Novo Nordisk. Ready for a new start INDEPENDENT RESEARCH UPDATE. Healthcare. BUY vs. NEUTRAL. Fair Value DKK1180 vs. DKK1000 (price DKK953.

Transcription

1 INDEPENDENT RESEARCH UPDATE 19th December 2013 Healthcare Ready for a new start Fair Value DKK1180 vs. DKK1000 (price DKK953.50) BUY vs. NEUTRAL Bloomberg NOVOB DC Reuters NOVOB.CO 12-month High / Low (DKK) 1,070 / Market capitalisation (DKKm) 421,936 Enterprise Value (BG estimates DKKm) 405,750 Avg. 6m daily volume ('000 shares) Free Float 71.4% 3y EPS CAGR 14.8% Gearing (12/12) -41% Dividend yield (12/13e) 2.23% We have had the opportunity to meet with vo-rdisk several times over the last month at our HealthCare Conference and during the Investor Day and, after a difficult 2013 fiscal year, we believe the stock should perform much better in the coming quarters. vo- rdisk has an impressive pipeline in diabetes, has several free options outside diabetes and enjoys a tremendous footprint in emerging markets. The upcoming share split 1:5 on 2 January 2014 offers an interesting entry point to play the come-back. YE December 12/12 12/13e 12/14e 12/15e Revenue (DKKm) 78,026 83,670 88,397 94,931 EBIT (DKKm) 29,474 31,498 33,599 37,850 Basic EPS (DKK) Diluted EPS (DKK) EV/Sales 5.2x 4.8x 4.5x 4.2x EV/EBITDA 12.5x 11.8x 11.0x 9.7x EV/EBIT 13.8x 12.9x 12.0x 10.4x P/E 24.4x 20.2x 18.7x 16.1x ROCE /12/13 Underperformer in vo-rdisk did badly in 2013 as very early in the year it received a CRL from the FDA about Tresiba. This delay by 3 years is a significant setback for the company as its main driver will not play out as expected. Concerns about the GLP1 class also made Victoza less buoyant whereas the slowdown in emerging markets somewhat affected the top-line. Late in the year, the company lost a big contract with Express Scripts in the US, saw the first Prandin generics launched and formed an uninspiring first guidance for By chance 2013 is coming to an end! NOVO NORDISK 'B' STOXX EUROPE 600 Source Thomson Reuters Rebased expectations. Behind our Neutral rating in 2013 were the overly optimistic view and too high expectations about both Tresiba and liraglutide 3mg. We were not comfortable about how the market was focusing on these two items. w everybody knows that vo-rdisk will struggle with Tresiba in Europe (to get a high price) and in Japan (to compete against strong Lantus) while it is postponed in the US until 2016 at best. Everybody can also judge the strengths and weaknesses of Lira in obesity and we are increasingly confident that it can be a drug. But beyond these two, IdegLira and semaglutide are also becoming stronger. Medium-term recovery, short-term catalyst. It is our understanding that vo-rdisk formed a cautious guidance for 2014 that it is likely to beat. We are making small adjustments, introducing Lira obesity (with 30% PoS), rolling over our numbers to 2014 and using a new hypothesis for the risk-free rate and risk premium. Our new FV is DKK1,180 (or DKK236 after the share split by 5 on 2 January 2014). This is a new BUY. Analyst: Sector Analyst Team: Eric Le Berrigaud Mathieu Chabert 33(0) Martial Descoutures eleberrigaud@bryangarnier.com r r

2 Company description vo-rdisk is the worldwide leader in Diabetes where it benefits from a strong underlying volume growth due to epidemiology, from geographic expansion, mix effects and innovation. Growth has further accelerated in recent quarters as the Danish company successfully added Victoza, a GLP-1 analog to its wide range of insulins to strengthen even further its positions within the Diabetes field. The drug is quickly gaining shares. w the company is mainly devoted to switching its insulins to a new generation of products. In parallel, the company is building a franchise in Hemophilia that could significantly leverage margins if voseven is transformed into a diversified product line. We see vo-rdisk as a success story that is unlikely to interrupt in the short term. Profit & Loss Account (DKKm) e 2014e 2015e 2016e Sales 60,776 66,346 78,026 83,670 88,397 94,931 98,143 Change (%) 19.0% 9.2% 17.6% 7.2% 5.6% 7.4% 3.4% EBITDA 21,358 25,074 32,374 34,498 36,599 40,850 43,001 EBIT 18,891 22,374 29,474 31,498 33,599 37,850 40,001 Change (%) 26.5% 18.4% 31.7% 6.9% 6.7% 12.6% 5.7% Financial result (605) (270) (1,663) 1, ,060 1,320 Pre-tax profit 18,286 22,104 27,811 32,598 34,499 38,910 41,321 Tax 3,883 5,007 6,379 7,400 7,762 8,560 8,884 Net profit 14,403 17,097 21,432 25,198 26,737 30,350 32,437 Restated net profit 14,403 17,097 21,432 25,198 26,737 30,350 32,437 Change (%) 33.8% 18.7% 25.4% 17.6% 6.1% 13.5% 6.9% Cash Flow Statement (DKKm) Operating cash flows 19,382 20,940 21,940 28,198 29,737 33,350 35,437 Change in working capital (85.0) 648 1, , Capex, net 5,579 3,459 4,070 3,500 3,500 3,000 3,000 Financial investments, net (8,820) (11,102) (12,398) (14,000) (10,000) (10,000) 0.0 Dividends 4,400 5,700 7,742 9,715 11,339 12,032 13,657 Net debt (13,827) (15,205) (16,488) (16,187) (20,361) (27,450) (45,262) Free cash flow 16,303 18,301 18,842 23,414 25,514 29,120 31,469 Balance sheet (DKKm) Shareholders' funds 36,965 37,448 40,632 42,115 47,513 55,831 74,610 +Provisions 10,101 11,909 11,055 11,055 11,055 11,055 11,055 +Net debt (13,827) (15,205) (16,488) (16,187) (20,361) (27,450) (45,262) =Invested capital 33,239 34,152 35,199 36,984 38,207 39,436 40,403 Fixed assets 24,109 25,107 25,506 26,006 26,506 26,506 26,506 + Working capital 9,130 9,045 9,693 10,978 11,701 12,930 13,897 =Capital employed 33,239 34,152 35,199 36,984 38,208 39,438 40,406 Total Balance sheet 61,402 64,698 65,669 67,547 72,727 80,857 97,714 Financial Ratios Operating margin Tax rate Net margin ROE (after tax) ROCE (after tax) Gearing Pay out ratio Number of shares, diluted (m) Data per Share (DKK) EPS Restated EPS Core EPS Change (%) 37.9% 22.1% 30.3% 20.8% 8.2% 15.8% 6.9% Book Value Operating Cash flows FCF Net dividend Source: Company Data; Bryan, Garnier & Co ests. 2

3 was a nightmare 1.1. Tresiba hit first and foremost A big setback in the US Already in 2012, there was some bad noise around Tresiba whose PDUFA date was delayed until after an Advisory Committee that was scheduled for 8 vember On that day, the experts from the EMDAC (Endocrinologic and Metabolic Advisory Committee) voted by 8 against 4 in favour of a marketing approval of Tresiba, recommending a post-marketing outcomes study. But unexpectedly, on 8 February 2013, vo-rdisk received a complete response letter (CRL) from the FDA in which the agency asked for additional data from a CV outcomes trial (CVOT). Given the split in sales for modern long-acting insulins (see Fig.1 with Lantus and Levemir), a 3-year delay in the US is a major setback for the company, all the more so that it had planned for a long time to fight against current price levels in Europe, asking for a significant premium that would dampen the take-off and ramp-up of the drug in those markets if not jeopardise the launch in some countries. Fig. 1: 9-month geographical sales split Lantus (left), Levemir (right) Emerging markets 15% Other 5% Western Europe 14% United States 66% Europe 26% rth America 58% International 11% Japan & Korea 3% Region China 2% Source: Company Data Since then, the DEVOTE cardiovascular outcomes trial was initiated and started enrolling patients in October This is a double-blinded randomised trial whose primary endpoint will be time to first MACE (CV death, non-fatal MI, non-fatal stroke). 7,500 type II diabetes people with moderate to high CV risk will be recruited and split into two arms to receive either Tresiba or Lantus daily in the evening. vo-rdisk believes that although the full trial will take 4 to 6 years to complete, a new filing could take place earlier, i.e. based on interim data which are expected to be available after 2 to 3 years. To be on the safe side, it is reasonable to expect a new NDA to be filed in H Fig. 2: reason to be too worried about CV risk with Tresiba Source: Company Data; Investor Day December

4 The trouble with the US delay for Tresiba is that should the drug enter this market in 2017, the competitive environment is likely to have significantly changed by then. With the original entry date in late 2013/early 2014, vo-rdisk would have faced a very classical competition with Sanofi s Lantus with real chances to win additional market share compared to what has been achieved with inferior Levemir. In 2017, the long-acting basal insulin market in the US is likely to be the scene of fiercer competition on two fronts: the first risk should come from Lilly one way or another, i.e. its own branded insulin currently in phase III but admittedly this seems to be facing some safety issues (liver enzyme elevation) and/or a first glargine biosimilar but which surprisingly has been filed in Europe but not yet in the US; the second risk might come from Sanofi itself as the market leader is developing a nextgeneration product known as U300. This new proposition, priced on par with Lantus, should see some arguments prevailing such as true once-a-day formulation, flexibility of use and at least a trend towards a decrease in severe hypoglycaemic events. Sanofi s marketing muscle will weigh as much as clinical data will but in the end all this should make vo-rdisk s task more difficult Green lights outside the US but limited opportunities It certainly does not offset the delay in the US, but it is worth saying that Tresiba was approved in 2013 in several other countries around the world, including in the European Union and Japan. As for Europe, vo-rdisk is engaged in a fight that it is unlikely to be successful in the majority of the countries where it is taking place. The group agrees to say that it is going to be a challenge in Europe but it wants to stay firm to address what sounds like a market abnormality: where the US pays USD8 per day for insulin treatment, Japan USD4 and China around USD3, Europe pays less than USD2. Innovation is not rewarded. Certainly true but very effective and safe drugs are available at that price and so why should Europe pay more unless for very significantly improved products? vo-rdisk must convince that Tresiba deserves such a claim. This is far from granted. As a reminder the price premium that is looked for is 70% (i.e. the same price as Levemir seven years ago). During the Investor Day, vo-rdisk expressed confidence and satisfaction about take-off in markets without access restrictions. Indeed Tresiba did start pretty well in Switzerland, Mexico and Japan, capturing 7-10% market share in the first few weeks. This is encouraging. Nevertheless, this tends to stabilise notably in Japan after the initial ramp-up and conversion of available patients, i.e. those looking for a new drug (the vast majority of patients switched from other insulins) and those treated by physicians convinced by the new proposition. Tresiba will be an opportunity outside the US, Japan and Europe and notably in emerging markets but it is unlikely to be the central piece of the strategy. Overcoming the difficulties in the US is thus key for the drug. We have significantly cut our estimates for the drug and only have USD850m in sales in 2020, which compares to EUR5.8bn for Lantus in We expect Levemir to come down from DKK11.6bn in 2013 to DKK9.7bn during the same period with a peak in 2016 of DKK14.8bn due to the continuous growth in the US. 4

5 1.2. But other events also made 2013 a difficult year A slowdown in GLP1 market growth In 2012, in order to help the investment community model the growth rate of Victoza, vo- rdisk s management said that it was expecting the drug to grow by DKK m sequentially from one quarter to the next. While it more or less achieved this goal in 2012, it did not in 2013 and missed the target in Q1 and Q3 with currencies further weighing on the reported numbers. The group said that requests for a global review of safety data for the whole class from regulatory authorities around the world generated some caution among physicians and weighed on prescriptions. Victoza further expanded its leadership however compared to former Amylin s products Byetta and Bydureon (now promoted by BMS and AstraZeneca without any big change in trends) while Lyxumia has just been launched in a few markets. However restrictions have now been removed and vo-rdisk is hopeful that it will be able to revive growth for Victoza in the coming quarters. Also to note is that additional promotional efforts have been made to support the brand, including an expanded sales force and a new DTC campaign and advertisement supports with very good first feedbacks and returns. This time, to its advantage, it is good to note that vo-rdisk will be free of any competition from Sanofi with Lyxumia in the US for another couple of years as the NDA was withdrawn and will be re-submitted in 2015 when interim ELIXA data are available. Only the loss of the contract with Express Scripts in the US (see below) will act negatively on growth in That said, in the GLP1 market, vo-rdisk has always said that competition would be more likely to come from Lilly than from Sanofi. In October, Lilly filed its once-weekly GLP1 dulaglutide in the US and in Europe thus making regulatory newsflow likely to happen sometime in H2 2014, including a very likely AdCom. Actually the very key question is whether Lilly and dulaglutide can succeed where Sanofi and Lyxumia failed, i.e. winning approval without safety data from an outcomes study. Although the entry of Lilly in the GLP1 market would be accretive for the class and overall good for its positioning in the treatment paradigm of type II diabetes, a once-weekly proposition backed by the FDA safety-wise would be more likely to be detrimental to Victoza in our view. With no doubt this is going to be one of the most important indirect events to pay attention to as catalysts for vo- rdisk The loss of a significant contract with Express Scripts The annual review of Express Scripts drug list made many victims this year as the biggest PBMs are big enough now to increase their pressure on drug companies any time there is an opportunity (i.e. each time there are alternatives within a class) to get lower prices. And it looks as if for each category at least one company is ready to agree on lower prices to get higher volumes. In this specific case, the selected drugs moved from a co-preferred status to an exclusive one after agreeing on higher discounts. So, in a surprising move we must say, Express Scripts decided to remove Victoza from the preferred drug list as of January 2014 in the GLP1 class in favour of Bydureon and Byetta (BMS/AstraZeneca), whereas volog was excluded in the short-acting insulin category to the benefit of Humalog (Lilly). 5

6 As the largest PBM in the US ahead of CVS Caremark, Express Scripts decision is expected to have a top line impact at the level of the whole group of 1-2% Guidance for 2014 seen as shy and not matching consensus Of course, this was included in the guidance formed at the time of Q3 press release for next year which calls for revenue and operating income growth in high single-digits in local currencies. Together with a meaningful negative currency impact, this was guiding towards numbers that were a bit short of what the market was expecting for 2014, in particular at the operating income level as it suggested a flat margins. Be it with Prandin generics or the loss of the Express Scripts contract, it was thought that margins could have been secured as the product mix could be further improved. w despite additional progress on COGS, development costs and marketing expenses could grow a touch faster than sales, hence the guidance for profits. It also factors in a flat year for very profitable voseven compared to bullish 2013 when the drug rose 11% (over the first 9 months). In our view, this is a conservative first guidance that is unlikely to be revised before the Q1 numbers are unveiled. A clear upside to the numbers would be another year of significant price increase for long-acting insulins in the US, driven by Lantus and followed by Levemir, which is neither in the guidance nor in our numbers but a reasonably likely scenario. Inroads with Tresiba and Ryzodeg in new markets will also be key to monitor. To a lesser extent new formulations of Bydureon, Lyxumia and dulaglutide might impact Victoza s growth rate. All in all, we have adjusted our numbers in accordance with the guidance which gives an EPS estimate of DKK51 to 52, up 9% to 10% in reported terms with a negative 2 to 3% impact from currencies. We see the consensus very much in line with this type of expectation and so no risk of disappointment barring unforeseen new events. 6

7 2. Focus is shifting towards new items 2.1. Liraglutide 3mg in obesity still a topic The big picture provides mixed feelings From a philosophical point of view, we are usually very reluctant to include obesity drugs in any sales forecast model given the unpredictability of regulatory decisions and the high probability of failures in the field historically. Moreover, although concerns about the GLP1 class have been removed and the benefit-risk balance reaffirmed as positive in type II diabetes, this will not necessarily be the case at a high dose in a perceived milder indication like obesity. Last but not least, available clinical data so far provide mixed feelings about the clinical value of liraglutide 3mg in terms of efficacy once it is placebo-adjusted. As seen in Fig.3 below, with the exception of the maintenance trial, differences between liraglutide and placebo are in a range of 3.9% to 5.4%. This is the most relevant indicator of efficacy in our view. vo-rdisk is also used to sharing figures that emphasise the much higher proportion of patients losing more than 5% and 10% of their body weight respectively. But should many more people lose 10% of their body weight with Lira 3mg, if the difference in the end is only 5%, then this could mean some are not benefiting at all from the drug. Fig. 3: SCALE trials demonstrate weight loss but of limited magnitude Source: Company Data, Investor Day December 2013 During its Investor Day in December however, vo-rdisk made decisive comments about its investigational drug for obesity, addressing questions about biomarkers. Although there is no relevant biomarker to detect beforehand who might benefit this drug, it has been observed that early response to treatment highly correlates with the highest weight loss levels at week 56 as shown in Fig.4. This could be a key observation and a major take-away when discussing the value proposition of the drug with healthcare authorities for approval, label and post-marketing requirements, but also from a marketing perspective as it is fair to expect the company to implement a strategy that, with regard to an expensive daily cost of treatment (around USD25/day in the US), could make the drug available for free during the first 12 weeks of treatment and then payable to responders. By the way, existing weight-loss drugs are already offering free sampling but usually for two weeks. 7

8 Fig. 4: Early detection of high responders Source: Company Data, Investor Day December Awareness is on the rise With 107 million obese people in the US using the strict definition of a BMI above 27 kg/m² or a prevalence of 38%, and also very high numbers in Central and Latin America as well as in the Middle East, obesity has become an epidemic, and associated with the increased prevalence of many other diseases with a high burden, thus representing a very significant total cost for society. And it is fair to say that after years and decades of arguments of whether to consider obesity seriously as a disease rather than as a condition, although it is still compulsory to first manage food intake and activity, significant progress has recently been made which could help introduce drugs more often into a treatment algorithm. It is probably not by chance that two drugs were approved in 2013 by the FDA to treat obesity: Vivus Qsymia and Arena s Belviq. Compared to these two recent newcomers which so far have registered limited success, vo- rdisk will try to position Lira 3mg in a much more medically-oriented way to build on its comprehensive and untapped clinical dossier but also to take into consideration the requirement for daily injections that de facto excludes weight loss management for cosmetic purposes. But on the other hand this is the best possible argument to prevent drug users from overdosing. We would advocate that, considering the more friendly and open-minded approach towards treating obesity with medication, the ideal target for a new entrant would be the upper part of the market for both physicians and patients. In other words, the more severely affected the patients, the higher the chances to make them become compliant to treatment. From a physician s perspective, it is in vo- rdisk s interest to target weight management specialists and frequent prescribers. And that is exactly what the group presented as a strategy, mentioning an initial 20,000 base of prescribers when crossing GLP1 use and scripts for anti-obesity medications Sales potential is (still) very significant Whatever the number of discounts and stratum of caution (probability of success, addressable market, market penetration ) we factor in to model revenues for Lira 3mg, it is a considerable market opportunity, notably in the US. 8

9 For that reason, but also for all the above-mentioned arguments and because in the end the slight increase in heart beat rates does not look as if it will call into question the global safety/tolerability profile of the drug, we have decided to introduce Lira 3mg into our model. We have tried to simplify the model which could become very complex in this type of indication. So starting with the 107 million obese patients in the US and using the estimated 5% target vo- rdisk spoke about, leads us to an addressable 5.35 million people for the US only. We took the 33% of patients losing at least 10% of their body weight with intend-to-treat methodology (SCALE Obesity and prediabetes) as the real target to assess the size of the population able to pay. Then we estimated that the penetration rate would progressively ramp up to reach 15% in 2020 and beyond. We set the price as a function of the current price once it is dose-adjusted, representing a total annual cost of USD7,908 without further increases. Lastly we assigned a 30% probability of success to reflect the still remaining uncertainty about approvability of a high-dose GLP1 in obesity. In the end, probability-adjusted, this nonetheless represents an additional extra USD628m in sales opportunity in 2020 or an upside to the FV of DKK55/share (while ex-us potential is excluded from our calculations). As a reminder, the filing for Lira 3mg is expected to take place sometime around the turn of the year IdegLira and semaglutide should become increasingly important IdegLira: Tresiba s delay should not hide strong clinical data Indeed Tresiba s delay in the US made another significant victim for vo-rdisk: the fixed-dose combination of the long-acting insulin with the GLP1 analog liraglutide: IdegLira. Rightly or wrongly, many observers including ourselves think that the combination of putting together the two injectable classes of drugs will play a very significant role in the coming years as the proportion of patients who need both products (whichever one is used first) will increase while this option is easier to use, improves compliance and reduces the cost for the patient and for society. However, such products have not generated much excitement in the investment community so far. It is fair to say that they have accumulated setbacks, with LixiLan dropping its Fix-Flex device, postponing its entry in phase III to H while IdegLira was de facto delayed by Tresiba s CRL. Within the medical community, at ADA and EASD where the DUAL I and DUAL II results were presented, two distinct categories emerged: bullish on one side, impressed by the strength of the data, and bearish which stated that few people deserved such powerful drugs, whereas dual-acting products increased the risk of malpractice. Actually IdegLira is unlikely to be first choice after OAD s failure. It makes no sense. Within the type II diabetes paradigm of treatment, vo-rdisk suggests that it might be positioned as an optimiser once both GLP-1 and long-acting insulin are used by a given patient and before he needs to intensify with a second insulin. We agree and would consider that this is possibly a worst case scenario as some patients might directly go from a GLP-1 to the fixed-dose combination unless they require some time to get accustomed with insulin titration. 9

10 We believe that clinical data will drive adoption and commercial success and they are simply amazing. From an efficacy perspective, as seen on Fig.5, they are untapped: no other single drug provided a 1.9% reduction in HbA1c and saw 81% of patients reaching a level of 7% or below. Simultaneously, it is reducing the GI side effects of Victoza (8.8% nausea vs 19.7%, 3.9% vomiting vs 8.5%) and offering a neutral to slightly positive impact on weight compared to insulins. Interestingly, the average dose of each component was reduced compared to what is needed when monotherapies are used. For instance, the average insulin daily dose was 38 units with IdelgLira compared to 53 units in the degludec arm after 26 weeks in DUAL I. Fig. 5: Impressive clinical data for IdegLira Source: Company Data, Investor Day December 2013 IdegLira was filed in Europe in May 2013 on the basis of the DUAL I and DUAL II results. However, key phase III trials are still ongoing which should provide very meaningful incremental data not only for the label of the drug but also for its intrinsic value and its commercial potential. We are talking here about data assessing the opportunity to switch from single therapies. DUAL III and DUAL V will be of particular interest from this perspective and the first of the two is expected to report sometime in H (switch from GL1, 26-week trial, 429 patients) and the second in We have assumed a very progressive ramp-up of IdegLira in the various markets where it will be sequentially approved and launched starting in Q We have USD150m in sales in 2016 and USD1bn in Semaglutide: A must have for the franchise The debate about how and when GLP1s will be used is far from over today and it is premature to make any formal and definitive statement about this topic as too many swing factors are still ahead of us. Short term, we would like to make clear whether the GLP1 market is a single and unique market or if there are different products within this family, calling for different strategies and positions. There is an ongoing debate in the space, led by vo-rdisk on one hand and Sanofi on the other. Their respective goals and interests are well understood but it is so far unclear which option should prevail, although we are progressing in our understanding of the matter. To make it simple, vo-rdisk does not consider the GLP1 class as different from any other segment of the diabetes market, although there are obvious differences among products. On the other hand Sanofi (and its partner Zealand) are trying to position Lyxumia at a different level compared to Victoza. When the first says that GLP1 is treatment of choice when OADs are no longer sufficient to 10

11 control glycaemia, as a first injectable drug, the second is preferentially opting for GLP1s as an add-on to long-acting insulins. But this is in total coherence with the respective strengths of the two companies, i.e. best-in-class GLP1 on one hand, market leading long-acting insulin on the other. Each one wants to protect its interests. But, precisely, our goal is to try understand what is behind the marketing strategies and what is the true underlying target market for GLP1 analogs five or ten years from now. When a market leader like Sanofi with Lantus tries to expand in the field where it is dominant as it does with Lyxumia, there are significantly increased chances of success. But rules in pharmaceutical space have dramatically changed in recent years and marketing no longer dictates what should happen. So, although we understand what Sanofi is trying to do starting from the strong base of patients that Lantus offers, our increasingly strong belief is that vo-rdisk has a superior product and that in the end the best product will prevail. In other words, we are less and less convinced that the market segment will split between long-acting and short-acting GLP1s: the biggest slice of the GLP1 pie will be the one after OAD failures whereas Sanofi will simply do its best to leverage on the Lantus big opportunity to take the biggest possible portion with Lyxumia as add-on therapy. But the GLP1 market will also face other challenges in the years to come, first being the exponential increase in the number of players, not only with lixisenatide but also with albiglutide (GSK), dulaglutide (B.I./Lilly) and semaglutide (vo-rdisk) while AstraZeneca/BMS should be inclined to put more resources behind the new formulation of Bydureon that is expected to be approved sometime in H And there is something common between all the above-mentioned drugs with the exception of lixisenatide: a once-weekly administration. This is not unrelated to the previous debate: if new GLP1s reaching the market massively are once-weekly ones, then the concept of a short-acting GLP1 would see its validity even further jeopardised and its market opportunity further reduced. But beyond this debate, what is at stake is to try to understand how easy-to-use and friendly formulations of weekly GLP1s could impact the class. Some might argue that Bydureon has already been available for years and is a benchmark but the product has just carried too many disadvantages to be used as a proxy: need for reconstitution, bigger needle size, head-to-head study against Victoza (DURATION 6) showing inferiority, change in control from Amylin to BMS/AstraZeneca. This is too much to be competitive. w better drugs are approaching. An improved version of Bydureon could be the first answer to the critics. However it will address formulation issues but will not change the product per se. The first possible significant and impactful competitor in a once-weekly form to reach the market is likely to be Lilly s dulaglutide. The drug was filed in the US in October 2013 and in Europe in vember One of the leaders in the diabetes field could then enter the GLP1 market in late 2014 and directly with a once-weekly formulation that it will promote as the best choice. This is a threat to Victoza. But vo-rdisk s answer is not far behind. The company had considered two options simultaneously: a long-acting version of liraglutide or a new molecular entity called semaglutide. In the end it opted for the second one. By selecting semaglutide, vo-rdisk did not only want to have the best possible long-acting product but also to have a chance to develop a better product based on key efficacy metrics. While it is unlikely to compete on PPG or FG considering its pharmacokinetic 11

12 profile, the ideal once-weekly GLP1 will preferentially target HbA1c and weight, trying to provide some kind of basal GLP1 across a full week just like long-acting insulin provides basal insulin across a full day. What phase II data have shown is that semaglutide has the potential to be superior to liraglutide in managing glucose levels (as measured by HbA1c reductions) and perhaps even more significantly to be superior on weight loss (see Fig. 6). Fig. 6: Encouraging clinical data for semaglutide in phase II Source: Company Data, Investor Day December 2013 The first studies of the planned SUSTAIN phase III clinical programme which is made up of six trials started in H2 2013, namely SUSTAIN 2, 3 and 6. For that reason, these three are now listed on whereas the other three should start sometime in 2014 with reporting timelines for all in 2015 with the exception of the outcomes trial which is due early in Fig. 7: The SUSTAIN phase III clinical programme Trial # of pts Duration Estimated completion date Design Primary endpoint SUSTAIN weeks - monotherapy SUSTAIN weeks August 2015 add-on to MET, vs Januvia change in HbA1c SUSTAIN weeks July 2015 add-on to OADs, vs Bydureon change in HbA1c SUSTAIN weeks - vs Lantus SUSTAIN weeks - add-on basal insulin SUSTAIN weeks January 2016 vs placebo time to first MACE Source: Company Data; clinicaltrials.gov, Bryan, Garnier & Co ests. We think that a filing of semaglutide is possible in 2015 with outcomes data from SUSTAIN 6 added to the package sometime later but in due time for discussions with healthcare authorities in We have a first minor contribution to sales in 2016 and USD1.8bn in 2020, with a 50% probability of success across the whole period. If anything, this could prove conservative. But it is also fair to consider semaglutide in the whole context of a GLP1 family which could be composed of Victoza, IdegLira and semaglutide from 2016 onwards. All in all, our 2020 GLP1 franchise s revenue estimate at vo-rdisk is DKK21,420m or roughly USD4bn using current currency rates, o/w Victoza is 52%. 12

13 2.3. Other assets are almost all free-options More opportunities within diabetes During its Investor Day, vo-rdisk also spent some time on other assets within the diabetes franchise which are at earlier stages of development for the majority of them but with one notable exception: the faster-acting insulin aspart which recently started three of the four planned phase III trials, forcing us to factor it into our model for the first time. The objective is simple and obvious: try to be even closer to what the pancreas is doing in healthy patients around mealtimes, i.e. act very fast to control glucose levels while avoiding hypoglycaemic events, compared to existing short-acting insulins. The idea would be to inject the drug as the meal starts rather than to inject it ahead of the meal which still requires planning, although modern insulins are already very effective compared to older products. By making small changes to the insulin aspart molecule, introducing two new excipients which are inactive but have interesting properties (nicotinamide is boosting speed of action while L-arginine is improving stability), vo-rdisk thinks it has a new proposition that could offer as significant a difference compared to vorapid as the one previously offered by vorapid vs NPH. Based on the strong and clear data collected from phase I trials, the group has decided to move directly into phase III with a so-called ONSET programme made of four trials whose read-out will take place in 2015 with a tentative filing timeframe in H We have included the drug in our revenue model for the first time, with caution, i.e. a 50% probability of success, a non-adjusted 2020 sales forecast of USD600m while also adjusting vorapid on the downside over the same period of time. Beyond this opportunity, none of the other assets are factored in as we write as they are too early in their development. However, it is worth mentioning that an oral form of semaglutide entered phase II with the objective to be ready for phase III in the next couple of years (with several back-ups in phase I) and that an ultra-long acting insulin called LAI287 together with several oral insulin candidates are in phase I. There is no value for any of them but they are now on the radar screens Beyond diabetes, this is a free call t a single word was said about haemophilia or inflammation during the Investor Day. This could either mean that there was enough to do and say about diabetes and obesity and/or that vo- rdisk is comfortable with the idea that the rest is simply optionality and would not like the consensus to put any value on it. Whatever the reason, we agree: too early-stage on average, too little history in the respective fields for vo-rdisk, too limited an upside anyway. We issued a report and made a call a couple of years ago about the perspective of the haemophilia franchise for the group but it is fair to say that it has not (yet) delivered on promises as both vothirteen and voeight faced setbacks and delays while vatreptacog alpha was dropped. Nevertheless, we still have combined sales for all new haemophilia products of USD m in 2020 which does neither seem too optimistic nor out of reach. We have zero for the portfolio in inflammation and we do consider that this is globally the case for the consensus too, which is fine. 13

14 3. New FV, new rating 3.1. A new FV of DKK1, Impact from change in hypothesis for FV methodology Twice a year we revisit the basic general hypothesis used by the analysts to calculate their FV, i.e. the risk-free rate (RFR) and the equity risk premium (ERP). Deliberately we at BG have opted for a longterm approach to the topic as RFR is the 7-year average of five different European national rates whereas ERP is the 3-year average of three different risk premiums of European indices. Nevertheless, entering into 2014 has a significant impact on these calculations as current low rates are replacing much higher levels. RFR is moving down from 3.0% to 2.8% and ERP from 6.1% to 6.0%. These changes have a positive impact on vo-rdisk s FV of DKK60/share. The new WACC for vo-rdisk will be 7.60% Impact from roll over The turn of the year is also the context to switch year 1 into the DCF and EVA models from 2013 into For growth stocks like vo-rdisk, this has a significant impact as the full sequence of fast-growing cash flows is brought forward by one year at a rate of 7.6%. This change has a positive impact on vo-rdisk s FV of DKK35/share Impact from new products included for the first time This note is opportunistically used to introduce two new drugs into our model for the purpose of revenue and profit estimates as well as for the FV calculation: - Lira 3mg in obesity is first. t only does it progress on schedule and should soon be filed but it was discussed in a fairly convincing manner during the recent Investor Day, thus prompting us to introduce it for the first time, with a still significant discount (30% PoS only) however. In the corresponding part of this note, we presented the various metrics (population, penetration, price ) that we used to try evaluate its potential. In the end, this change has a positive impact on vo-rdisk s FV of DKK55/share. - The faster-acting insulin aspart is second. This has quickly moved into phase III and as such deserves value, here again with a significant discount (50% PoS). Although it is easier to monitor and to assess compared to a new compound in obesity, it will have to demonstrate superiority over existing vorapid which is far from being granted, hence the discount. This change has a positive impact on vo-rdisk s FV of DKK20/share. Considering the various elements mentioned, as well as minor changes to the underlying business and to currency rates estimates, the new FV stands at DKK1,180/share, up 18% compared to the previous DKK1,

15 3.2. Short-term catalysts may offer a good entry point A 1:5 share split is to take place on 2 January 2014 vo-rdisk is used to splitting its stock and has done so several times over the last 20 years: - On 2 January 1997, B shares were split 1:2; - On 4 April 2001, B shares were split 1:5; - On 3 December 2007, B shares were split 1:2; - On 2 January 2014, B shares will be split 1:5. As can be seen in the table below, the stock outperformed the Stoxx50 over 1, 3, 6 and 12 months after each of the three splits in 15 out of 16 cases, and in 50% of the cases by more than 10%: Date Type 1 month 3 months 6 months 12 months 02/01/1997 Div 1:2 Stock change Stoxx /04/2001 Div 1:5 Stock change Stoxx /12/2007 Div 1:2 Stock change Source: Bryan, Garnier & Co ests. Stoxx What else in January? On 30 January 2014 vo-rdisk will release its 2013 full-year results. The question could be: is the group able as early in the year as end January to change the full-year 2014 guidance? The answer to the question is probably no because moving parts and swing factors for 2014 are somewhat back-loaded in the year but a clearer view should be obtained about the ramp-up of Tresiba, revival of Victoza and potentially the new price increase of long-acting insulins in the US. Last year, vo-rdisk first said that revenues would grow high single-digits and in connection with the full-year results said 8-11% whereas operating profit went from growth to 10% growth. We cannot rule out a similar situation in 2013/2014. After Q3, the group said that both revenues and operating profits would grow in high single-digits. Should a range including double-digits be provided, this could be a first positive signal for an upside to numbers. By then, or at the same time, vo-rdisk should confirm that it has filed Lira 3mg in obesity with the FDA. This would not come as a surprise but would bring the candidate another step forward. For all these reasons (and some we have not developed here like the amazing and so convincing footprint the group has in emerging markets, where much of its growth prospects stand, making possible the challenging objective of bringing the total number of patients treated by vo products from 23m to 40m in 2020), we are confident in adopting a new BUY rating. It could have been a medium-term move but the share split, the Lira 3mg filing and full-year results suggest they could act as positive catalysts for the stock, hence speeding up our conversion. 15

16 DKKm 2012 Chg. 2013e Chg 2014e Chg 2015e Chg 2016e Chg 2017e Chg 2018e Chg Net turnover total % % % % % % % Insulin analogs % % vorapid/volog % rth America % % % % % % % Europe % % % % % % % International % % % % % % % Japan & Korea % 923-1% 831-2% 806-3% 766-5% 705-8% % Region China % % % % % % % vomix % rth America % % % % % % % Europe % % % % % % % International % % % % % % % Japan & Korea % 767-6% 694-8% % % % % Region China % % % % % % % Levemir % rth America % % % % % % % Europe % % % % % % % International % % % % % % % Japan & Korea 386-1% 285-7% 238-9% % % % % Region China % % % % % % % Tresiba rth America Europe % % % International % % % Japan & Korea % % % Region China % % % Ryzodeg Ultra fast-acting Human insulin & related % GLP1 family Victoza % rth America % % % % % % % Europe % % % % % % % International % % % % % % % Japan & Korea % 332-8% 290-5% 290 0% 275-5% 253-8% 241-5% Region China 70 NS % % % % % 417 5% IdegLira semaglutide Lira obesity Protein-related % % % % % % % Oral antidiabetic products % % % % % % % Diabetes Care total % % % % % % % Source: Company data; Bryan, Garnier & Co ests. 16

17 Price Chart and Rating History /12/ NOVO NORDISK 'B' Source: Thomson Reuters Datastream Ratings Date Ratings Price 30/03/2012 NEUTRAL DKK773 18/07/2011 BUY DKK653 Target Price Date Target price 01/11/2013 DKK /09/2013 DKK /08/2013 DKK /05/2013 DKK980 02/05/2013 DKK925 11/02/2013 DKK920 17/01/2013 DKK /10/2012 DKK920 01/10/2012 DKK885 10/08/2012 DKK910 08/06/2012 DKK760 30/03/2012 DKK770 03/02/2012 DKK725 30/01/2012 DKK705 28/10/2011 DKK660 13/09/2011 DKK655 05/08/2011 DKK715 18/07/2011 DKK720 17

18 18

19 BUY NEUTRAL SELL Bryan Garnier stock rating system For the purposes of this Report, the Bryan Garnier stock rating system is defined as follows: Stock rating Positive opinion for a stock where we expect a favourable performance in absolute terms over a period of 6 months from the publication of a recommendation. This opinion is based not only on the FV (the potential upside based on valuation), but also takes into account a number of elements including a SWOT analysis, positive momentum, technical aspects and the sector backdrop. Every subsequent published update on the stock will feature an introduction outlining the key reasons behind the opinion. Opinion recommending not to trade in a stock short-term, neither as a BUYER or a SELLER, due to a specific set of factors. This view is intended to be temporary. It may reflect different situations, but in particular those where a fair value shows no significant potential or where an upcoming binary event constitutes a high-risk that is difficult to quantify. Every subsequent published update on the stock will feature an introduction outlining the key reasons behind the opinion. Negative opinion for a stock where we expect an unfavourable performance in absolute terms over a period of 6 months from the publication of a recommendation. This opinion is based not only on the FV (the potential downside based on valuation), but also takes into account a number of elements including a SWOT analysis, positive momentum, technical aspects and the sector backdrop. Every subsequent published update on the stock will feature an introduction outlining the key reasons behind the opinion. Distribution of stock ratings BUY ratings 55.8% NEUTRAL ratings 31% SELL ratings 13.3% 1 Bryan Garnier shareholding in Issuer 2 Issuer shareholding in Bryan Garnier Research Disclosure Legend Bryan Garnier & Co Limited or another company in its group (together, the Bryan Garnier Group ) has a shareholding that, individually or combined, exceeds 5% of the paid up and issued share capital of a company that is the subject of this Report (the Issuer ). The Issuer has a shareholding that exceeds 5% of the paid up and issued share capital of one or more members of the Bryan Garnier Group. 3 Financial interest A member of the Bryan Garnier Group holds one or more financial interests in relation to the Issuer which are significant in relation to this report 4 Market maker or liquidity provider A member of the Bryan Garnier Group is a market maker or liquidity provider in the securities of the Issuer or in any related derivatives. 5 Lead/co-lead manager In the past twelve months, a member of the Bryan Garnier Group has been lead manager or co-lead manager of one or more publicly disclosed offers of securities of the Issuer or in any related derivatives. 6 Investment banking agreement A member of the Bryan Garnier Group is or has in the past twelve months been party to an agreement with the Issuer relating to the provision of investment banking services, or has in that period received payment or been promised payment in respect of such services. 7 Research agreement A member of the Bryan Garnier Group is party to an agreement with the Issuer relating to the production of this Report. 8 Analyst receipt or purchase of shares in Issuer The investment analyst or another person involved in the preparation of this Report has received or purchased shares of the Issuer prior to a public offering of those shares. 9 Remuneration of analyst The remuneration of the investment analyst or other persons involved in the preparation of this Report is tied to investment banking transactions performed by the Bryan Garnier Group. 10 Corporate finance client In the past twelve months a member of the Bryan Garnier Group has been remunerated for providing corporate finance services to the issuer or may expect to receive or intend to seek remuneration for corporate finance services from the Issuer in the next six months. 11 Analyst has short position The investment analyst or another person involved in the preparation of this Report has a short position in the securities or derivatives of the Issuer. 12 Analyst has long position The investment analyst or another person involved in the preparation of this Report has a long position in the securities or derivatives of the Issuer. 13 Bryan Garnier executive is an officer A partner, director, officer, employee or agent of the Bryan Garnier Group, or a member of such person s household, is a partner, director, officer or an employee of, or adviser to, the Issuer or one of its parents or subsidiaries. The name of such person or persons is disclosed above. 14 Analyst disclosure The analyst hereby certifies that neither the views expressed in the research, nor the timing of the publication of the research has been influenced by any knowledge of clients positions and that the views expressed in the report accurately reflect his/her personal views about the investment and issuer to which the report relates and that no part of his/her remuneration was, is or will be, directly or indirectly, related to the specific recommendations or views expressed in the report. 15 Other disclosures Other specific disclosures: Report sent to Issuer to verify factual accuracy (with the recommendation/rating, price target/spread and summary of conclusions removed). A copy of the Bryan Garnier & Co Limited conflicts policy in relation to the production of research is available at Yes 19