comparison.

EBITDA multiple
 The EBITDA is capital-structure neutral and is thought to be a more robust measure for cross
companies application.

2008(trailing) 2009(forward)
EBTTDA 5833 6195
Multiple
Amgen 9.5 9.3
Gilead 18.3 16.2
Average 13.9 12.75
Industry 12 Il.3
median
Enterprise
value
Amgen 55413.5 57613.5
Gilead 106743.9 100359
Average 81078.7 78986.25
Industry 69996 70003.5
median
The results show quite different enterpnse va lues. "A" and "G" shows total different forward values
from the DCF evaluation . But looking closer at the industry median gives an enterprise value that is
about 2.7% different from the original calculations in the DCF model.

Enterprise Value/Revenue

2008(trailing 2009(forward)
Revenue 13418 13535
Multiple
Amgen 4.1 4
, Average 6.7 5.9
Industry median 4.8 4.3
Enterprise value
Amgen 55013.8 54140
Gilead 124787.4 105573
Average 89900.6 79856.5
Industry median 64406.4 58200.5
The revenue multiple is poor comparable and wou ld not be taken i nto account. Th.is multiple is
usually used for companies with no earni ngs.

PEG

It is not recommended to use the PEG multiple.

Summary: The DCF calculated in question 4 gave an estimated enterprise value of 72.676.
Using Gilead's P/E multiple (same expected growth) gave an enterprise value of 74.818. The
EBITDA multiple gave different enterprise values ranging from 57.000 - I00.000, but the
industry median
multiple implied a value of 70.000. The big differences in valuation multiples high light the difficu lt
task of using comparable's as a proxy for enterprise valuation. This is especially true since the
companies used in the calculations are not adequate in terms of size and risk. Despite these flaws, the
multiples confirm that the injtial valuation of Genentech as a stand-alone company seems to be witmn
range, and that the offer of $89 per share is considered to be a "fair" price.

In addition, the differences in multiples also confirms the vast differences in Wall street's consensus
about price targets (exhibit 14).

More overly, the offer of $89 per share for Genentech is within range of how much Roche
would be recommended to pay:

Iron law of Merger&Acquisitions

Closing Price 81.82
Offer 89
Shares left to buy 462.88
Value of offered price 41196.32
Value at closing price 37872.8416
Value of premium 3323.4784
Value of synergies 4953.205073
Syn. Excess of premium 1629.726673

There is excess over premium of 1629. This means that Roche 's Shareholders don 't
lose value on the deal as the expected synergies of the acquisition is bigger than the
premium paid.

6a) how has thefinancial crisis affe cted Genentech's value?

The demand oflife-saving drugs was not cycl ical, thus Genentech's near-term revenues
and earnings were not much affected by the financial turmoil. However, the stock price
dropped in line with their industry. From exhibit 15: Genentech stock has peeked in
August 2008 (94$ per share) after the first Roche's offer, and fell sharply from September
on with bottom as low as $70 in October and November.
Crisis had an impact on Genentech's cost of capital and cost of debt: As cost of capital
increased, WACC with which the companies' cash flows are discounted increase as well. In
addition, risky markets make borrowing more expensive as there is an »investor's flight« to
safer assets. In this case the analysis should be adjusted and discounted with appropriate
discount factor.

6.b) What changes in valuation assumptions occured between June 2008 (LPR) and
November 2009 (NFM)?

There were many differences in NFM fi-om the LPR valuation assumptions. The first year of
forecast began from a higher base. Total revenue was $14.118b in NFM vs. $13.535b in LPR.
In NFM valuation the period of the detailed forecast was extended from 2018 to 2024 with
6.9% cumulative average growth rate assumed to prevail over the 16 year period. The
specific cost (9.5% vs. 10.2%) and expense ratios differed but total cost was almost the same
(52.1% vs.52.3%).The assumed tax rate (35% vs.30.7%), depreciation expense (3.3% vs.
2.9%), changing in net working capital (1.3% vs. 0.7%), and capital expenditures were all
lower (3.9% vs. 3.1%). NFM DCF valuation included $l 1.4b change reflecting the
capitalized value on this expense. NFM valuation included an $8.2b positive adjustment to
enterprise value reflecting the value of >>opt-in« rights to the product pipeline when the
current licencing agreement expired (2015).
Terminal value $ 180,359.58
Discounted $ 49,515.56
NPV $ 57,069.83
Stock option expense $ 11,420.00
opt in rights $ 8,190.00
Enterprise value $ 103,355.40
S.T debt $ 500.00
LT.debt $ 2,329.00
Investment and Sec. $ 9,000.00
Value of equity $ 106,296.40
Shares $ 1,052.00
Value per share $ 101.04
Annu al-Growth/L.T 1.5% 2% 2.5%
1!:rowtb
6% 89.32 92.27 95.68
6.9% 97.69 101.04 104.91
8% 109 113.07 117.58

The new Enterprise value is up from 72.676 to 103.355 in the new NMP plan. Adjusting for
debt and cash we get value of equity equal to 106.296 and a value per share of 101.04. This is
a substantial difference from the init ial LRP as a stand-alone company. This plan may be a bit
more biased than the in itial LRP plan, but we th ink that the opt in rights of 8.190 is a
welcoming addition to the DCF model. This can be seen as an extra value from Roche point
of view and should be taken into account.

The value-matrix adjusted for different levels of annual growth and L.T growth gives us a
range rrom 89.32$ to 117.58$. The special board's appraisal of a price of around $112 per
share seems, in this case, is in the upper levels of the valuation.

7) How did Genentech 's board of and management respond to Roche's off er of USD
89per share?

To protect the minority shareholders, Genentech's board of directors delegated responsibility

for appraising the offer to a special committee (three independent directors). The committee
quickly rejected Roche's offer and waited until November to put forward its own view of
price: $112 to $115 per share. In addition, the special committee retained Goldman Sachs as
financial adviser and Latham and Watkins as legal counsel. Their first action was to
recommend employees retention and severance plans; Roche quickly agreed. But the
committee announced that its members had unanimously determined to reject the 89$ per
share proposal as substantially undervaluing the company. Genentech decided to left the
doors open to future offer.

In November Genentech senior managers presented a new financial forecast to Roche's senior
managers (see 6b). Goldman Sachs provided a new cash flow analysis based on new data,
resulting in a valuation rrom 112$ to 115$ per share. The NFM was updated after Roche's
offer and was concluded only five months after Genentech submitted its LRP to Roche in
June. Some person felt like it was a biased plan and that LRP was more reasonable and
unbiased view of future prospects.

8a) what should Franz Humer do? Specifically, should he launch a tender offer for
Genentech 'sshares?

Humer has three alternatives:

First he could raise the offer price close to 112$ even if there was no guarantee that
special committee wou ld agree to any compromise.

Raising the offer to 112$ per share would conflict with the Iron mle of Mergers &
Acqu isitions, where the value of synergies needs to be bigger than the premium. With
a current market price at 82-84 per share, the premium wou ld be around 16$ per share
and thereby substantially overpricing the synergies. In addition, according to our own
calculations, the proposed price of 112$ seems to be in the higher levels of our range
matrix (see 6b).

Roche could make a tender offer for Genentech's shares; so take the issue of valuation
directly to Genentech's shareholders, bypassing the board and the special committee.

A tender offer would be made on the assumption that takes the LRP plan into account.
The NMP plan is, according to our assumptions, a bit biased. First of all, according
with our calculations, the range between 90$- 117$ conflicts with the wall -streets own
price targets. In addition, the extended long term plan to 15 years from the original I 0
years is more prone to errors since uncertainty increases with each year of additional
cash-flow projections. As a result of this, the cash-flow NPV changes from around
53.000 to 35.000. Moreover, the new tax rate at 30.7% heavily conflicts with
Genentech's own average tax rate of 63% over the last 11 years. The median rate of
36% clearly h ighlights the difference between the two assumptions. Given a
normal ized tax-rate of 35% (as in the initial calculations), the base-case value per
share drops from $101 to $94.55.

Given the LRP plan as a base case for our valuation, and given that there is to be a
tender offer, we would recommend a price per share based on a base-case range of
$70- $83 per share. With synergies of $I 0.7 per share the maximum tendered offer
should be in the range of $70-$93.7. Since the min imum offer is the market price
($82-$84), the tender offer should be in the range of $82 to $93. 7 per share.

Risks:
•A tender offer would be viewed as hostile move and might solidify opposition to
Roche among Genentech's managers and employees. Th is would make later
integration of the two companies more difficult and costly, and might drive some of
Genentech's star scientists and managers into the arms of riva ls (loss of Human
Capital).
•Roche wou ld be able to buy only shares actually tendered and wou ld need to comply
with the terms of the affiliation agreement in order to squeze-out the remaining
shareholders in a second step merger.
•Some shareholders might decide to not tender their shares, even if they thought the
price was fair: Risk of not taking over.