CFA - The Pharmaceutical Industry

CFA INSTITUTE INDUSTRY GUIDES

THE PHARMACEUTICAL INDUSTRY ISBN 978-0-938367-81-9

9 780938 367819

90000

CFA INSTITUTE INDUSTRY GUIDES THE PHARMACEUTICAL INDUSTRY by Marietta Miemietz, CFA

©2013 CFA Institute This publication is designed to provide accurate and authoritative information with regard to the subject matter covered as of the date of publication. It is distributed with the understanding that the publisher is not engaged in rendering legal, accounting, or other professional services. If legal advice or other expert assistance is required, the services of a competent professional should be sought.

978-0-938367-81-9 October 2013

ABOUT THE AUTHOR Marietta Miemietz, CFA, is a cofounder of and director of pharmaceutical advisory services at Primavenue. Previously, she spent 13 years as a sell-side analyst of the European pharmaceutical and health care industries and was consistently rated among the top 10 pharmaceutical analysts from 2006. Ms. Miemietz holds an MBA with a concentration in finance from WHU–Otto Beisheim School of Management and the Belgian Diplôme d’Etudes Spécialisées in biotechnology from Université Libre de Bruxelles.

©2013 CFA INSTITUTE

iii

CONTENTS Introduction 1 Industry Overview

2

The Drug Discovery, Development, and Approval Process

9

Intellectual Property: Patents, Regulatory Exclusivities, and Other Forms of Protection 18 Business Models Industry Consolidation Notable Trends

22 23 25

Financial Statement Analysis Forecasting Drug Sales and Company Profits

27 31

Valuation of Pharmaceutical Firms Portfolio Considerations

39 42

Industry Resources Regulatory Agencies Other Resources Major Medical Conferences

46 46 46 48


v

INTRODUCTION Among the most distinctive features of the pharmaceutical industry are the complexity and length of the product development cycle and the independence of pharmaceutical companies’ operating performance from industry trends. The earnings outlook of individual companies is determined, first and foremost, by the products they develop and market. Consequently, the industry is characterized by exceptional heterogeneity that notably complicates peer group analyses and often entails divergent share price performance. These unusual characteristics are attributable to the fact that the prospects of each player are linked to the prospects of the drugs to which it has full or partial commercialization rights. Drugs are approved by regulatory agencies for specific indications, and their peak sales potential depends on the prevalence of the conditions they are intended to treat, their efficacy and safety, and the competitive landscape. In this context, it is worth noting that the conditions a medicine is intended to treat are often narrowly defined; for example, a drug that is licensed for the treatment of colorectal cancer is unlikely to compete with a blood cancer drug, and even two different blood cancer drugs might be targeted at separate patient populations. These dynamics have profound implications for pharmaceutical industry analysis and investing. Top-down analytical approaches based on overall market growth rates and market share development, often a good starting point in other industries, add limited value at best and may often be misleading. The quality of bottom-up analyses that take into account the clinical utility, cost-effectiveness, competitive landscape, intellectual property, and economics of individual drugs is typically the main success factor in selecting pharmaceutical and biotechnology stocks. Following a brief overview of the industry, this primer delineates the determinants of success in drug development and marketing and then reviews the implications for financial statement analysis and forecasting, as well as valuation and portfolio considerations. Unless noted otherwise, analysis is confined to branded drugs for human use, and conclusions may not apply to other areas of health care, such as consumer and animal health care products or generics. The aim of this report is to provide a general understanding of the complexity of pharmaceutical industry analysis and the main issues involved. It is designed to enable the reader to critically appraise research, corporate presentations, and other communications with respect to drugs, companies, and the industry. It is by no means exhaustive. Myriad issues may arise— issues that are deeply ingrained in the scientific aspects of a molecule, the clinical considerations pertaining to a particular disease, regional clinical practice and regulatory legislation, or specific patents—that must be reviewed on a case-by-case basis. ©2013 CFA INSTITUTE

1

INDUSTRY OVERVIEW Given that introduction, it should come as no surprise that the pharmaceutical market is large and highly fragmented. In 2012, the global market for human prescription pharmaceuticals was valued at more than $850 billion. The four largest market categories—the central nervous system, cancer, metabolic and gastrointestinal diseases, and cardiovascular disorders—accounted for slightly more than half the market in terms of value; each of these categories can be subdivided into numerous conditions that require separate treatment approaches. As depicted in Figure 1, the largest players hold mere single-digit market shares, and many of them are active in other segments of the health care market (note that revenues from activities other than health care are not shown). Successful drug development today requires a unique skill set that cannot be transferred to other industrial activities. Nonetheless, many of the leading pharmaceutical players are exposed to other areas of health care for historical reasons and with a view to smoothing out the growth profile and cash generation on the group level as well as exploiting the modest synergies with regard to, for example, target markets, research and development (R&D),

80 70 60 50 40 30 20 10 0

7% 6% 5% 4% 3% 2% 1% 0%

Pfi

ze Ro r c N he M ova er r ck s & Gl Co ax oS San m ofi Jo As ithK hn tr lin so aZ e n en & e Jo ca Br h ist ns ol on -M ye rs Lilly Sq ui b Ab b bV Am i e ge No B n vo ay No er rd i Ta sk ke da

2012 sales in $bn

Figure 1. Branded Human Prescription Drugs: Key Players

Pharmaceucals (human branded prescripon drugs excluding vaccines) Other healthcare Pharmaceucal market share (RHS)

2

WWW.CFAINSTITUTE.ORG

Industry Overview

or production. These ancillary health care activities primarily include consumer health, animal health, generics, diagnostics, and medical technology. As noted previously, the dynamics of the market for patented prescription drugs for human use are such that drug-specific attributes are a far more important determinant of success for individual companies than are general industry trends. In most of the key pharmaceutical markets of the developed world, the majority of patients are able to obtain the drugs they need; for the most part, their treatment is paid directly or reimbursed by third parties, although restrictions often apply (e.g., mandatory generic substitution or requirements to initiate therapy with the lowest-cost medicines). Consequently, the pharmaceutical industry is among the least cyclical of industries, but recessions may entail such austerity measures as drug price cuts, and dwindling consumer confidence may, to some extent, result in fewer physician office visits by patients or “drug holidays” (discussed later). Most innovative drugs enjoy a period of market exclusivity—because of either patents or regulatory exclusivities—that may span many years, implying that a new molecular entity will have monopoly status for a certain period. Thus, a pharmaceutical company that is commercializing a highly effective, patent-protected drug in a therapeutic area of high unmet need may be able to generate strong sales growth in the same year that a competitor faces a rapidly declining top line—for example, owing to patent loss and ensuing generic competition or as a result of emerging branded competition or safety concerns about its main products. Under the pharmaceutical industry’s cost structure, positive revenue developments translate into significant operating leverage. Companies incur substantial R&D and marketing expenses that are largely fixed in the short term; the variable cost of producing and distributing higher volumes of any given drug is comparatively low. Consequently, the accuracy of forecasts of a pharmaceutical company’s profits hinges on the analyst’s ability to predict the future sales of each drug in the company’s portfolio and pipeline. Substantial errors in forecasting a company’s top line will almost invariably lead to even greater errors in forecasting the bottom line and, thus, a “bad call.” Therefore, a thorough analysis of a company’s drug portfolio, which frequently requires expert knowledge in various therapeutic areas, is of paramount importance. In addition, various megatrends and industry-specific themes affect the dynamics of the industry to a meaningful extent. In light of the pronounced changes in pharmaceutical business models that have been implemented over the last five years, a brief history of the pharmaceutical industry is in order before reviewing its dynamics. Although the discovery of the first drugs was largely the result of serendipity, increasing levels of insight into disease biology and the mechanism of the action of drugs on the molecular level resulted in ever more targeted drug discovery efforts, which bore fruit initially. The latter part of the 20th century saw ©2013 CFA INSTITUTE

3

CFA Institute Industry Guides: The Pharmaceutical Industry

step changes in medical innovation; given the dearth of effective drugs available at the time to treat such widespread conditions as diabetes and hypertension, many newly launched drugs became blockbusters, attaining peak sales of $1 billion or more. The pharmaceutical industry enjoyed high earnings growth, and the investment community’s expectation that the industry would continue to innovate at the same pace was reflected in the industry’s valuation: P/E multiples often reached the high teens or greater. Many companies put in place significant production, marketing, and administrative infrastructure in an effort to maximize the top line; as recently as 10 years ago, some of the most successful primary care drugs were each promoted by thousands of sales representatives in the United States alone. Most major pharmaceutical firms dedicated substantial resources to such life-cycle management (LCM) activities as the development of new formulations of existing drugs or clinical trials in additional indications or patient subgroups, all with a view to extending the lives of the drugs’ patents. Although pharmaceutical companies generally do not disclose the proportion of their R&D expenses attributable to LCM as opposed to the discovery and development of new molecular entities, evidence suggests that LCM activities proved to be highly lucrative. The decline in new-drug approvals observed through much of the last decade, despite rising absolute R&D spending, may be attributable in large part to the focus on product LCM. Other possible contributory factors include rising hurdles for some of the largest indications, such as diabetes and hypertension, in which improving on existing drugs has become increasingly challenging, as well as a delay in the adaptation of R&D and business models to a changing regulatory and payer environment. At the beginning of the new millennium, the industry placed much emphasis on the development of primary care drugs with billion-dollar sales potential in order to leverage their existing infrastructure and replace older drugs that were approaching patent expiration. To minimize the perceived risks of the costly clinical development phase of new drugs—whereby pharmaceutical firms test the efficacy and safety of new-drug candidates in hundreds, if not thousands, of patients over many years—many firms developed new molecular entities that displayed only modest structural variation and only minor therapeutic advantages over existing drugs. In referring to these products, critics used the derogatory term “me-too” drugs. Shifts in the regulatory and payer environment eventually derailed the industry. In particular, the US Food and Drug Administration (FDA) displayed heightened risk aversion in the wake of the withdrawal of Merck’s painkiller Vioxx from the market in 2004 owing to side effects, thus raising the bar for the approval of newdrug candidates targeted at non-life-threatening conditions. Payers grew increasingly reluctant to reimburse for expensive new drugs that offered only a modest perceived benefit over older drugs, which were losing their patent protection and 4


Industry Overview

becoming available generically at much lower cost, a trend that was exacerbated in the financial crisis of 2008. Collectively, all these trends resulted in high attrition rates for new-drug candidates as well as some commercial failures. Toward the end of the last decade, many large pharmaceutical stocks were trading on single-digit forward P/E multiples as “patent cliffs” loomed and investors’ confidence that the industry’s R&D engines would yield novel agents to offset the imminent revenue loss plummeted. The realization that the past strategy might no longer be viable, coupled with the market’s disenchantment, triggered the industry’s quest for a new commercial model. Large pharmaceutical corporations recruited managers who were industry novices but possessed extensive experience in such fields as marketing and operational excellence. The major players embarked on large-scale cost-reducing initiatives to ensure acceptable levels of profitability beyond the patent expirations of key blockbusters; many companies reduced their cost base by billions of dollars in a matter of years. Although some of these cost reductions were attributable to synergies in the context of “megamergers,” various companies achieved multi-billion-dollar savings in the absence of material M&A activity. This outcome was achieved in part by scaling back primary care field forces in Western markets, a step that was accompanied by changes in the commercial model: Firms relied increasingly on key account management to drive the top line and shifted their R&D efforts toward specialty care. The relative attractiveness of specialty care over primary care lies in the fact that it can be served by a smaller sales force and thus at a smaller fixed cost; at the same time, clinical and regulatory success rates tend to be higher for drug candidates that target underserved niches of debilitating and potentially life-threatening indications, such as cancer. Previously neglected aspects of cost control were also addressed, including excessive procurement bills that resulted largely from a lack of coordination of group-wide purchasing activities. Furthermore, the industry reduced fixed costs, especially those related to R&D, by outsourcing various activities. It proactively identified incremental business opportunities; many companies rediscovered ancillary activities (consumer and animal health) that are less prone to patent loss and notably increased their presence in emerging markets. Although the drug industry is global in nature—regulatory approvals all over the world can usually be obtained for drugs that have proved safe and effective in the treatment of the targeted conditions—many pharmaceutical firms have focused primarily on Western markets in the past. And although Western markets continue to dominate in absolute terms—with the United States estimated to account for approximately 40% of the global pharmaceutical market in 2012—much of the majors’ growth is now coming from the emerging markets. Many pharmaceutical companies have made steady progress toward replenishing their pipelines, partly by tapping into external innovation provided, to a large ©2013 CFA INSTITUTE

5


extent, by midsize and biotechnology companies. In the aggregate, pipelines comprise numerous molecules with novel modes of action that target medical areas of high unmet need and are tailored to well-defined patient populations, thus implying that a genuine market opportunity is likely to materialize for compounds that prove safe and effective. In fact, many recently launched drugs and compounds in development are so highly targeted that a debate has ensued whether “personalized medicine,” one of the most extensively discussed megatrends of the health care sector, represents an opportunity or a risk for the pharmaceutical industry. Proponents argue that the more clearly a drug’s target population is defined and the more easily it is identified—for example, by the use of biomarkers that confirm the presence or absence of a mutation—the higher the chance that such a drug can be developed. This approach may, in turn, increase the chance of successful clinical trials and save the firm the considerable expense of conducting negative trials. Although skeptics are concerned that personalized medicine might shrink a drug’s target market to relatively small patient subgroups, proponents contend that highly targeted drugs may gain traction rapidly in the treatment of conditions characterized by a significant genetic component, such as tumors that bear certain mutations, whereas patients suffering from such multifactorial illnesses as diabetes may be more difficult to stratify. Thus, personalized medicine should not be regarded as a threat per se to profitability; drugs that are more narrowly targeted and thus more effective have a higher chance of obtaining regulatory approval as well as reimbursement at attractive price points, which may compensate for the potential volume tradeoff. Furthermore, payers are increasingly embracing value- and outcome-based pricing models. For instance, health technology assessment bodies assign grades to a drug’s benefit in certain indications and patient subgroups, and some pharmaceutical companies have agreed in some regions to refund the cost of drug treatment if the patient does not benefit meaningfully, suggesting that they have nothing to gain from the administration of medicines to patients who are unlikely to respond. In this context, it is worth noting that disease prevalence is only one of the factors that determine a drug’s peak sales. A drug’s efficacy, both in absolute terms and relative to competing drugs, and the severity of the condition it is used to treat have a significant impact on its price point as well as its penetration rate. Similarly, the notion that the aging population in Western markets will be a key driver for the pharmaceutical industry is a fallacy. A drug designed to treat conditions that arise more frequently in old age, such as diabetes, will not necessarily achieve greater commercial success than a drug aimed at severe illnesses that may manifest themselves earlier in life (e.g., multiple sclerosis). Some senior citizens may be well controlled on generic drugs or considered too frail for intensive treatment, whereas 6


Industry Overview

a young patient suffering from a devastating disease may receive an expensive drug for extended periods in the absence of generics. Effective treatments for some of the most debilitating conditions frequently associated with old age, such as Alzheimer’s disease, are still largely lacking. Price pressure is another trend that has persisted for many years and has accelerated recently as government-linked payers have felt the brunt of austerity while private payers have been affected indirectly by austerity and recessionary tendencies. Although one or more price increases a year remain the norm for many effective drugs marketed in the United States, price cuts and concessions occur with some regularity in many other regions. Pharmacoeconomic studies that demonstrate an overall cost benefit to the health care system are gaining in importance. Although they raise the cost of drug development and may be fraught with methodological challenges, they also allow for the price differentiation of highly effective new drugs. Recent and upcoming expirations of blockbusters’ patents are expected to result in tens of billions of dollars in savings for health care systems globally, thus providing some flexibility to reward innovation in areas of high unmet need. Austerity notwithstanding, it thus appears fair to assume that a safe and highly effective drug for the treatment of a serious, underserved condition, such as Alzheimer’s disease or heart failure, would likely achieve peak sales well in excess of $1 billion. Where clinical differentiation is lacking, price pressure is likely to intensify further. For example, some generics markets, including Germany’s, are increasingly tender driven, with significant negative effects on price and profitability. In the United States, where substitution of generics is common, generics companies rely heavily on first-to-market strategies that afford short windows of opportunity to maintain relatively high prices until the onset of multi-source generics. The rising price premium of truly innovative drugs that address medical areas of high unmet need over interchangeable products and those conferring only a modest benefit has resulted in increasingly focused strategies, with the most innovative pharmaceutical companies pursuing differentiation rather than cost strategies in Western markets. Today, only a few of the pharmaceutical majors have significant generics operations in developed markets because the key determinants of success—including time to market, breadth of portfolio, and logistical capabilities—differ markedly from the core competencies required in the branded pharmaceutical space. With respect to general industry dynamics, the relative lack of seasonality and cyclicality is worth noting. With the exception of some categories (e.g., allergy treatments and flu vaccines), most drugs are prescribed and administered yearround. The months of summer vacation as well as the holiday season are typically somewhat lighter than other times of the year. To the extent that there is an element of patient self-pay, drug sales may be moderately cyclical. For example, patient co-pay requirements per drug pack may induce some patients to take “drug ©2013 CFA INSTITUTE

7


holidays” in a tough economic climate, while those patients who have to pay part of the fee for physician office visits out of pocket may postpone a health checkup and, by implication, the purchase of medicines for the treatment of conditions they are unaware of. Where employers are a key source of private health insurance, rising unemployment rates may negatively affect consumption. Self-medicating individuals may replace expensive over-the-counter (OTC) brands with white label products in times of declining consumer confidence. However, all these factors tend to have only a very modest impact on industry sales; rarely do they visibly affect any sets of quarterly results. Health care reform, usually in the form of price cuts, typically represents the main fallout from a bleak macroeconomic picture. In recent years, many companies have experienced low- to mid-single-digit annual price pressure across their European drug portfolios in the wake of austerity measures implemented by various governments. The US market is dominated by private health insurance and has thus been largely immune to government initiatives in most years. Beginning in 2010–2011, however, the US Affordable Care Act reduced companies’ US drug sales by a low single-digit percentage as a result of increases in the mandatory Medicaid rebate rates and similar measures. It remains to be seen whether improved access to health insurance will have a positive effect on industry growth rates in the longer term. Two key considerations in forecasting the impact of health care reform are worth highlighting. First, reform measures usually need to be ratified by legislatures and their implementation can be time consuming; it is possible that the first impact on the industry will not be felt until a year or more after the first observation of a decline in macroeconomic indicators. Second, the sensitivity of branded drug sales to reform measures rarely differs among the majors; although some drugs may be more negatively affected than others, the drug portfolios of the industry majors tend to be sufficiently diversified for the net effect on branded human drug sales to be similar across companies. Of course, exposure at the group level is partly determined by diversification into areas other than patented human prescription drugs. However, care must be taken when assessing the impact of any reform measures on small and midsize players, which may face substantial exposure with respect to a particular drug or region. In extreme cases, health care reform has the potential to negatively transform the earnings of such companies.

8


THE DRUG DISCOVERY, DEVELOPMENT, AND APPROVAL PROCESS Prior to launching a new drug, the sponsor (i.e., the pharmaceutical company or companies that own the rights to the compound) must extensively evaluate its efficacy and safety in order to obtain approval from the appropriate regulatory authority in each jurisdiction where it intends to market the drug. For example, US approval must be obtained from the FDA; European approval is usually obtained from the European Medicines Agency (EMA); and Japanese approval may be granted only by the Japanese Ministry of Health, Labor, and Welfare. The drug development process is lengthy, costly, and fraught with risks. Concerns over a compound’s efficacy, safety, or commercial viability may emerge at any point in the process; if such concerns are sufficiently serious, the company may decide to terminate development, which implies that the investment in the compound’s development will never be recouped. Pharmaceutical analysts regularly revise their forecasts in response to R&D-related news—notably revising sales forecasts for drug candidates to reflect their launch probability (which rises as drugs progress through development), the expected commercial positioning in light of emerging scientific data, and any potential changes to launch timelines. The progression of drug candidates to the costly advanced stages of development may also have a bearing on short- and medium-term R&D expenses. Since continuous rejuvenation of the drug portfolio is of paramount importance to a pharmaceutical company’s profitability, a thorough understanding of the drug discovery, development, and regulatory process is crucial to the accuracy of forecasts. This section explains the basic drug development process and the regulatory process in key geographic regions. Special emphasis is placed on the key US market, where the regulatory process is highly transparent and usually relatively speedy. Drug discovery generally starts with ideas for a drug target and a lead molecule. The choice of target (e.g., a cell surface receptor involved in sending messages into the cell nucleus or a messenger molecule that binds to receptors as a ligand) is typically driven by a company’s understanding of the biology of a particular disease. For example, a tumor might express cell surface receptors that are absent in healthy tissues, and a pharmaceutical company might endeavor to develop a medicine that selectively targets this receptor. Although many pharmaceutical majors work on elucidating disease mechanisms, substantial outside work—performed, for ©2013 CFA INSTITUTE

9


example, by academic institutions—is taken into account when choosing a target. Translational medicine is the branch of science concerned with the clinical applications of basic research. Once a target has been identified, the company’s scientists study various approaches to blocking or modulating the target to reduce or ablate disease activity. For example, they may use a compound that binds to and blocks the cell surface receptors found on cancer cells, thereby preventing them from receiving further growth signals. The compounds identified as having potential utility are classified as leads. Leads must satisfy various requirements—for example, they need to interact effectively with their target, but interaction with other molecules in the human body should be kept to a minimum in order to avoid side effects that may arise from off-target activity. Leads are often identified by screening molecules from existing “libraries” against the target. Optimizing the most suitable compounds identified in this manner typically requires substantial knowledge of chemistry or biochemistry. Owing to the complexity of the process, it is not feasible for any company to identify and study each target and lead compound in-house. Therefore, co-operations between firms are announced with some regularity. For example, one pharmaceutical or biotechnology company may supply the library to be tested against a target supplied by another company, or two companies that have identified drug candidates targeting the same pathway may join forces to develop these compounds together. Collaborations between the industry and academia are also common. Furthermore, pharmaceutical companies frequently in-license drug candidates that originated at biotechnology companies; the larger company thus bolsters its pipeline while providing financing as well as clinical and regulatory expertise and marketing prowess. Once a lead compound has been identified, it proceeds to the pre-clinical stage, which comprises various tests in vitro as well as in relevant animal models. Testing in humans, also known as the clinical stage, requires approval from ethics committees, which is granted only after a compound has been fully characterized in the pre-clinical setting. Clinical trials are designed to fully elucidate a drug’s safety, efficacy, and key characteristics in humans. Thus, the design of clinical studies and the scope of the program vary with the type of compound being studied and the condition it is intended to treat. Simplistically, the clinical trial process may be divided into three phases. In phase I, the compound is typically tested in healthy volunteers. Phase II studies enroll patients and are usually designed to give preliminary evidence of efficacy and safety while determining the best dose(s) to test in phase III. Phase III trials are designed as pivotal, or registration, trials and are powered to yield statistically significant results on a drug’s efficacy and safety. Although regulators review the totality of the data (including the full safety database available at the time) before approving or rejecting a new-drug application, 10


The Drug Discovery, Development, andIndustry ApprovalOverview Process

successful completion of phase III is usually a prerequisite for approval. Once a drug has been approved, its safety is monitored regularly and the sponsor may be asked to fulfill post-approval commitments; this post-marketing phase is sometimes referred to as phase IV. Exhibit 1 outlines the basic process of getting a drug to market. Exhibit 1 presents a conceptual framework; in practice, the drug development process may vary substantially. For example, most cancer drugs are too toxic to be given to healthy volunteers. Drug candidates for the treatment of conditions with a poor prognosis may receive regulatory approval after a pivotal phase II study. An adaptive trial design (e.g., a phase II study is rolled over into a phase III trial if certain criteria are met) is increasingly being used for some diseases. Regulators frequently provide sponsors with guidance and feedback on trial design in an effort to minimize the risk of inadequate trial design. Clinical trials are often large, with hundreds of patients typically enrolled in phase II studies and thousands, occasionally tens of thousands, recruited into phase III. Hence, the cost of clinical development is high: Depending on the therapeutic area under study, phase II trials may cost tens of millions of dollars, and the bill for a phase III program often amounts to hundreds of millions of dollars. The increasing focus on “outcomes” trials (discussed later) puts further upward pressure on the cost of clinical development. Consequently, the industry has concluded that the cost of failure in phase III is unacceptably high and has put in place extensive measures designed to identify potential problems in phase II or earlier. Although the industry now appears to Exhibit 1. Getting a Drug to Market Drug discovery and pre-clinical phase

Phase I

Phase II

Phase III

Target and lead identification; testing in vitro and in vivo (animal model)

Tests in healthy volunteers

Efficacy and safety in patients; dose selection

Registration studies to establish safety and efficacy

Multi-year process

Cost: millions of dollars

Cost: millions of dollars


Duration: months

Duration: typically >1year Cost: at least tens of millions of dollars

Duration: about 2 years (varies)

Regulatory review Duration: typically 1year; expedited review maybe available

Launch

Within days or weeks of approval if reimbursement negotiations are not necessary; otherwise, up to a year or more

Cost: often hundreds of millions of dollars

11


appraise the clinical and commercial potential of drug candidates in early- and midstage development more critically than in the past, the possibility of failure in phase III can never be ruled out. It is therefore prudent to risk-adjust sales forecasts until a drug has passed phase III and, ideally, received regulatory approval. Owing to the high cost of clinical development, studies are typically sponsored by the pharmaceutical firms themselves. Occasionally, investigators or cooperative groups may sponsor trials based on their own hypotheses. Although these studies may occasionally produce intriguing results, caution is warranted because they are not always comparable in size and quality to industry-sponsored trials and are often unsuitable as registration trials. Because share prices tend to react to the results of pivotal trials, it is worthwhile to briefly review the design of typical phase III trials and touch on the interpretation of results. Many pivotal studies are designed as global trials, with clinical centers across the United States, the EU, Eastern Europe, Asia Pacific, Latin America, and other key regions. The FDA usually requires the inclusion of a meaningful number of US patients in the pivotal study, whereas other regulators may accept data from a smaller local study in addition to the pivotal data in order to ascertain that the drug is safe and efficacious in the local patient population. Generally speaking, the patient population enrolled in the clinical program must be representative of the patient population that will receive the drug after its approval because such factors as ethnicity and standard of care may have a bearing on patients’ responses to a drug. Often, a phase III program consists of two studies; however, regulators frequently accept a sole pivotal study, notably for indications that require large and complex trials. For other indications, it may be advisable to conduct more than two phase III studies in order to demonstrate the drug’s compatibility with other frequently used drugs in different patient populations—diabetes being a prime example. The study sponsor typically selects one primary endpoint, although co-primary endpoints are occasionally chosen in complex settings, such as acute care. The primary endpoint is usually an efficacy endpoint and reflects the main hypothesis that the trial has been designed to test. For example, the primary endpoint of a diabetes trial may be a reduction in blood sugar or a composite score of heart health. Trials that assess a drug’s impact on the mortality and morbidity (M&M) of the patient population are often referred to as “outcomes” trials. Endpoints related to M&M are considered “harder” than so-called surrogate endpoints, which merely measure changes in a marker of disease severity, such as blood glucose or blood pressure. However, M&M trials tend to be lengthy owing to the requirement to enroll a very large number of patients and to follow them for a long period in order to observe statistically significant differences in rare events, such as death. Therefore, they are not usually part of the initial registration package. In addition to the primary endpoint, the sponsor chooses secondary endpoints, which often include safety. 12



Simply put, a study is considered positive if the primary endpoint is met— that is, if the main hypothesis is proved and the result is statistically significant. Although a positive study bodes well for approval of the drug, the regulators evaluate the totality of the evidence and may reject a drug for other reasons, such as observed safety signals or weak results on secondary efficacy endpoints. If the primary endpoint is not met, the study is considered negative, making regulatory approval extremely unlikely. Choosing the primary endpoint well and optimizing other aspects of the trial design are thus of paramount importance to the success or failure of a drug. Secondary endpoints are generally considered merely supportive, and even resounding success with regard to each secondary endpoint usually does not make up for failure to meet the primary endpoint. Some drug candidates have been doomed by poor trial design rather than by an intrinsic lack of efficacy or safety. Other aspects of trial design that may have a bearing on a drug’s chances for approval include whether the design is for an open-label trial or a double-blind trial (in which patients and physicians do and do not know, respectively, whether they are receiving the study drug, a placebo, or another comparator); the choice of comparator (a placebo or an active comparator that is commonly used to treat the disease); the inclusion and exclusion criteria; the statistical analysis plan; and numerous other factors. Once the full clinical development of a drug candidate has been completed, the sponsor usually submits the entire dossier to the relevant regulatory authorities in the jurisdictions where it intends to market the product. Submission in the United States, the EU, and Japan is now virtually simultaneous for many drug candidates, although timing differences can arise from minor or major variations in regulatory requirements with respect to either the clinical development plan or the data analysis. The regulatory review process starts upon receipt of the dossier by the agency. From the public’s point of view, the FDA offers the most transparent process. Normally, the agency formally accepts an NDA (new-drug application) file for review shortly after its submission. On the rare occasions when a dossier is rejected— usually for technical reasons—the sponsor typically resubmits within a relatively short period. Under Prescription Drug User Fee Act (PDUFA) regulations, the FDA’s standard review time is 10 months, although the agency may extend the review period by up to 3 months if it requires more time to consider the vast amounts of data that generally need to be analyzed as part of an NDA review. Expedited review procedures may be available for drug candidates that target a medical area of very high unmet need, such as rare and lethal forms of cancer. Irrespective of the type of review process, the FDA sets an action date, also known as the drug candidate’s PDUFA date, by which the agency must communicate its regulatory decision. The FDA may either approve a drug or send a Complete Response Letter (CRL) stating that a drug application cannot be approved in its present form. On some occasions, ©2013 CFA INSTITUTE

13


the deficiencies raised in the CRL may be addressed fairly quickly and the drug may be resubmitted for approval within a relatively short space of time. In other cases, new clinical trials may be necessary to establish a compound’s efficacy and safety to the FDA’s satisfaction, which can delay the product launch by years. Before making a regulatory decision, the FDA may convene a panel of experts (also known as an advisory committee) who publicly share their views on the drug’s efficacy, safety, and overall approvability. The amount of drug-specific data and other information made publicly available in the context of advisory committee meetings typically far exceeds the amount of data that can be gleaned from any other source. Extensive briefing documents are posted on the FDA’s website, usually 48 hours before the start of the panel’s meeting. These documents contain the FDA’s questions to the panel (tough questions have, on occasion, rattled investors’ nerves), both the sponsor’s and the FDA’s detailed review of the data, and a preliminary assessment by the FDA reviewer—all spread over hundreds of pages. The meeting itself typically lasts a full day, with presentations by the sponsor and the FDA as well as questions by the panel to both the sponsor and the FDA. The meeting also includes an open public hearing—where other stakeholders, such as patients and patient organizations, may express their views on the suitability of the drug for the targeted patient population—and a debate by the panel members on nonvoting questions, followed by yes/no/abstain votes on the voting questions. Typical voting questions seek to ascertain whether the drug’s efficacy and safety have been established and whether the drug should be approved. The entire meeting is usually webcast and provides not only a glimpse of the FDA’s main concerns and the likelihood of approval but also a general sense of factors that may have a bearing on the drug’s commercial potential. Nonetheless, the outcome of an advisory committee meeting should be interpreted with caution. Importantly, the FDA retains ultimate responsibility for the approval of a drug; a positive “adcom” vote does not guarantee approval, nor does a negative vote necessarily herald rejection. Pharmaceutical companies usually issue a press release on the voting results shortly after the panel adjourns; however, the votes may not give a full picture of the panel’s views on a drug. It is therefore advisable to watch the panel itself and to note the explanations of panel members for their votes. Some yes votes may be heavily “caveated,” while some no votes may relate to concerns that are easily addressed. The panel members’ opinions may not reflect those of attending physicians in the field; the panel members represent different areas of expertise and may include statisticians and practitioners of other disciplines who would not necessarily prescribe the drug after approval. Conflicts of interest, such as extensive consulting agreements with the pharmaceutical industry, may keep some of the most renowned opinion leaders off the panel. Finally, the committee is merely expected to weigh in on the compound’s approvability in general terms 14



and does not consult directly on the label, although panel members periodically point out that they struggle to discuss the issue of approvability in a vacuum. For example, if a panel member believes that a drug should be withheld from patients with renal failure and that it should be approved if appropriate monitoring of renal function is mandated, that panel member would be expected to vote in favor of approval and to rely on the FDA to address contraindications and requirements for monitoring on the label. Therefore, a drug may receive approval, but a restrictive label may effectively relegate it to later lines of therapy and thus limit its peak sales potential. The approval process of the European Medicines Agency (EMA) differs from that of the FDA with respect to various administrative aspects and is often less transparent to the public. Many drugs are submitted to the FDA as part of the centralized authorization procedure, which results in a single marketing authorization that is valid throughout the European Union, Iceland, Liechtenstein, and Norway. Occasionally, national approval procedures may be chosen—either the decentralized procedure, whereby sponsors may file simultaneously in more than one country, or the mutual recognition procedure, whereby a drug is approved in one country with an option to subsequently request recognition of that authorization in other EU member states. Like the FDA, the EMA formally accepts or rejects the dossier. The actual review process takes up to 10 months; however, questions from the EMA to the sponsor trigger “clock-stops” until receipt of the answers. These interruptions are not formally communicated to the public, making the timing of the EU decision on the approval of a new drug difficult to predict. If the EMA’s queries cannot be addressed in the time frame specified, the dossier is typically withdrawn and later resubmitted. The EMA’s Committee for Medicinal Products for Human Use (CHMP) convenes monthly, usually after the 20th day of each month. Unlike FDA panel meetings, CHMP meetings are nonpublic to shield the committee from any lobbying efforts on the part of stakeholders. Following completion of the process, the EMA issues a European Public Assessment Report (EPAR), which summarizes its conclusions with respect to a compound’s risk–benefit profile. At the end of the review process, the CHMP issues a recommendation to the EU to approve or reject the drug. The EU generally follows this recommendation within three months of issue. Whether the regulators, especially the FDA, have become more exacting and possibly more politicized is a subject of intense debate. The FDA might be forgiven for being gun-shy, having taken flak from the US Congress in the wake of post-approval safety concerns that have led to product withdrawals. In contrast, the EMA appears to be relatively insulated from politics. The ethical dilemma faced by regulators is inherent in their mandate to make new drugs available to patients to halt or slow down disease progression and reduce sequelae while shielding them from druginduced harm. Even the largest clinical trials may not unearth all the side effects ©2013 CFA INSTITUTE

15


that may arise in the field, and imbalances in serious adverse events observed in clinical trials between patients receiving a study drug and those on a placebo or other comparator could either signal a potential safety issue or merely reflect the play of chance. Ruling out unacceptable safety risks is thus one of the main challenges of both drug development and regulatory review. Regulatory guidance documents that lay down the specific requirements to establish a drug candidate’s safety to the agencies’ satisfaction have greatly clarified the statistical aspects of trial design and interpretation. For example, in the wake of concerns over heart risks associated with anti-diabetic agents, the FDA established clear rules on demonstrating the absence of unacceptable cardiovascular risk. Another point of contention is the responsibility that regulators are expected to assume for protecting patients from themselves. Although the EMA appears largely to trust physicians to prescribe drugs to suitable patients only and patients to take their drugs as prescribed, the FDA’s role in this regard appears more ambiguous. The question of whether FDA advisory committees ought to base their recommendations, in part, on the risks that may arise from off-label use and drug overdose has cropped up repeatedly but has never been met with a definitive answer. The extent to which sponsors are wary of FDA concerns in this regard is illustrated by the availability of safety studies of drugs intended for use in chronic obstructive pulmonary disease (COPD) in asthma patients, who might conceivably be prescribed the drug for off-label use. It is important to note that the mandate of both the FDA and the EMA encompasses only the assessment of a drug’s clinical risk–benefit. Economic considerations are outside the scope of the regulatory review process, and regulatory approval does not guarantee that a drug will receive reimbursement at a price acceptable to the sponsor. The drug discovery, development, and approval process is lengthy: More than 10 years can elapse between the first description of a potential drug target in the literature and the launch of the first drug to interact with that target. Clinical development alone is a multi-year process; the duration depends on the scope of the clinical and analytical work to be performed, drug firms’ decision processes, and possible delays caused by such things as problems with the stability of a drug’s formulation or having to put a trial on “clinical hold” while an observed imbalance in adverse events is being investigated. Although timelines may vary widely as a function of various requirements, the following guide may be used as a starting point: Phase I studies can usually be conducted and analyzed in a matter of months, whereas a full phase II program can rarely be completed in less than a year because it often comprises multiple studies, with treatment durations of up to six months relatively common and even longer durations under certain circumstances. The phase III program usually lasts at least two years, with treatment durations of at least one year 16



and additional time for patient accrual, follow-up, and data analysis. It may take significantly longer in the case of very large trials for which patient recruitment takes a long time or in the event of very long treatment durations or the need for extensive patient follow-up. But pivotal trials can also be much shorter (e.g., for anti-cancer drugs targeted at particularly lethal tumors). Most clinical trials that are relevant to the analysis of the pharmaceutical industry majors are listed at http://clinicaltrials.gov, where expected timelines are usually provided. The regulatory review process may amount to six months or less if an expedited or priority review is granted—for example, for drugs that have received the FDA “fast track” or “breakthrough therapy” designation. Standard review processes tend to take approximately one year in most key territories. In the event of a rejection based on major clinical deficiencies, it may take years to address the regulators’ concerns. Although a drug may be launched within days or weeks of approval, a delay of one year is not uncommon in regions known for drawn-out reimbursement negotiations.


17

INTELLECTUAL PROPERTY: PATENTS, REGULATORY EXCLUSIVITIES, AND OTHER FORMS OF PROTECTION As discussed in the previous section, discovering a drug and getting it to market is a lengthy and resource-consuming process. Manufacturing and distributing a drug are relatively straightforward by comparison, although pharmaceutical production has its own challenges, notably in the context of the industry’s shift toward biopharmaceuticals, which are typically produced by genetically engineered cells. To perform the significant amount of pre-clinical and clinical work required to establish a drug’s safety and efficacy, originators of new drugs require incentives in the form of periods of market exclusivity during which they can earn a return on their investment. Essentially, there are two levels of protection: patents and regulatory exclusivities. These forms of protection run in parallel—that is, an off-patent drug may not be copied by generics players while regulatory exclusivities are in place, and generics companies must demonstrate that existing patents are invalid or not infringed by their product if they wish to launch a generic once regulatory exclusivities run out. The rules and legislation around both forms of protection are exceptionally complex, and readers should be aware that the following discussion merely scratches the surface. Patents are issued by patent offices. The strongest protection is typically afforded by the patent on the active ingredient by which a drug exerts its biological effect. So long as the active ingredient is protected, the drug itself is protected; any drug containing a different active moiety would be considered a different drug, not a generic, and would need to complete a full clinical program before obtaining approval. However, substance patents tend to be the first to expire in the patent estate surrounding a drug. Although substance patents have a life of 20 years from the date of issue, they are normally granted at an early stage of the lengthy drug discovery/development process. By the time a drug launches, the active-ingredient patent is often less than 10 years from expiration. In the event of severe delays in the drug development process, a substance patent can even expire prior to launch. Various provisions allow drug makers to extend their drugs’ patents by a number of years. For example, products marketed in Europe may receive supplementary 18


Intellectual Industry Overview Property

protection certificates (SPCs) that add up to five years of protection; under the Hatch–Waxman Act patent term extension provisions, US patents may be extended by up to five years to compensate drug firms for some of the time that compounds spend in development or registration. Even so, the active-ingredient patent is likely to expire earlier than weaker forms of patents, such as formulation, process, or use patents. A generics company may be able to circumvent these other patents—for example, by changing the drug’s excipients or key steps in the production process. Use patents, which preclude generics from being used in certain disease settings, can be difficult to enforce. For completeness, trademarks are worth mentioning. Although they do not play a pivotal role in the protection of most drugs against erosion by generics, they may add an extra level of protection in some cases, especially for drugs administered in a device, such as an injection pen or inhaler. Over time, patients may become loyal to their device and balk at the notion of having to use a generic that comes in a device with a different “look and feel.” It is incumbent on the sponsor of a generic to assert that its product is not infringing any valid patents. For example, when a company submits an abbreviated new-drug application (ANDA), or generic file, to the FDA, the application must contain either a paragraph III or a paragraph IV certification. In the case of a paragraph III certification, the FDA holds off on final approval until all the patents listed in its Orange Book database have expired; a paragraph IV filing reflects the generic sponsor’s conviction that unexpired Orange Book patents are either invalid or not infringed. The branded drug company is informed of all paragraph IV filings that are based on one of its brands as a reference product and may sue a generics company within 45 days of such notification if it concludes that its patents are valid and would be infringed by the generic. In the event of a lawsuit, the FDA is banned for 30 months from approving the generic unless there is an earlier court decision in favor of the generic’s company. This stay is often referred to as a Hatch–Waxman stay. The validity or invalidity, as well as the infringement or non-infringement, of patents is determined by the courts. A court may invalidate patents on such grounds as obviousness or prior art, or it may rule that a patent is unenforceable owing to inequitable conduct. If a court finds patents to be valid and enforceable, the generic may be launched only if it does not infringe them. The court’s ruling may be appealed. If a generic is launched while litigation remains ongoing, the launch is considered “at-risk,” meaning that the generic’s company may be liable for damages if it is later found to have infringed any valid and enforceable patents. Owing to the high level of uncertainty around the outcome of litigation, it is not uncommon for the makers of the branded drug and the generic to settle their litigation. Settlements typically result in a launch date for the generic that ©2013 CFA INSTITUTE

19


falls somewhere in between the assumed launch dates under various hypothetical court judgments. Settlement agreements must be structured so as to ensure that the health care system or consumer is not disadvantaged; “pay to delay” deals, in which the branded drug maker pays the generic’s company to hold off on a launch and delay the legitimate entry of its generic, are unacceptable because they deprive the health care system of potential savings. A pharmaceutical company that embarks on the discovery and development of a new drug thus faces substantial uncertainty about its patent estate. The post-launch life of the active-ingredient patent may be difficult to predict, and there are no guarantees that the company’s patents will be upheld in court or that key patents will not be circumvented by generics companies. This uncertainty might conceivably deter the drug maker from investing in large-scale clinical trials, especially if the patent estate appears relatively weak (e.g., in the case of a molecule that was discovered and patented early), with the result that the active-ingredient patent might expire before or shortly after launch. Similarly, a drug maker might refrain from developing drugs for niche indications if there is a high risk that the drugs’ sales might be too low to earn an adequate return on investment before their patents expire. Of course, decisions against the development of drugs that hold promise from a medical perspective risk being detrimental to patients who might face a dearth of treatment options. Regulatory exclusivities offer intellectual property protection independent of patents in order to incentivize drug firms to invest in drug candidates. A plethora of regulatory exclusivities are available; the following discussion is confined to the most common forms in the United States and the EU. The FDA awards five years of exclusivity for new chemicals and three years for a “change,” such as a new formulation. If a drug maker establishes the efficacy and safety of its drugs in children, pediatric exclusivity is awarded, adding another six months to the exclusivity period conferred by patents or other regulatory exclusivities. So-called orphan drugs, targeted at conditions that affect fewer than 200,000 people in the United States, receive seven years of exclusivity. The EMA awards 10 years of exclusivity to new drugs, including 8 years of data exclusivity, during which generics companies may not reference the originator’s data, and 2 years of market protection, during which generics may not be approved; new indications may entitle the drug firm to a 1-year extension. Successful development of a drug for pediatric patients renders the applicant eligible for a six-month patent term extension. Orphan drug exclusivity—granted for drugs that target indications affecting fewer than 5 in 10,000 Europeans or that would be unlikely to yield a sufficient return on investment for other reasons—lasts for 10 years, with a 2-year extension possible if a new orphan indication is added. Table 1 summarizes the main forms of protection for new drugs. 20


Intellectual Industry Overview Property

Table 1. Summary of the Main Forms of Protection of New Drugs against Generics Patents and Trademarks Active-ingredient patents Formulation patents Process patents Use patents Patent term extensions Trademarks


Regulatory Exclusivities New-drug exclusivities Exclusivities with respect to changes (e.g., new formulations or indications) Pediatric exclusivity Orphan drug exclusivity

21

BUSINESS MODELS The industry majors vary with respect to their degree of and approach to diversification. Some firms are essentially “pure plays” that offer almost exclusively branded prescription medicines for human use. Their success is thus inextricably linked to the growth trajectory and longevity of their marketed drugs as well as the success of their pipelines. The potential rewards are high, but so are the risks. Many players therefore seek to balance their business through exposure to other, more predictable segments of the health care sector. Closely related fields that can be somewhat synergistic with the business of branded human prescription drugs include OTC drugs and products for animal health. These businesses are characterized by limited patent exposure and significantly reduced R&D risk, with shorter payback periods for R&D spend, steady growth, and organic growth rates typically in the low to midsingle digits. Thus, the management teams of large corporations tend to view these businesses as a natural extension of their core pharmaceutical operations. In contrast, the core competencies required to compete effectively in other fields of health care, such as generics, diagnostics, and medical technology, may differ substantially from the skill set and know-how acquired in the course of branded drug development and marketing. Consequently, the industry majors tend to enter these arenas selectively, and many firms operate in these market segments for historical reasons rather than as a deliberate move in recent years. Industrial conglomerates comprising both pharmaceutical and non-health-care operations are rare and usually result from historical developments. For example, the pharmaceutical industry was, to some extent, born out of the chemical industry; as a result, some pharmaceutical players retain chemical operations today. In such cases, the gradual divestment of non-health-care activities over time tends to be more common than deliberate moves to diversify away from health care. Table 2 reports the degree of diversification of various relevant business models. Table 2. Pharmaceutical and Health Care Business Models: Degree of Diversification Business Model Pharmaceutical “pure play”

Brief Description Focus is almost exclusively on branded prescription drugs for human use Balanced drug portfolio May include OTC drugs and products for animal health in addition to branded human prescription drugs Diversified health care company Includes other health care segments (e.g., generics, diagnostics, medical technology) in addition to branded drugs Diversified industrial company Includes activities unrelated to pharmaceuticals with a focus on pharmaceuticals (e.g., agriculture, chemicals)

22

Company Examples AstraZeneca, Bristol-Myers Squibb, Novo Nordisk Eli Lilly, GlaxoSmithKline, Merck & Co., Pfizer, Sanofi Johnson & Johnson, Novartis, Roche Bayer, Merck KGA


Business Models

INDUSTRY CONSOLIDATION Consolidation has been a long-standing theme of the pharmaceutical industry and has given rise to a number of very large organizations with tens of billions of dollars in annual revenues. Nonetheless, the industry remains fragmented, with even the largest players commanding mere single-digit market shares. Although this situation leaves room for future megamergers, the combination of two pharmaceutical giants is rarely the alternative preferred by the companies’ management teams. Managing extremely large organizations is fraught with challenges. Although synergies may substantially decrease the combined cost base—notably, selling, general, and administrative expenses (SG&A)—the absolute amount of revenue expected to be lost to patent expirations in future years rises dramatically, thereby increasing the pressure on the combined pipelines to deliver. (For this reason, an “ideal” takeover candidate would typically be the mirror image of the industry majors, in the sense that it would combine a modest current revenue base with a potentially transformational pipeline. However, such acquisition targets are few and far between.) Antitrust considerations are another potential obstacle to M&A that should not be underrated. Owing to the high degree of fragmentation of the pharmaceutical market, there is a high risk that the combination of two pharmaceutical firms could result in their having a dominant position in one or more market subsegments—for example, multiple sclerosis or a specific type of tumor. This result would trigger mandatory divestment of one of the drugs targeting that disease area, which could dilute the shareholder value of the merger transaction. Antitrust considerations are not limited to classical pharmaceutical businesses. In fact, the combination of two animal health businesses could be exceptionally problematic from an antitrust point of view, with mandatory divestments particularly hard to execute. A number of the major players have exposure to animal health, and a situation could conceivably arise in which that business activity would effectively block mergers that might otherwise appear attractive with respect to the firms’ human drug businesses. In light of these considerations, many pharmaceutical executives in recent years have expressed a preference for “bolt-on” acquisitions—for example, to reach critical mass in areas where the company had previously been underrepresented—and for drug-licensing deals over large transactions. Owing to an abundance of small and midsize drug companies, this approach also offers more choice than megamergers do. Numerous biotechnology companies are working on drugs for the treatment of conditions with high unmet need or offer platform technologies that may enhance the drug discovery process or the features of drug candidates themselves. These players range in size from small venture-capital-funded start-ups and vehicles spun out of universities to midsize and large companies that own the rights to cash-generative marketed drugs wholly or in part. Depending on the scale of their operations as well as their focus and outlook, these firms may rely to a significant ©2013 CFA INSTITUTE

23


extent on large pharmaceutical companies to provide funding and clinical expertise in the drug development process and a commercial engine to sell the drug following approval. Occasionally, the industry majors opt for the outright acquisition or purchase of a stake in biotechnology companies. In general, however, full ownership of R&D assets is not considered of paramount importance. In light of the plethora of scientific approaches being explored, it seems impossible to predict which technologies, mechanisms of action, and therapeutic categories hold the greatest promise. Unsurprisingly, many large pharmaceutical companies seem content to “part-own” a large number of different assets and are reluctant to place large bets on unproven scientific approaches. Therefore, licensing arrangements are common in the industry. In a typical licensing deal, a large pharmaceutical player acquires commercialization rights to a biotechnology company’s drug candidate globally or in key territories. In return, the pharmaceutical company makes an upfront payment to the biotechnology firm and commits to further success-based milestone payments and a royalty on future drug sales. The upfront payment usually compensates the biotechnology firm for work carried out to date and often amounts to millions or tens of millions of dollars. Milestone payments are typically linked to clinical, regulatory, or commercial milestones (e.g., the transition to advanced stages of clinical development, regulatory approval, or the achievement of prespecified sales targets following launch). Milestones may be structured to compensate the biotechnology firm for future costs (e.g., expenses for clinical development if the originator has agreed to bear part of the clinical costs under a co-development clause) or as success-based milestones. Royalty rates are subject to negotiation and are determined by various factors, most importantly the drug candidate’s development stage, the quality of the preclinical and clinical data available at the time the agreement is entered into, the compound’s commercial potential, and the relative balance of the total economics between upfront payments, milestones, and royalties. Tiered royalties, in which the royalty rate itself depends on product sales, are relatively common. Singledigit royalties are often agreed on for compounds licensed at an early stage of development, whereas double-digit royalties—usually in the teens, occasionally in the twenties or higher—are the standard for compounds in the advanced stages of development that come with impressive sets of clinical data. Although some biotechnology companies prefer to leave drug commercialization to their pharmaceutical partners, others are keen to field a small force of their own, especially on their home turf, and may negotiate co-commercialization and profit-sharing rights or options. Licensing deals based on 50/50 profit sharing in major territories are rare but not unprecedented. 24


Business Models

A major pharmaceutical firm’s activity with respect to the licensing of drug candidates and other collaborations with the biotechnology industry and academia reflects its commitment and approach to innovation. Across the industry, the majors acknowledge the impossibility of building in-house expertise in every scientific hot spot and increasingly rely on external innovation for that reason alone. In light of the restructuring initiatives taken in recent years in anticipation of the “patent cliff,” many players have sought to rebalance between fixed and variable R&D spend and thus, by extension, internal and external R&D expenses. In addition, the jury is still out on “optimal” levels of R&D spend, both in absolute terms and relative to pharmaceutical sales. Some players reinvest more than 20% of their branded prescription drug sales in R&D each year, whereas others opt for an R&Dto-sales ratio in the low teens, with obvious implications for the budget amount allocated to drug licensing.

NOTABLE TRENDS Three general trends with respect to business models have been observed across the pharmaceutical industry. First, geographic diversification has become a higher priority for the majors in recent years, as a combination of such factors as patent expirations and a changing regulatory and payer environment has called into question the viability of a strategy that has historically been heavily reliant on Western markets. Second, the composition of portfolios of marketed drugs as well as pipelines has shifted away from primary care and toward specialty pharmaceuticals. Even such therapeutic niche areas as rare cancers and orphan diseases have lately attracted significant interest from the pharmaceutical industry. Although many firms remain committed to the primary care business, they have embraced specialty care as a second pillar, with its lower marketing spend, higher margins, and often higher success rate in clinical trials making up for the slightly more modest peak sales potential. Third, the industry has reduced its infrastructure and generally cracked down on fixed cost in recognizing that the top line may fluctuate over the years as a result of patent expirations. The industry majors are now increasingly relying on external suppliers while seeking ways to achieve more with fewer resources. Excess production capacity, duplication of infrastructure across countries, and even such comparatively minor items as the discretionary purchase of excess R&D supplies by individual teams are now buried in the past. Some companies continue to generate hundreds of millions of dollars in procurement savings each quarter—an impressive demonstration of the extent to which these firms are eliminating waste. Whether this newly found cost-consciousness bears risks is open to debate. The management teams of pharmaceutical firms maintain that adequate quality control procedures reduce the risk of outsourcing to the same level of risk as in-house ©2013 CFA INSTITUTE

25


activities and attribute glitches (e.g., FDA warning letters or supply disruptions) to such factors as human error and heightened regulatory scrutiny. In general, any efforts to reduce the cost base or render it more flexible may carry an inherent risk of opportunity costs and diminished control. Consequently, a modest reversal of recent trends may be in the offing as the industry leaves its “patent cliff” behind and as pipelines deliver.

26


FINANCIAL STATEMENT ANALYSIS The structure of the drug industry’s financial statements is relatively straightforward. The main challenge for the analyst is not the interpretation of historical data but, rather, accurately forecasting the myriad influences on each line item. Sets of financial results typically include detailed reports of sales by drug and region as well as extensive product-related disclosures, such as key drivers of the sales development of individual drugs in the period under review and a pipeline update. Annual reports usually also contain information on patents and regulatory exclusivities. Owing to the sheer size of the industry majors as well as the nature of the business, quarterly results are often heavily distorted by such “one-time items” as product-related impairment or restructuring charges. Many players have made large acquisitions and recognize significant amortization charges, which limits the comparability of results. Thus, in addition to the statutory accounts, most of the majors provide adjusted results, often referred to as “non-GAAP” results by US companies and as “core” by their European counterparts. Although adjustments usually include the add-back of impairment charges, other exceptional items, and amortization charges, subtle differences may exist between different players’ methodologies. For example, different companies might use different criteria (e.g., different materiality thresholds) in labeling items as exceptional; some might add back all amortization charges, whereas others might confine adjustments to charges concerning the acquisition of companies. Analysts should refer to the notes on each company’s methodology to ensure the correct interpretation of adjusted results; analysts may consider making additional adjustments to ensure adequate comparability for their particular analytical purposes. Even if two companies adjust their statutory results in a similar manner, their operating margins and margin structures are not normally comparable. As previously mentioned, pharmaceutical companies rarely own 100% of all their drugs and drug candidates. Depending on the specific clauses of licensing, collaboration, co-development, and other agreements, a company may book some or all of a drug’s in-market sales and may make or receive royalty and other payments with respect to a drug under existing cost- and profit-sharing agreements. Accounting rules for individual line items may also vary across geographic regions, especially for R&D spend. Although most stakeholders tend to think of the industry’s R&D activities as an “investment,” in-house R&D spend is generally expensed in the period when it is incurred under both US GAAP and International Financial Reporting Standards (IFRS). Rules vary, however, with respect to “external” R&D spend, such as upfront and milestone payments made under licensing agreements. Although such payments may be expensed immediately ©2013 CFA INSTITUTE

27


under US GAAP if they do not relate to an acquisition, they are usually capitalized and amortized over their useful lives under IFRS. Thus, an “all-in” measure of cash R&D spend in any given period may not be readily available for European companies. A rough estimate may be gleaned from asset schedules. Perusal of the press releases and financial statements issued by a drug firm’s biotechnology partners (who may be required to disclose the receipt of milestone payments that would be considered immaterial by the pharmaceutical firm that owes the payments) is a cumbersome and incomplete approach. R&D productivity is a chief concern of analysts and investors who seek reassurance that the billions of dollars many of the majors plow into drug discovery and development each year is money well spent. Unfortunately, disclosure around R&D expenses is usually vastly insufficient for this purpose, and any measure of R&D productivity based on financial statement analysis alone is bound to be a crude approximation at best. Companies rarely provide a detailed breakdown of their R&D spend by drug, indication, or type of molecule. Occasionally, drug makers may separate out the contribution of the “R” (early research) and the “D” (clinical development) to overall spend. It is virtually impossible to estimate the expense incurred with respect to new molecular entities as opposed to such LCM initiatives as new formulations. Furthermore, drugs may not attain their full sales potential on the sole basis of clinical data from registration trials, and a significant portion of a company’s R&D budget may be dedicated to further characterizing a drug that is already commercially available (e.g., by conducting “outcomes” trials designed to prove that it increases life expectancy or delays heart attacks). Analysts should also be mindful of the time lag between R&D efforts and regulatory approval. The mere approval of a drug does not necessarily mean that a company is about to earn an adequate return on the R&D spend incurred with respect to the molecule; for example, modest market uptake or premature genericization may have a dampening effect on its net present value (NPV). In light of these considerations, jumping to conclusions about trends in R&D productivity on the basis of such simplistic and static measures as contrasting industry R&D spend in any given year with the number of new drugs approved by the FDA in the same year is ill advised. It follows from the discussion thus far that a pharmaceutical company’s operating margin is heavily dependent on key drugs’ stages in their life cycles, the proportion of drugs’ cash flows owned by the company, and the accounting recognition of drug sales and profits, as well as the amount and recognition of discretionary R&D spend that may or may not yield an attractive return many years down the road. These factors render the comparison of margins and margin structures across companies largely meaningless, although benchmarking individual line items, such as the R&D-to-sales ratio or the administrative spend across companies, may occasionally be useful in identifying potential inefficiencies. Business mix further 28


Financial Statement Analysis

complicates such comparisons; for example, some of the majors generate significant sales with vaccines, a segment characterized by both lower gross margins and lower R&D spend than most other branded prescription drug businesses. Similarly, year-over-year comparisons of financial results are hardly insightful because most pharmaceutical businesses are not seasonal, and year-over-year growth rates may be heavily distorted by such factors as generics competition for key drugs that may have set in over the past 12 months. As discussed later in more detail in the section on forecasting, the most viable analytical approach is to track the main influences on each profit and loss line item and ascertain whether any future change should be expected. For example, gross margins of “classical” pharmaceutical businesses often exceed 70%. The absolute margin level is determined primarily by the operating leverage arising from blockbuster drugs that are simple to produce and pay-aways to third parties under licensing and similar agreements. Such factors as product and geographic mix, inventory adjustments, variations in biopharmaceutical production yields, and fluctuating raw material prices may entail modest fluctuations in gross margins; however, structural change should usually be expected only in the event of significant sales growth or decline—or if pay-aways change as a result of a significant growth or decline in the underlying sales of the related products or upon coming to a contractual end. R&D spend is largely discretionary in the sense that it is not directly linked to revenues. (Note that much of it is fixed, and a significant reduction would typically require major restructuring.) For most of the majors, the R&D-to-sales ratio is in the teens; occasionally, it may exceed 20%. R&D spend that remains far in excess of 20% of sales over time would likely invite investor scrutiny as to whether that level of spend is sustainable. SG&A spend, which may exceed 30% of sales, has traditionally been high in the pharmaceutical industry and has become a key source of productivity savings in recent years. Analysts should regularly ascertain the scope of further reductions in the absence of major investments in new-drug launches or infrastructure in fast-growing geographic regions, such as some emerging markets. Underlying operating margins north of 30% are by no means rare, and the margins of the most successful companies that own most of their assets and/or receive substantial royalty income may even reach or exceed 40%. At the same time, operating margins in the 20s are frequently observed in companies that have had major patent expirations, owe significant royalty payments to third parties, or have a strong commitment to innovation that translates into high R&D spend. Items “below the line” warrant careful analysis. Some companies may have significant associate income or minorities that cannot simply be extrapolated. As discussed later in the context of balance sheet structures, leverage differs widely, suggesting wide variations in the net financial result. Tax rates may also vary among companies for structural reasons. Finally, most diversified health care companies ©2013 CFA INSTITUTE

29


report segments in substantial detail. Owing to the significant differences in dynamics and margin structure between segments, analysts would be well advised to model each segment separately. Analysis of the balance sheet and cash flow statement is typically straightforward. Most successful pharmaceutical businesses are characterized by low capital intensity, high cash generation, and conservatism with respect to financing structures. Although the required level of investment in new production sites has increased somewhat as a result of a shift in product mix away from simple “small molecules” to more complex biotechnologically produced drugs, growth is often possible in the absence of major investments in property, plant, and equipment, and capital expenditures are rarely a concern. It is worth noting that many firms have reduced or are in the process of reducing their manufacturing footprint pursuant to the operational excellence initiatives they embarked on to brace for the patent cliff. Of all the restructuring measures announced and implemented by the majors, site closures and divestments will require the most time because the continued supply of medicines must be maintained and the transition to other factories or external organizations requires regulatory approval. Many firms, especially those that have made large acquisitions, have significant intangible assets. It is not unusual for the majors to carry intangibles worth tens of billions of dollars on their books, often dwarfing the value of tangible assets. Intangibles other than goodwill are amortized over their useful lives, often giving rise to substantial amortization charges over extended periods—hence, the industry’s focus on non-GAAP or core results. In contrast, the impairment of goodwill does not normally occur with high frequency. The book value of the intangible assets pertaining to the major individual products that were purchased in an acquisition is typically provided in annual reports, giving the analyst a rough idea of the potential impairment charges in the event of a severe problem with a flagship product. When analyzing and comparing working-capital requirements, analysts should be mindful of the length of biopharmaceutical production cycles—biotechnologically produced medicines and other complex drugs may take nine months or more to produce—the high opportunity cost of supply disruptions, and differences in payment terms between geographic regions and distribution channels (e.g., retail or hospital pharmacies). Owing to the relatively low variable cost of the production and distribution of blockbuster drugs, successful pharmaceutical companies tend to be highly cash generative, with free cash flow before acquisitions often on the level of net income. It is not rare for drug firms to accumulate significant net cash positions in the absence of M&A and to deleverage rapidly in the wake of major acquisitions. Consequently, it may be possible for pharmaceutical firms to gear up with respect to acquisitions without suffering major downgrades by the rating agencies. Share buyback programs 30



are a popular tool for raising structurally low debt-to-equity levels. At the same time, most pharmaceutical firms are conservatively financed to ensure that they would remain solvent under scenarios of extreme stress (e.g., the sudden withdrawal from the market of a multi-billion-dollar product). It is also important to keep in mind that bouts of innovation may not always coincide with major patent expirations; executives thus prudently brace for potential periods of much lower sales and cash flows. Many drug firms carry relatively high provisions (e.g., for pensions and other post-employment benefits for their large workforces). Under IFRS accounting, European companies may also provision extensively for the expected cost of litigation. Industry novices are often puzzled by the very high return on capital employed (ROCE) generated by many of the industry majors despite sizable positions in intangibles; some even regard these supernormal returns (ROCE in excess of 20% is not uncommon) as an indicator that drug prices are too high and thus unsustainable. When analyzing return measures, it is important to recall that success or failure in the pharmaceutical industry is a function of R&D productivity. By definition, the industry majors are those that generate the highest drug sales and are thus the winners in the race to develop new drugs. Because pharmaceutical operations are not capital intensive and a significant part of a company’s R&D spend is expensed and may thus be thought of as a “sunk” cost rather than invested capital, the denominator in return-on-capital calculations tends to be relatively low and the numerator is essentially a function of past R&D success. The industry is highly fragmented; although the largest and most successful players may be earning supernormal returns, those that have remained subscale because of pipeline failures may struggle. Elsewhere, past R&D success does not guarantee future success, and the ROCE of companies that fail to continually innovate would likely dwindle in the wake of patent expirations. This situation was illustrated by the investment community’s “confidence crisis” in 2008, when the valuations of the majors plummeted despite high levels of profitability, and some investors exerted substantial pressure on company management to downsize or shut down their R&D engines. Thus, it seems a foregone conclusion that the gap between the industry “winners” and “losers” will always remain wide and that the most important task of a pharmaceutical analyst is to identify the most innovative companies that own the most promising drugs, as discussed in the next section.

FORECASTING DRUG SALES AND COMPANY PROFITS It follows from the previous discussion that forecasting drug sales and profits of pharmaceutical companies requires a “bottom-up” approach to modeling, with ©2013 CFA INSTITUTE

31


“top-down” approaches merely serving as a cross-check. Forecasting drug sales product by product—and if companies provide a geographic breakdown, territory by territory—should be the starting point of pharmaceutical modeling. To the extent that off-patent products that have passed their peak sales make a meaningful contribution to group revenues, it may be sufficient to extrapolate the modest growth or decline they have exhibited in recent years. In the case of products that are approaching their peak sales, considering the erosion to be expected upon the expiration of key patents and monitoring the competitive landscape for emerging threats may constitute an adequate approach. However, products at a relatively early stage of their life cycle and pipeline projects require a full analysis if the accuracy of forecasts is to be maximized. A complete analysis of a drug’s or drug candidate’s potential requires an analyst to consider all commercial, regulatory, and legal aspects of drug development and marketing, including the timing of market entry in each region where the drug is intended to be launched, the addressable target market in each of the indications for which the drug is being developed, the drug’s price point, the average treatment duration and amount of drug taken, its penetration rate at peak, the growth trajectory to peak sales and beyond, the timing of loss of exclusivity, the expected decay curve following the generic’s entry, and the likelihood that the drug will reach the market. Once the projected sales trajectory has been included in an analyst’s model, the analyst must continually revisit the underlying assumptions and actively monitor all trends and news flow that may have a bearing on these inputs. For example, an evolving competitive landscape or an emerging safety issue may alter a drug’s projected peak penetration rates. The same is true for slower or faster uptake in the marketplace than originally expected. In fact, many analysts track launches closely. For the United States, weekly and monthly prescription data are available from various sources and are widely used to form an opinion of a drug’s acceptance in the marketplace ahead of formal financial reporting by the company. A firm may develop a drug in more than one indication, and the success probability in each indication must be adjusted as the clinical development program progresses. All these inputs must be derived from the myriad pieces of publicly available information that are characteristic of the pharmaceutical industry (please see the section on industry resources). Although the analysis and conclusions vary by indication, type of molecule, and other factors—and an exhaustive review is thus beyond the scope of this report—it is worthwhile to briefly touch on the key determinants of a drug’s commercial success, chiefly to enable users of research to challenge the peak sales assumptions of consensus and individual analysts. The expected timing of market entry must be ascertained separately for each indication and territory. Although the pharmaceutical industry is increasingly aiming to conduct global trials and to submit dossiers to regulators in key territories almost 32



simultaneously, regulatory or commercial considerations may result in significant time lags. Because the efficacy and safety of new molecular entities must be ascertained separately for each targeted indication, a drug may initially be approved for use in one indication only (e.g., rheumatoid arthritis), with other indications added to the label at a later stage (e.g., psoriasis and other autoimmune conditions). Many companies routinely provide information on the expected launch timelines of a drug candidate for key regions and the most advanced indications. Although there is usually a reasonable basis for this information, a quick plausibility check—based on the expected accrual rates, treatment duration, follow-up in the main clinical trials, and expected regulatory review period—is always warranted. Although drugs are typically launched within days or weeks of approval in some countries, especially the United States, drug launches are preceded by lengthy reimbursement negotiations in others. In some EU countries, drug launches may be delayed by months—occasionally, even by a year or more—as price and reimbursement negotiations drag on. As a result, drug companies typically refer to the European “rollout” rather than “launch” of a drug. Analysts are usually left to their own devices when forecasting the price point of a new drug because pharmaceutical firms typically regard pricing as a highly sensitive detail of their marketing strategy. As a rule, companies will aim for the highest price at which broad reimbursement can be secured. (Most drugs are reimbursed in most key territories, with the exception of certain therapeutic categories. For example, most consumers pay out of pocket for many women’s health products, such as contraceptives or fertility treatments, implying that consumer behavior and price elasticity play an important role in setting the prices for such medicines.) Pharmaceutical companies perform pharmacoeconomic studies to tag the savings that health care systems stand to make by reimbursing drugs that can reduce the latestage complications of diseases, keep patients out of hospitals or nursing homes, or reduce the length of hospital and intensive-care-unit stays. In most countries, health technology assessment bodies and payer organizations verify these assumptions or conduct their own analyses. For drugs that target conditions that were not previously amenable to drug treatment, the cost of such alternative treatment options as surgery may give a first indication of a realistic price point. Where competing drugs are already commercially available, their price points may be used as a benchmark. New entrants may be priced at a premium to the incumbents to reflect a superior value proposition—or in line with the incumbents or at a discount designed to gain rapid acceptance despite a late entry. Because many drugs are launched globally, pharmaceutical companies must strike a balance between, on the one hand, harmonizing prices across regions in an effort to minimize parallel imports, price pressure from interwoven national reference price systems, and general frustration on the part of patients ©2013 CFA INSTITUTE

33


and prescribers and, on the other hand, implementing regional pricing strategies in order to tap into additional volumes. Since rebates and access programs for uninsured or underinsured patients are common across the industry, analysts also need to be mindful of the differences that may arise between list prices and average net realized prices. In some territories, notably the EU, reimbursement of additional indications that were added to the label after the launch in the lead indication typically requires price concessions. Many drugs tap into potentially very large markets; numerous conditions affect millions of patients in the United States and the EU alone, with tens or even hundreds of millions of sufferers around the world. Multiplying global disease prevalence with a reasonable price may thus yield target markets valued at tens of billions of dollars. This approach thus carries a high risk of overstating a drug’s commercial potential. Eliminating patient populations that are unlikely to receive the drug is of paramount importance in deducing the addressable target patient population. Estimates of disease prevalence furnished by various organizations often include patients who have not been diagnosed. There are many reasons why conditions may go undiagnosed; for example, the patient may live in an area where access to health care is poor, or the patient may be asymptomatic and the condition may not be diagnosed during routine exams. Unless there are sound reasons to believe that future diagnosis rates will increase substantially, it is prudent to use the number of diagnosed patients as a starting point for any market-sizing exercise. The inclusion and exclusion criteria of clinical trials and drug labels provide valuable clues to patient eligibility. For example, use of a drug may be confined to patients with a moderately severe disease who present with a specific disease phenotype and certain biomarkers, who are in good overall health with good liver and renal function, and who receive concomitant therapy that has been tested in combination with the new drug in clinical trials, such as certain chemotherapy backbones in cancer treatment. It is equally important to keep in mind that not every patient who is eligible to receive a drug from a certain class will necessarily be treated. Although physician awareness of new drugs tends to be high because clinical results are routinely presented at key medical congresses and are published in peer-reviewed journals—and drug firms employ field forces that provide detail to individual prescribers—there are numerous reasons why patients and their physicians may decide against a certain course of treatment. Poor compliance may shorten treatment duration or dose intensity relative to that tested in clinical trials. Depending on the condition, disease awareness campaigns may be required before drugs gain broad acceptance. Competing drugs may either vie for share or expand the market. Usually, various drugs in development have the potential to displace the incumbents, and analysts 34



need to monitor clinical trial news flow and the scientific community’s reaction to clinical results. The attribution of success probabilities to any drug candidate in any indication is highly subjective. Although historical average values (less than 40% of phase II and less than 80% of phase III assets typically reach the market) may be used as a guide, analysts may choose higher or lower values for any drug candidate depending on the published pre-clinical and clinical data, the therapeutic category and history of the therapeutic class, the design of the pivotal trials, the company’s track record, and various other factors. It is important to remember that the success or launch probability of a drug candidate is only an analytical tool to capture the risk associated with clinical development. The events are binary: A clinical trial will either succeed or fail, and regulatory approval will either be granted or denied, implying that the ex post success probability is always 100% or 0%. In the case of the industry majors, whose pipelines typically comprise dozens of assets in phase II or III, the law of averages may apply, suggesting that actual peak sales for the pipeline as a whole may approach risk-adjusted forecast peak sales, provided that all other assumptions also prove correct. However, for a small biotechnology company, the success or failure of a drug may “make or break” the investment case. The purely mechanistic view that multiplying “peak sales before risk adjustment” by an assumed success probability yields “peak sales after risk adjustment” does not truly reflect the ambiguity associated with forecasting drug sales, especially in the early stages of development. Analysts and investors should be mindful of the feedback loop between success probabilities and other inputs, such as the addressable target market, price, and penetration rates. For example, a phase III trial may succeed in the sense that the primary efficacy endpoint is met, but the side effect profile may be such that analysts feel compelled to revisit their assumptions with respect to the compound’s suitability for different patient populations and price point. In fact, in the early stages of development, the limited clinical information available can make it hard to predict whether the drug is more likely to be a mainstream or a niche product or whether it merits a price premium or discount to competitors. Essentially, two avenues are available for dealing with this ambiguity. The first is to model a drug’s risk-adjusted peak sales under various scenarios and then either pick a base-case scenario or probability-weight the scenarios. This approach may be suitable for drug candidates that have the potential to transform earnings in a bestcase scenario and that have a wide range of potential outcomes. A simpler approach is to reflect the ambiguity with respect to such parameters as patient populations, price, and penetration in the choice of success probability, which implicitly accounts for the possibility that a drug may reach the market but fall short of its maximum commercial potential. Table 3 presents an example of forecasting a drug’s peak sales. ©2013 CFA INSTITUTE

35


Table 3. Forecasting the Peak Sales of a Fictitious Drug for Patients with a Certain Mutation

Disease prevalence (in millions of patients) (A) Proportion of patients harboring the relevant mutation (B) Target population in millions (C = A × B) Expected average drug price/month (D)

Expected average treatment duration in months (E) Expected revenue per patient (F = D × E) Potential target market (millions) (G = C × F) Expected peak penetration rate (H)

Expected peak sales before risk adjustment (millions) (I)

Success probability (J)

Expected peak sales after risk adjustment (millions) (K = I × J)

US 1.0

EU 1.4

Rest of World

Commentary

A companion diagnostic test to screen for the mutation is under development

10%

10%

0.1

0.1

$3,000

$2,500

7

7

$21,000

$17,500

$2,100

$2,450

Based on market share analysis in light of competing products in development Potential assumed to be similar to Europe; bottom-up modeling not feasible for rest-of-world region Based on clinical data presented at a recent medical congress

25%

$525

40%

$210

20%

$490

40%

$196

Based on competitor product targeting a different mutation of the same disease Based on treatment duration in clinical trials

$490

40%

$196

Included in analyst’s financial forecasts

As a rule of thumb, a typical drug attains its peak sales within five years of launch. However, the growth trajectory may vary between indications. For example, cancer medicines may reach their peak rapidly because the patient pool turns over rapidly, whereas the uptake of new drugs for the treatment of such chronic conditions as asthma tends to be slow owing to low switching rates between products. Predicting the timing of a generic’s entry and the original brand’s subsequent revenue decline can be a daunting task, with significant implications for profit forecasts. Determining the 36



earliest possible date of a generic’s entry on the basis of a drug’s patent estate alone may be anything but straightforward. The duration of patents and exclusivity periods is not always known with certainty, and the level of protection afforded by formulation and process patents, which often expire after the compound patent, may be difficult to gauge. To the extent that “at-risk” launches by generics companies are possible, the question arises whether they are likely to embark on such a course of action. Once a drug loses protection, the question remains whether any copycat versions can clear regulatory hurdles—a subject of intense debate with respect to complex “biologic” drugs that may be difficult to copy. Although some “biosimilars” legislation is now in place in the main territories, various biosimilars players have recently exited the space, at least partly, but even seemingly simple generics of “small-molecule” drugs may be delayed by supply disruptions and other factors. The erosion of the branded franchise following patent expiration typically varies with the intensity of generics competition. As a rule, revenue streams for such simple, small-molecule drugs as pills and tablets tend to largely disappear in key Western markets within a year of the expiration of the main patent or regulatory exclusivity, whereas injectable “biologic” drugs tend to face slower and potentially delayed erosion. It is worth noting that in some emerging markets, notably those characterized by significant patient self-pay and high brand awareness, original brands may continue to grow even after the entry of competing generics. LCM initiatives designed to improve the convenience or tolerability of medicines (e.g., once-daily dosage forms to replace twice-daily medicines or patches to replace pills) may go a long way in protecting a franchise from a competing generic, provided that a large number of patients can be switched to the enhanced product prior to the generic’s entry. Brand loyalty may also partly shield a drug maker from competing generics and may be especially high for drugs that are self-administered in trademarked devices. Accurately forecasting the top line is by far the most challenging task pharmaceutical analysts face. Assuming variable costs of 30% for a fully owned drug, a 1% error on the top line may translate into an error of 2% to 3% on the bottom line. For the reasons already delineated, the peak sales of drugs at an early stage of their life cycle are notoriously hard to forecast, and the gap between drug sales under a realistic best- and worst-case scenario may be extremely wide for some companies. Consequently, an analyst’s level of confidence in a set of forecasts may be of paramount importance to an investment decision. It is equally important to realize that not every drug is fully owned by one company. Collaborations and licensing agreements are common in the pharmaceutical industry, and analysts must track the proportion of in-market sales and profits that accrues to each company. In some cases, a company may book only a portion of in-market sales; in other cases, it may book 100% of sales but pay royalties to a third party. And a company’s economic interest in a drug may be limited to royalty ©2013 CFA INSTITUTE

37


income. To the extent that such pay-aways or income streams are material, forecasting them separately is recommended. For example, if a company is known to pay a mid-teens royalty rate on the sales of a sizable drug to the originator through the cost of goods sold (COGS) line, this aspect should be modeled separately. Extremely lucrative drugs that are simple to make may have gross margins in excess of 95%, whereas incremental COGS as a percentage of sales may exceed 20% for complex biologics. Marginal COGS should also be modeled separately when the impact of a top-line surprise on COGS would be material. It may be impractical to attempt to model the COGS of other drugs with any precision because companies do not normally disclose the exact manufacturing costs or minor pay-aways. Similarly, it is virtually impossible to ascertain variable marketing and R&D expenses associated with specific drugs. On a short-term basis, many of the large pharmaceutical companies provide a fair amount of guidance on the evolution of the individual line items. Analysts should think of much of the operating cost as “fixed,” in the sense that the absolute amount is unlikely to change dramatically from one year to the next absent major developments or restructuring. For example, R&D spend remains relatively constant over the years. If a large number of pipeline projects were to show promise in early clinical development, management would likely prioritize the development of some assets over others, and R&D expenses may rise only moderately. Similarly, analysts should not erroneously assume that R&D spend would fall precipitously upon completion of large, ongoing phase III trials because the clinical work conducted on any one drug rarely ends with submission to the regulators and other assets may be progressing through the pipeline. Most companies control SG&A expenses tightly; nonetheless, new-drug launches may result in upward pressure on spend. Most of the majors operate globally, often in more than a hundred countries, and are thus affected by currency fluctuations. Although many companies operate R&D centers and production sites in various locations around the globe, it is impractical to cover the entire value chain of pharmaceutical operations in each country where the company’s drugs are prescribed, thus limiting the scope for natural hedges. Although many companies hedge transaction exposure in major markets to some extent, appropriate hedging instruments may be unavailable at an acceptable cost in every market. Most companies provide “ready reckoners” and other measures of currency sensitivity to enable analysts to gauge the impact of currency movements on reported sales and profits. In contrast, the currency impact on individual line items can be difficult to forecast. Most companies clearly communicate their dividend policy and, where applicable, their share buyback program. Pharmaceutical companies typically opt for a “progressive” dividend policy, meaning that dividends per share will remain at least at the prior year’s level even in “down years” that may arise as a result of patent expirations. 38


VALUATION OF PHARMACEUTICAL FIRMS The value that investors ascribe to pharmaceutical companies and the industry as a whole is based on perceived innovative power to a significant extent. With the exception of smaller, frequently off-patent “tail” products and ancillary activities, most revenue streams are finite. In fact, the largest revenue and profit streams usually face rapid erosion upon the expiration of patents or regulatory exclusivities, with an inherent risk of diseconomies of scale unless they can be replaced through portfolio rejuvenation or the company substantially reduces its fixed costs in the wake of a pronounced top-line decline. Until the middle of the last decade, one-year-forward P/E multiples in the mid- to high teens reflected the market’s expectation that the industry would be able to innovate so as to maintain the historical double-digit growth rates indefinitely. With patent expirations looming and following numerous clinical setbacks that had left pipelines depleted, many of the majors traded on single-digit or lowdouble-digit one-year-forward P/E multiples in 2008 because the attractive cash flows that the industry continued to generate were widely expected to come to an end in the near future. The structural issues described earlier decrease the utility of many of the standard approaches to valuation that are used routinely in other industries. Peer group analyses are complicated by the heterogeneity of the industry: The marked differences in patent exposure, competitive threats, and pipeline maturity and depth translate into pronounced differences in the predicted growth trajectory and risk–reward profile of individual players. Furthermore, there are relatively few “pure plays” whose exposure is limited to branded prescription drugs for human use; many of the industry leaders are engaged in miscellaneous ancillary activities, including OTC drugs, diagnostics, generics, and products for animal health. Some of the midsize players even have exposure to the fields of chemicals and agriculture, which may merit different sales and profit multiples. Owing to the substantial differences in margins and margin structure that arise principally from the manner in which drug profits are shared between companies and recognized in their financial statements, enterprise value (EV)/ sales ratios are rarely used in the valuation of pharmaceutical companies. All things being equal, a company that books substantially all of the sales of the products for which it has marketing rights but pays high royalties to the ©2013 CFA INSTITUTE

39


originators of those drugs does not merit the same EV/sales multiple as a company that owns 100% of its assets. Also, care should be taken in the case of companies that receive substantial drug-related profits that they book through the “other revenues” or “other operating income” line. Substantial differences in profitability that may render the EV/sales multiple meaningless also arise below the earnings before interest and taxes (EBIT) line, as in the case of stakes in other companies and also tax rates, which can vary dramatically between companies and fluctuate over time. EV/EBIT multiples account for differences in margin structure but may be distorted by high levels of amortization in the case of companies that have made major acquisitions. EV/EBITDA (earnings before interest, taxes, depreciation, and amortization) may be a more suitable measure because it reflects all aspects of current operating performance, with the caveat that the selection of a “fair” discount or premium of a company relative to its peer group requires a judgment call on the company’s future growth and risk profile relative to its peer group, in addition to a view on such “below the line” items as taxes. There is no standard approach to deriving the premium or discount a stock merits; relative valuations reflect both the perceived longevity of current cash flows and the shareholders’ faith in a company’s pipeline and innovative power in general. P/E multiples are generally the most popular metric used to assess relative valuation, and stocks often revert to historical average premiums or discounts unless a company’s fundamentals have improved or deteriorated markedly relative to its peer group. Like EV/EBITDA, P/E multiples reflect the market’s view of a company’s current profitability and future fundamental prospects, with the added advantage of incorporating such nonoperating items as taxes and income from associates. Financial leverage tends to be relatively low in the pharmaceutical industry and is unlikely to distort P/E ratios significantly. A word of caution: There are slight differences between companies’ adjusted EPS figures, and not every company provides an adjusted EPS figure. Brokers who contribute to consensus forecasts may overlay their own standards for adjusting and harmonizing EPS figures. Thus, care must be taken to ensure that P/E multiples are calculated on the basis of, or applied to, comparable EPS figures. Cash-flow-based valuation approaches sidestep the issues arising from the lack of comparability between pharmaceutical companies that hamper peer group analyses. Such approaches capture the medium-term profit outlook with a relatively high degree of reliability because the visibility on peak sales and patent loss of key profit drivers is fairly high. Inevitably, they require a judgment call on the quality of a company’s pipeline and its ability to innovate. By 40


Valuation Financial of Pharmaceutical Statement Analysis Firms

their very nature, methodologies based on discounted cash flow (DCF) tend to be highly sensitive to assumptions about the terminal growth rate, which include implicit assumptions about the long-term growth of cash flows from ancillary activities as well as the ability of the company’s branded prescription drug business to continually reinvent itself following patent expirations. In conclusion, DCF-based valuation approaches may yield the best approximation to “fair” value, provided that investors and analysts have a high level of confidence in the company-specific forecasts that serve as inputs. Reverse-DCF analyses that gauge the terminal growth rate implied in a stock’s current share price reveal the market’s view of a company’s long-term innovative capacity. Investors may consider the likelihood that a company will meet or exceed market expectations in the long term as well as the plausibility of differences in long-term growth prospects priced into the shares of pharmaceutical companies. Owing to drug firms’ relatively high dependence on individual drugs and the operating leverage associated with large assets, project-specific NPV models are popular with analysts and investors. It is important, however, to be aware of their limitations, especially concerning the uncertainty of forecasting drugs—at an early stage of development, a drug’s peak sales potential and its effective period of exclusivity may be very difficult to predict—and the inherent challenge of allocating the substantial fixed costs that continue to characterize drug firms even in this era of in-licensing and outsourcing to individual projects. Bottom-up valuation approaches that consist of adding up project NPVs may constitute a viable approach for biotechnology firms with limited infrastructure and few drugs in development. However, in the case of large pharmaceutical companies that rely on the continued success of marketed products as well as permanent rejuvenation of their portfolios in order to maintain their existing infrastructure, NPV models are better suited to sensitivity analyses. For example, NPV models of potential blockbuster drug candidates or key marketed products that take into account only the variable cost associated with these assets may provide a rough idea of the implications of clinical success or failure—or of the potential withdrawal of a product from the market owing to safety issues—for a stock’s valuation and the business decisions a company might be forced to make in the event of failure. Table 4 reports on the suitability of various valuation methodologies for the industry.


41


Table 4. Suitability of Standard Valuation Methodologies for the Pharmaceutical Industry Method EV/sales

EV/EBITDA

Suitability Low

Medium

P/E

High

DCF

High

Reverse-DCF

High

NPV models

Medium

Commentary This multiple does not adequately capture differences in margins that may arise as a result of drug profit sharing between companies. This multiple does not capture substantial differences “below the line” (e.g., relating to associate income and variations in tax rates). This is the most widely used multiple; because individual companies’ prospects may differ materially, it requires the user to consider the premium or discount merited by a stock relative to its peer group. This approach captures the analyst’s specific views on a company’s prospects and thus sidesteps issues that may arise from the industry’s heterogeneity and resulting lack of comparability across peer groups. This methodology gauges the terminal growth rate implied in a stock’s current share price and thus reveals the market’s view of a company’s long-term innovative capacity. The use of drug- or project-specific NPV models should be reserved for sensitivity analyses. This approach does not lend itself to bottom-up valuation, owing to the difficulties of allocating fixed costs to a large number of individual projects.

PORTFOLIO CONSIDERATIONS Because a pharmaceutical company’s operating performance is determined largely by the success of its drugs and drug candidates and may often be influenced only marginally by general industry trends, stock picking is of the utmost importance. The stocks of companies that generate top-line surprises and positive clinical news flow usually tend to outperform the industry. Even so, it is advisable to spread pharmaceutical investments over a number of stocks so as to minimize idiosyncratic risk. Even the best-managed and best-positioned company is not immune to the risk of extreme events that are impossible to foresee or hedge. Such events can entail extremely large cash outflows and can severely dent a company’s profitability for extended periods. 42



Examples include, but are not limited to, supply disruptions, the emergence of safety concerns in relation to a large product, the failure of high-profile pipeline compounds, product liability suits, and litigation of marketing practices. True “hedges” are rare, in the sense that it is often impossible to identify companies that would benefit directly and proportionally from a competitor’s setbacks, suggesting that diversification is often the only viable strategy to reduce stock-specific risks. It is worth noting that the heterogeneity of the industry, where fundamentals are concerned, also extends to the characteristics of pharmaceutical company shares. The listing currencies of the largest stocks include the USD, EUR, GBP, CHF, DKK, and JPY, which further complicates hedging decisions. Size differences can also be extreme. For example, the market capitalization of each of the five largest companies exceeds $100 billion, but there is an abundance of small and midsize companies valued at mere billions, hundreds of millions, and even millions of dollars. This fragmentation reflects the high barriers to entry into any one segment of the pharmaceutical industry; many biotechnology players are one-product companies that own the intellectual property to a specific molecule, which prevents their larger peers from developing the same molecule and putting them out of business. In light of the pronounced impact that such product-related news flow as clinical data and regulatory approval decisions may have on a company’s future profitability and investor sentiment, an understanding of key upcoming catalysts is a prerequisite for well-informed investment decisions. Uncertainty with respect to both the timing and the outcome of any product-related news flow is one of the main challenges of investing in pharmaceutical firms. Companies often provide a rough guide to the expected timing of news flow; for example, they might communicate that headline results from a phase III trial are expected in the second quarter or second half of the year. It is rarely possible to specify the month, let alone the day, when key news flow is expected. Consequently, it is often virtually impossible for investors to avoid exposure to news-flow risk when taking positions in pharmaceutical stocks. The likelihood of success may be even more difficult to predict than the timing. Many analysts and investors avidly study the publicly available clinical data, trial design, and historical precedents to gain a better understanding of the potential pitfalls; however, there are no crystal balls to predict the outcome of any one event. Consequently, short-term trading strategies are often fraught with risks. Longerterm investment decisions should ideally be based on a high level of comfort around the totality of the news flow to be expected over a time horizon of 12–18 months. There are no compelling reasons to shun stocks that are devoid of large individual catalysts. In fact, stocks with a large number of expected news-flow items that may not be transformational in and of themselves but have the potential to drive longerterm consensus earnings upgrades and to positively affect sentiment may offer an attractive risk–reward profile. ©2013 CFA INSTITUTE

43


However, caution is warranted in the case of upcoming catalysts related to projects that are advertised as “free options” by the sell side. Although many analysts may prudently refrain from including the potential upside in their forecasts, it is often included “in the whisper.” By the time a sufficient number of investors have taken positions to benefit from the potential upside, enough of it is priced in to trigger a negative share price reaction in the event of failure, whereas the rally to be expected in case of success may be limited. A particularity of the pharmaceutical industry is that news flow without the potential to influence near-term forecasts may still trigger significant share price reactions. For example, a stock may rally on the announcement of positive phase III data for a potential blockbuster drug candidate, even though the achievement of peak sales may still be more than five years away and the drug may not break even until one to two years after the launch. The relatively modest impact of industry trends on companies’ operating performance largely precludes trading strategies centered on quarterly earnings releases. A firm’s operating margins primarily reflect the growth trajectories or decay curves of its most profitable assets (i.e., their stage in the product life cycle) and the timing of major expense items, which is largely discretionary. With the industry’s rising exposure to emerging markets, where business may be more tender driven and conducted in a large number of different currencies (many successful drugs are sold in more than 80 countries), even the top line is increasingly difficult to predict. (Note that US sales of key drugs can usually be forecasted fairly accurately owing to the wide availability of weekly and monthly prescription data.) An earnings “beat” or “miss” is thus not necessarily indicative of a change in company fundamentals; it may simply reflect fluctuations from one quarter to another. Rarely is there a significant “read-across” from one company’s set of results to those of its peers reporting later. Pipeline news flow represents a risk with respect to earnings releases to the extent that the release contains a pipeline update and may, on balance, be slightly more negative than positive. The reason is that the results of pivotal trials are seldom communicated in the context of earnings releases. Headline results are usually announced by means of a press release as soon as they become available, whereas the presentation of details must often await a medical congress that may take place many months later. Medical congress rules typically prevent company management from divulging any further information ahead of the event. At best, a company may announce positive internal decisions about its pipeline, such as phase III transitions of drug candidates. An important corollary is the possibility that the termination of development programs or other negative news may be announced in earnings releases. Although divergent share price performance may be more common and more pronounced in the pharmaceutical industry than in other industries for the reasons mentioned, there have been extended periods when stocks have moved as a group, 44



often linked to industry rotations coupled with a news-flow void. Owing to their relatively pronounced lack of cyclicality, pharmaceutical stocks are perceived as defensive in the classical sense (i.e., as offering relatively low beta). (As discussed previously, pipeline failures and various issues that may arise with respect to a company’s current operations have the potential to send individual stocks plummeting. In the latter part of the last decade, many of the majors suffered a string of setbacks, prompting some investors to exclaim, “This industry is not defensive!”) Because cash flows are relatively stable and predictable during the exclusivity periods of key products, many companies offer generous dividend payout ratios and have adopted a progressive dividend policy. Therefore, rotations into the pharmaceutical industry may at times be driven by considerations relating to dividend yields. Although the rewards for successful stock picking may vary as a function of the market environment, it appears intuitive that a lack of due diligence with respect to any pharmaceutical investment invariably entails severe risks to the performance of a portfolio.


45

INDUSTRY RESOURCES REGULATORY AGENCIES US Food and Drug Administration 10903 New Hampshire Avenue Silver Spring, MD 20993 (888) 463-6332 www.fda.gov European Medicines Agency 7 Westferry Circus Canary Wharf London E14 4HB United Kingdom 44 (0)20 7418 8400 www.ema.europa.eu/ema Ministry of Health, Labour and Welfare 1-2-2 Kasumigaseki Chiyoda-ku Tokyo, 100-8916 Japan 03-5253-1111 www.mhlw.go.jp/english

OTHER RESOURCES FDA ORANGE BOOK A useful tool to track drugs approved in the United States, this resource is searchable by brand name, active ingredient, and various other parameters. It also contains such information as the date of approval, the approved dosage forms, and patents and regulatory exclusivities. www.accessdata.fda.gov/scripts/cder/ob/default.cfm

US PATENT AND TRADEMARK OFFICE The federal agency for granting US patents and registering trademarks, the USPTO provides a searchable database of patents. www.uspto.gov 46


Industry Resources

FDA ADVISORY COMMITTEES Announces upcoming panel meetings and provides briefing materials. www.fda.gov/AdvisoryCommittees/WhatsNew/default.htm

FDA WARNING LETTERS Communications to drug manufacturers and other stakeholders, which specify deficiencies in production systems that need to be addressed. www.fda.gov/ICECI/EnforcementActions/WarningLetters/default.htm

FDA PARAGRAPH IV PATENT CERTIFICATIONS Provides a list of all the drugs that have received a paragraph IV patent challenge from a generics company. www.fda.gov/Drugs/DevelopmentApprovalProcess/ HowDrugsareDevelopedandApproved/ApprovalApplications/ AbbreviatedNewDrugApplicationANDAGenerics/ucm047676.htm

EMA FIND MEDICINES Allows users to view approved as well as submitted medicines; following approval, a European Public Assessment Report detailing the EMA’s views on the drug’s risks and benefits is also made available in this section of the website. A number of e-mail subscription services are available on agencies’ websites for those who wish to receive regulatory news directly. www.ema.europa.eu/ema/index.jsp?curl=pages/includes/medicines/ medicines_landing_page.jsp&mid=

CLINICAL TRIALS A widely used resource on completed and ongoing clinical trials. It is searchable by drug, sponsor, condition, and various other parameters and includes details on each study, such as the start date and expected completion date, the targeted patient enrollment, the design (including endpoints), patient inclusion and exclusion criteria, and the geographic location of the study centers. http://clinicaltrials.gov

PUBMED A medical search tool that returns abstracts and articles (some of which are free) on drugs, therapeutic classes, and medical conditions. www.ncbi.nlm.nih.gov/pubmed ©2013 CFA INSTITUTE

47


BLOOMBERG Users will find a wealth of information, including prescription data and various analyses, by typing in industry functions (e.g., BI PHRMG for the major pharmaceutical companies).

MAJOR MEDICAL CONFERENCES Tens of thousands of physicians and other stakeholders convene annually to stay current on medical developments in their particular field of practice. These events are often chosen as a venue for the presentation of important new clinical data on individual drugs. Registration is fee based and is open to the public; abstracts are typically made available free of charge on the organizations’ websites. American Society of Clinical Oncology www.asco.org European Society of Medical Oncology www.esmo.org San Antonio Breast Cancer Symposium www.sabcs.org American Society of Hematology www.hematology.org American College of Cardiology www.cardiosource.org/acc American Heart Association www.heart.org/HEARTORG European Society of Cardiology www.escardio.org

48


CFA INSTITUTE INDUSTRY GUIDES

THE PHARMACEUTICAL INDUSTRY ISBN 978-0-938367-81-9

9 780938 367819

90000

CFA - The Pharmaceutical Industry

Recommend Documents