A Token Design for Decentralized Insurance on the Blockchain
March 6, 2020
Alexander Braun and Niklas Haeusle
Imagine you are a smallholder paddy field farmer. Your crops are vulnerable to natural disasters such as floods or droughts, even more so in the face of climate change. Since your entire existence depends on the integrity of your crops, you would like to have insurance in place that reimburses your losses in case your harvest gets destroyed. However, what if the insurance premium is unaffordable? What if coverage is not being offered, because your small premium cannot cover the costs of traditional insurance companies? What if it takes too long to receive a compensation payment after the loss event occurred?
In developed economies, these problems arise less frequently. In these places, insurance penetration tends to be high, monopolies are rare, and coverage for a wide range of perils is available. Furthermore, customers are protected by regulations and may have other assets that help them bridge the delay between the occurrence of the loss and the indemnification by their insurer. The situation is different in emerging economies, however, where the insurance sector is often not properly developed or regulated. This means that rent-seeking monopolistic enterprises can take advantage of the insured or, even worse, no insurance solution is available at all. When Hurricane Maria struck Puerto Rico in September 2017, for example, most citizens were uncovered. Of those that had insurance, 11,000 were still waiting for their claims to be settled more than one year after the event.
Decentralized Insurance Platforms on the blockchain can help to mitigate these problems. Blockchains are a form of distributed ledger technology that enables record keeping by a network of nodes rather than a trusted intermediary. Blockchain databases are highly transparent, tamper-proof and openly accessible. A number of experts have already described this technology as the foundation of the world’s next-generation financial services infrastructure. The latest public blockchain systems such as Ethereum can run smart contracts. The latter are pieces of code that execute if-then-else instructions. Thus, they are well-suited for the development of a Decentralized Insurance Protocol (DIP), i.e., a digital mechanism that allows for the decomposition of the classical insurance value chain. Based on a DIP, insurance risks can be priced, transferred, and managed by a group of independent service providers without having to rely on a central intermediary or a traditional insurance company balance sheet.
Decentralized insurance relies on a high degree of automation, e.g., in sales and underwriting and therefore makes policies affordable, even for very low sums insured. In addition, parametric triggers based on physical variables such as rainfall or wind speed ensure a rapid payout of claims: it usually takes one day instead of several months to reimburse the customers. The model ultimately becomes viable, since tail risk, such as an accumulation of claims due to a large-scale natural disaster, can be tokenized and transferred to the capital markets.
Evidently, this new institutional arrangement of the “decentralized insurance organization” may provide many benefits for customers. However, a number of critical problems have to be solved before practical implementation becomes feasible. One of the most pressing questions is the incentivization of “workers” in the decentralized setting, that is, those individuals or entities that are responsible for the production of the product or service. Since decentralized insurance works on a pure transactional basis, classical economic incentivization through hierarchies and managerial oversight cannot be applied.
In a recent research project, we addressed exactly this problem. We closely cooperated with the Decentralized Insurance Foundation, a Swiss non-profit organization to develop a feasible incentivization mechanism for their DIP on the Ethereum blockchain. The result of this collaboration is a new token design with strong foundations in microeconomic theory that also exhibits a high degree of practical feasibility. At the center are cryptographic tokens, i.e., digital assets issued to the anonymous blockchain network participants who do not trust each other. Due to the common asset and a shared interest in its value, their interests align along the goals of the project. A main feature of our token design is staking, a digital form of collateralization, which requires network participants to deposit tokens into a smart contract or bound wallet. In case of an insufficient effort and, in turn, a faulty result, the tokens are redistributed to the customers to compensate them for their losses. Our results show that this concept can effectively incentivize workers in the decentralized insurance organization and therefore prevent malfeasance that could endanger the product quality and ultimately the whole market.
Furthermore, we determine the optimal stake to be put up by workers who deliver a component or service of the digital insurance product in the DIP network. We are able to show that the welfare-maximizing stake is much lower than the maximum loss that could occur for the customer. Hence, partial collateralization is sufficient to reach an optimal outcome for all token holders. Our results exhibit a high practical relevance for the projects run by the Decentralized Insurance Foundation such as parametric hurricane coverage or crop insurance. The latter has already been piloted by Etherisc, Oxfam, and the global insurance broker AON with rice farmers in Sri Lanka and was highly acclaimed by the media.
This post is based on a forthcoming working paper by Alexander Braun, Niklas Haeusle & Stephan Karpischek. Alexander Braun is a Senior Fellow of the Wharton Risk Center.