Main

It is estimated that 6 million people die each year from AIDS, tuberculosis (TB) and malaria. The social benefit of creating vaccines to prevent these diseases is tremendous; unfortunately, however, the private benefit to companies for development in these areas does not mirror this. Developing countries do not possess the capacity to pay for drug development, which current estimates put at between US$750 million and $1 billion1. In order to meet the Millennium Development Goals (see Glossary box) associated with preventing and treating diseases such as HIV, malaria and TB, the Global Fund estimates that annual international resources put into organizations combating the spread of these diseases should be in the range of $15 billion in 20072. To overcome this last barrier, the Bill and Melinda Gates Foundation, the Global Alliance for Vaccines and Immunizations (GAVI), the Global Alliance for TB Drug Development and many others have made major strides in committing money to vaccinations for children. The most recently established of these international organizations, the International Finance Facility for Immunization (IFFIm), was announced recently by Gordon Brown, UK Chancellor to the Treasury, and consists of a multi-country effort to raise money for GAVI. Initial pledges by the UK, France, Italy, Spain and Sweden total more than $4 billion over a period of 10 years3.

So although financial resources are currently available to address the issue, the money still needs to be distributed in a manner that ensures the rapid and safe development of effective vaccines through sustained research and development. Traditionally, methods to provide incentives to encourage research for neglected drugs have fallen into two main types: 'push' and 'pull' methods. This Perspective article examines the shortcomings of traditional push and pull methods in the context of corporate project valuation techniques. We propose a new incentive mechanism that combines push and pull elements and is based loosely on the principles of call options in equity markets, and which we therefore term the Call Options for Vaccines (COV) model. In a typical call option, an investor can purchase the right to buy a share of stock at a later date for a fixed price. The idea behind an option is that an investor can pay a premium now for the potential to profit later. However, the potential to profit later is not guaranteed, and so risk is involved in the payment of the premium. The seller of the call option also undertakes some risk, because any potential profit will come at the seller's expense.

Here we aim to demonstrate that an initial investment in exchange for a discounted future price could provide the proper incentives to make neglected disease research attractive and profitable. Although our model focuses on vaccine research, some of the principles could, in theory, be applied for other drugs targeted toward similarly neglected diseases. For the sake of simplicity, we limit the scope of this discussion to vaccine development.

An incentive problem

As stated above, the main obstacle to developing vaccines for neglected diseases is the alignment of corporate and social incentives. Current and past solutions addressing this issue have fallen into two broad categories: push and pull mechanisms. The advantages and disadvantages of each method are briefly described below.

Push mechanisms. Push mechanisms consist of giving money to academic, or government, scientists to facilitate scientific research on a particular problem. By reducing the costs of inputs and advancing the state of basic science, push mechanisms aim to make vaccine development cheaper by reducing the costs of research and development to a particular institution or laboratory by providing grants4. Other examples include tax credits from governments, fast-track regulatory review, partnerships between public and private entities, and developing clinical trial infrastructure5. One criticism of push mechanisms in general is that academic and government scientists might be unlikely to engage in the necessary activities to take a research project from bench to bedside. This is less of a problem in an organization such as the Global Alliance for TB Drug Development, which engages in the active management of all phases of the development process to ensure that the priority is effective vaccine development. However, in most other basic research labs, there is little impetus and/or experience to move into translational or applied research6. Money is spent with no guarantee of future benefits. Similarly, if programme directors are incapable of judging which avenues of research hold the most promise, poor project selection could hinder the success of direct funding programmes7. Poor allocation of funds could result if lobbying interferes with funding decisions, or governments value products differently than consumers7.

Pull mechanisms. The existence of a large and profitable market is central to the concept of pull mechanisms, the other major type of research incentive (see Box 1 for an example). Pull mechanisms seek to induce private companies to develop vaccines by promising a reward if the goal is achieved. One of the more attractive pull mechanisms is that of advance purchase commitments8. These are based on the premise that companies are more likely to invest in development if they can be guaranteed profitable returns. Developed countries possess the money to pay for drugs that provide a therapeutic benefit at a rate that protects profits, whereas undeveloped countries often do not. By agreeing beforehand to pay a specified amount for a vaccine, if developed, governments or other large purchasers can create strong financial incentives to develop a viable product. No money is paid to the company if the drug is not developed, therefore minimizing the financial risk to the purchaser, but increasing the downside for the developer. This unequal distribution of risk is one of the greatest limitations to pull mechanisms.

There are also a number of design issues that must be considered when developing purchase commitments. To begin with, the commitment will not be effective unless the sponsor stays committed. Governments might change funding priorities after several years9. Similarly, countries might attempt to avoid adherence to patents or to the terms of the commitment. Also, an advance purchase commitment requires advance specification of the vaccine or drug that will qualify for purchase. The vaccine might work differently in certain populations, or provide risk:benefit ratios acceptable in a severely stricken country that might be unacceptable in more developed nations10. Regulatory approval could be used as the overriding qualification for purchase, but will not necessarily determine efficacy requirements. Finally, the value of the purchase commitment must be large enough to stimulate development, but still smaller than its social value (the maximum gain from the vaccine for society).

An approach that tries to address the limitations outlined above has been developed in a working group assembled by the Center for Global Development (CGD). The resultant report11 outlined a plan to create a guaranteed market for vaccines. In this approach, CGD guarantee a certain price, subsidized by co-payments from national purchasers, for a specified number of units of a qualifying vaccine (whether or not a vaccine qualifies is decided by an independent committee). Although the price is guaranteed for any vaccine, the quantity is not, as countries are not obligated to purchase developed vaccines. This guarantee of price (but not quantity) removes the need to explicitly specify the conditions of an acceptable vaccine, but still maintains a credible commitment to purchase. The CGD estimates that 200 million doses of vaccine could be bought at a price of $15 per dose, resulting in the creation of a $3-billion market. This estimate underscores the implicit assumption that the key to successful development is to translate social benefits into corporate profits, as the next section explains. Pull mechanisms hinge on the decision of companies to undertake R&D projects, given an expected future cash flow. To understand these decisions, it is worth reviewing the basics of corporate finance and project valuation.

Investment decisions and project valuation

The valuation of biotechnology and pharmaceutical projects has gained much attention in the past decade. Historically, decisions about corporate projects have been undertaken using a discounted cash flow model (DCF, see Box 2), but Stewart12,13 proposes that the risk-adjusted net present value (NPV; see Box 2) should be used when valuing biotechnology projects. The technique is similar to that of traditional DCF; however, all cash flows would be discounted by a factor that appropriately reflects the chances of success and failure for the project. Yet this does nothing to accurately characterize the decision structure needed for pharmaceutical R&D because it does not consider the value of the abandonment option — that firms can cut their losses and do not need to continue pouring money into a losing project14,15.

Dixit et al.16 examined the optimal investment rule that firms should use in deciding whether or not to invest in a project with fluctuating payoffs. They describe the scope to invest as similar to that of a call option and suggest that a firm will invest only when the returns exceed the costs — a practice similar to the exercise of a call option that is well above its strike price. Damodoran15 also suggests that R&D can be thought of as a call option, although he characterizes the entire amount spent on R&D as the cost of the option, and the resulting product the payoff. Therefore, an alternative to the DCF method described above is that of real options valuation (ROV).

Real options are essentially an opportunity to invest in a product as a result of development or research in that product17. It is widely accepted that the development of a drug by a corporation can be thought of as a series of decisions about whether to continue or discard a particular project. Having an opportunity to invest means that firms have the potential for a profitable return on investment as well as scope to minimize loss by abandoning a project.

Schwartz18 used ROV analysis to determine the real options value of a pharmaceutical project, along with the value of its abandonment option. Rogers14 also argues for the use of real options valuation (ROV) analysis for decision-making in the pharmaceutical industry (Fig. 1). ROV allows companies to quantify the value that is added to a project by having the potential to invest, and the potential to abandon, the project. Furthermore, ROV could be applied to the evaluation of a portfolio of projects19 or to rank projects within a portfolio20. A typical NPV calculation might yield a different investment decision than one undertaken with an ROV analysis. Real options demonstrate that projects can have greater value than previously determined under DCF analysis because the distribution of expected project returns is shifted more towards the positive side21. This would be primarily due to the fact that a negative NPV project would treat all investment decisions as determined at the present time; by contrast, an options approach would leave these decisions open and flexible, and would therefore add value to the overall project.

Figure 1: Decision tree outlining the pre-market authorization process, including transition probabilities of success (P).
figure 1

Fig. 1 illustrates the decision process that might currently be employed in an ROV analysis and could potentially be used in a Call Options for Vaccines analysis to determine the appropriate amount and timing of the incentive payment. Probabilities taken from Struck36.

Therefore the idea of financial options concepts being applied to pharmaceutical practice is not a new one. However, previous analyses were undertaken from the perspective of the pharmaceutical companies and considered as a method to determine whether to continue or end R&D for a particular drug. It is possible to extend this concept to the traditional push and pull programmes to create a new mechanism to stimulate vaccines research for neglected diseases.

The COV model

As discussed previously, push and pull mechanisms both have potential drawbacks and face barriers to their successful implementation that the COV model, which combines elements of both push and pull mechanisms, might overcome. The COV model is not a completely novel concept; the potential merits of combined programmes have been discussed previously4,6. One of the best examples of such a combination is the US Orphan Drug Act, which offers companies 7 years of market exclusivity along with tax credits for drugs that target diseases afflicting less than 200,000 people22. Additionally, some evidence is available to support the idea of combining push and pull mechanisms. Hsu and Schwartz23 created an ROV model to evaluate the capacity of different incentive mechanisms (push, pull and hybrid) to stimulate the development of effective treatments. They concluded that a hybrid mechanism of payments to the developer to help cover R&D costs combined with purchase commitments offered the greatest hope. This combined method spurs additional research as the initial payment increases. It also addresses two conflicting problems. First, the cost of the firm's early-stage R&D can be covered (at least in part) by the early payment. This drives more initial R&D but gives no incentive to create a final product. However, the purchase commitment solves this second problem by providing a market and profits only for working vaccines. Hsu and Schwartz discuss this in their analysis and recommendations, but give few suggestions about how such a scheme could be implemented.

In our COV model, a potential purchaser would buy the right (during development) to purchase a specified amount of the vaccine at a later date, for a specified price. If the vaccine never makes it to market, the purchaser only pays a premium equal to the cost of the initial 'option' contract. A fair valuation of an option will make the current value of the premium equal to the expected future profit from holding the option — the purchaser is therefore protected from the downside risk of the cost of development, and the developing company is given an additional, earlier incentive to continue development. The greatest challenge is to persuade companies to invest in a vaccines market with low returns. Conventional thinking suggests that if it is possible to increase returns, at the very least giving the project a positive NPV that meets a predetermined threshold, then profit-maximizing companies will always invest.

There are two ways to increase the NPV of a project involving large development costs: increase expected future cash flows or lower current costs. So whereas pull mechanisms seek to increase future payouts, and push mechanisms help to lower current costs, our strategy does both. Our method hopes to incorporate financial incentives, while assuming a sense of corporate responsibility. Few companies will undertake projects in which the NPV is negative; however, it is naive to think that the NPV threshold for neglected vaccines can be as high as those for projects in developed nations. Therefore any viable solution must incorporate incentives to generate appropriate financial returns to developers, while relying on the goodwill of these companies to maintain a lower profit threshold for neglected vaccine projects. This assumption of corporate responsibility is supported by several examples of corporations donating drugs to treat neglected diseases24,25.

Once a company brings a drug to a specific phase of testing — for instance, Phase I clinical trials — a potential purchaser could be allowed to examine all of the data on the vaccine and make an independent assessment of its potential. Ideally, this would be an international non-governmental organization or charitable foundation, with adequate funding to make several investment decisions and create a credible investment commitment, such as The Global Fund or GAVI. If the purchaser believed the drug to be a good investment, it would pay an agreed upon amount (the methodology for determining this will be discussed later) and in exchange would obtain the contractual right to buy a certain amount of the vaccine at a reduced price if it made it to market. If the vaccine was to run into problems during clinical trials and did not receive marketing approval, then the development company would retain the initial investment and the purchaser would have neither an obligation to buy nor derive a benefit from the investment. However, any contract negotiated would need provisions ensuring access and ownership of all vaccines developed from the initial, funded line of research. If one avenue proved promising, only to spawn a successful vaccine from a related mechanism, the purchaser would have an equal right to the new vaccine, as it came from the intellectual property of the funded research. Likewise, if a company were to acquire the development project and wanted to stop development, the contract could call for financial penalties.

The final price, for the holder of the options contract, could be lower in this method than in conventional pull mechanisms because the company would have had the additional incentive of the earlier investment to ensure participation. The international purchaser could then distribute the vaccine to developing countries on the basis of their capacity to pay. This is similar to tiered pricing6,26, also referred to as Ramsey pricing27. A market price and a discounted price (for the holder of the options contract) could be fixed for the vaccine, negating the effects of parallel importing, while ensuring access to essential medicines in the countries that need them the most. If the contractually specified amount of vaccine is purchased, but more is demanded, the purchaser could then continue to buy vaccines at the higher full-market price. All countries would be eligible for the purchase of vaccines through the non-governmental organization, although the price for each one could be negotiated and kept confidential. The payment for the vaccine by the purchasing country would offset the cost of the investment for the NGO and ensure that the country does not seek a vaccine that is not appropriate for its needs11 — that is, not effective for their population or particular strain of disease. As long as the pricing is tied to the capacity to pay, equity and efficiency objectives would seem to be satisfied.

The evaluation of the proposed vaccines will be of crucial importance to making the system work efficiently. Financiers, economists and scientists would all be needed to determine whether the options contract provided good value for money and was worth investing in. Full disclosure by the company of all test results (both from animal models and clinical trials) would be necessary, in a manner similar to that required for licensing approval. Reluctance to disclose such proprietary information should be overcome by the desire to obtain preliminary funding; confidentiality would of course be essential. The group to evaluate the drugs could be an extension of the purchasing organization.

If the Global Fund were to be involved in such endeavours, a separate committee would need to be set up with special expertise in this arena. According to the Global Fund framework, 50% of the committed funds are set aside to purchase drugs and other commodities28. The framework also emphasizes that the fund is a financial instrument, not an implementation entity, and so this additional committee should use the resources of the fund to make investment decisions to fulfil the goal of purchasing medicines. The purchase of vaccines for the benefit of participating countries would help solve some of the logistical problems faced by countries with under-funded and over-stretched healthcare systems, by eliminating regional approaches to disease cure and prevention. The current approach to the distribution of funds requires a great deal of accountability and coordination in many countries that simply struggle to balance all of the available donor agencies29. Effective vaccines could negate some of these problems by forming a centralized solution, although problems in effective distribution of the vaccine would still remain. Box 3 and Fig. 2 explain how the COV model could work in practice.

Figure 2: Binomial model representing investment with two payoff states.
figure 2

In the higher state, which has a 95% chance of success, the cash flow will be $5,000. In the lower state (symbolizing a failure in development), the cash flow is $0. If the cost of undertaking the project is $5,000, then the developer will only break even if there is a 100% chance of obtaining the higher cash flow. Therefore in this case, the developer will choose not to take the project, because the expected cash flow is equal to the probability of each outcome multiplied by the value of each outcome: 0.05 × 0 + 0.95 × 5,000 = $4,750. To make the expected cash flow high enough to cover the $5,000 development cost, the higher-state payoff could be increased to $5,263. This would give an expected cash flow of 0.05 × 0 + 0.95 × 5,263 = $5,000. Alternatively, the costs to the developer could be reduced by an initial payment of $250. If the cost-reducing payment were made before undertaking the project, the original expected cash flow of the project would be enough to cover the total costs of $4,750. If a positive NPV were desired (instead of an NPV of $0), payments on either end could be increased slightly. Overall, there is a saving of $13 ($250 versus $263) if the payment is made earlier and this stems directly from the fact that the investor is taking on some of the risk in the project development. NPV, net present value.

Practical challenges to the COV model

Despite the many benefits of the COV model (Box 4), there would be several challenges in trying to put the model into practice. In a typical option valuation, a number of pieces of information are used to find the value of the call option15: current value of the underlying asset; time to expiration on the option; variance in value of the underlying asset; dividends paid on the underlying asset; strike price of the option; and the riskless interest rate. Not all of these have direct pharmaceutical corollaries, but many similarities can be seen. To better understand this, assume the hypothetical development of vaccine X. A proper valuation of vaccine X for the COV model would require knowledge of the volatility of development (variance in value of the underlying asset) and the length of time development might take, from the bench to the bedside (time to expiration). No dividends are paid on the asset, because a vaccine can only produce cash upon successful market authorization. The strike price for the vaccine is the cost of vaccine X after market authorization to the holder of the contract. The riskless interest rate is the opportunity cost of investing money in development of the vaccine, but an appropriate discount rate should reflect the costs of other missed opportunities, as it does when using the weighted average cost of capital (WACC; see Box 2). The challenge exists in trying to find a proper method of calculating the monetary payment that accurately reflects the additional risk assumed by the investor and compensates them accordingly with a reduced vaccine price. As discussed above, ROV seems to provide an excellent method of calculating the overall expected value of a given project and could be used to determine how much additional money would be needed at any given stage to make the project attractive to finance executives. Figure 1 illustrates the decision process that could be used in ROV to determine the appropriate amount and timing of the incentive payment.

A fair assessment of this payment, however, would require knowledge of the costs and risks associated with each stage of development considered during the investment. A more accurate figure for specific vaccines would be dependent on the type developed, any prior history of failures or successes in animal models as well as the track record of vaccines in similar therapeutic groups, or with similar biochemical mechanisms. The incentive payment (that is, the price of the option) must also be weighed against the savings on the final price of the vaccine for the purchasing organization. This price must be sufficiently below market to encourage initial investment by a purchaser, yet still high enough to cover the marginal cost of production, with additional incentives to encourage the company to develop the drug to full production.

This last point brings up a challenging question: what might keep the pharmaceutical company from only developing far enough to gain the incentive payment? If the costs of initial research are minor compared with the value of the options contract, companies might be motivated to compete for the contract, turn a profit by gaining the contract and never follow through. However, this can be avoided through several different design methods. If the contract is undertaken far enough along in the development phase, there will be less incentive to undertake research for the sake of winning the contract. A careful balance of the amount of the option cost and the risk that must be incurred to obtain it must be found. If this is combined with competition from other pharmaceutical firms to develop similar drugs, then the advantage of simply investing up to the contract point will diminish, as not all companies' research costs will be recouped by the contract money. Similarly, if a single purchaser is in charge of such a system, gaming the system in the above manner will cause the purchaser to lose faith in the developer, which could potentially lead to denial of contracts in the future. The importance of this 'good faith' relationship should not be underestimated — top drug makers have seen a decline in prescriptions and a loss in revenues due to rising mistrust among consumers30. Bad publicity is a powerful incentive, and the transparency necessitated by the COV model, with regular progress reports to COV purchasers, could improve the overall disposition of the industry.

A very real downside to this approach is that the cost of the option contract payment would also need to include the time value of the money paid. Purchasers would therefore lose the investment potential of that sum. Similarly, bad investment decisions could quickly lead to a large loss of money with no real benefits. This is similar to one of the critiques of push mechanisms — that project managers might prove incapable of deciding on the most promising research plan. This could be rectified by the formation of an independent body to assess which ideas/vaccines to fund. Such an approach relies heavily on the full disclosure of all relevant documents, which might not be achievable. Given this, great emphasis should be placed on the skill and qualifications of the drug evaluators, as well as on the free exchange of information between the companies and potential purchasers. International purchasers would need to hire people with specific skills in financing, project valuation and/or ROV assessment. However, it is reasonable to think that, if information is shared appropriately, the skills of the purchasers could closely match those of the vaccine developers in this area.

One could also envision a situation in which additional regulatory trials mandated for marketing approval would increase the total cost of development. This possibility might deter some companies from engaging in a contract with a predetermined price. This could be solved in one of two ways: first, by setting aside a special emergency fund (paid for by the purchaser) that could only be accessed in the specific case in which additional government trials are required; and second, the probability and cost of this scenario could simply be built into the model, which might slightly increase the premium for the call option.

Another criticism of this approach is that the number of projects could be limited, which in turn might narrow the number of potential vaccines in development. However, it could be argued that exactly the opposite would occur. If companies were contractually obligated to sell their first vaccine at a lower price, they might also intensely work on more effective vaccines in parallel, with the hopes that purchasers would have to buy the better vaccine at full price due to public pressure. They would still have an incentive to make a good-faith effort to bring the original to market, because if they did not the global purchaser would see this as a breach of contract and be less likely to invest in them in the future. Some might also point out that this competitive environment could hinder the sharing of ideas, particularly when compared with the free exchange of data and information seen in academic and government-funded laboratories. Although this is true, it is no different than the current business climate of for-profit scientific research. So although this model would not help academic laboratories, it should spur on private development, and that is its intent. If it was felt that data sharing among companies was absolutely crucial, particularly for a product funded in part by international organizations, then the potential loss in sales from generic manufacturing would need to be compensated for by an increased premium price for the drug. The key is to maintain a financial incentive for the developer, and so the release of proprietary information would and should come at a high cost.

Conclusions

Debate continues about the most efficient method of spurring research and development for neglected diseases for which normal market incentives do not exist. Traditional push and pull measures have many shortcomings and do not seem to offer viable options for accomplishing the goal of developing new vaccines and drugs for diseases and patient populations that otherwise are unattractive for industry to pursue. The hybrid COV strategy combines both push and pull mechanisms by offering potential purchasers the right to purchase a drug in the future at a discounted price in exchange for a payment now, and is essentially a risk-sharing scheme in which the initial investment of money is relied upon to signal a good-faith interest in future drugs that meet therapeutic needs. Therefore, the necessity to specify the requirements of a vaccine — one of the biggest obstacles to purchase commitments — is avoided. The success of this approach will depend largely on the evaluation of the possible drugs for investment, along with the amount of payment incentive and the reduction in price afforded to the purchaser at market launch.

Further research on this subject is necessary to determine the optimal mix of these conditions to stimulate the maximum amount of development in the most efficient manner. Specifically, the valuation methodology for the intended investment projects would need to be developed, and the most challenging part of developing that methodology would be to ensure that the predetermined strike price did not undermine incentives for pharmaceutical companies to fully develop the vaccines. Clearly, the higher the initial payment is, then the greater the incentive for producers and the greater the potential for waste by the purchaser. A delicate balance must be struck, one that accounts for the risks and benefits faced by each stakeholder. Perhaps the world of finance might offer capable assessments and valuations in this area. Once the underlying mathematical valuation method has been developed, it should be tested with real data from recent development projects to determine how accurately it can predict option values and strike prices that actively encourage development. The COV model is an idea that complements and challenges conventional practice, but we acknowledge that it needs more development before it is practically implemented.