The broker model for peer-to-peer insurance: an analysis of its value


The increasing role of technology is generating new challenges in fostering innovative insurance products and in offering new means of distribution for existing products. Herein, we focus on peer-to-peer insurance, where technology is used to connect policyholders and the insurance structure recalls its roots based on organised mutual solidarity. The aim is to highlight strengths and weaknesses of the broker model of peer-to-peer insurance. We address the main legal and regulatory issues by referring to European Union regulation on insurance and we develop an actuarial model to support our conclusions and evaluate the convenience of such a business model for policyholders. Although peer-to-peer insurance was created as a fairer alternative to the pool of traditional insurance companies, we see major challenges that need to be solved. The evidence sheds doubt on the convenience of the broker model for peer-to-peer insurance while the regulatory risks seem, paradoxically, certain.

This is a preview of subscription content, log in to check access.

Fig. 1
Fig. 2
Fig. 3


  1. 1.

    The pioneer in the development of this model of insurance was Friendsurance (D) in 2010, followed by Guevara (U.K.) and Inspeer (FR). TongJuBao (CN) and Lemonade (U.S.) followed.

  2. 2.

    This definition was later reproduced by IAIS (2017), p. 40.

  3. 3.

    The Institute of International Finance (IIF 2016) p.11, outlines that “Unlike P2PI brokers, P2PI carriers do the actual underwriting and offer their policies directly to consumers online. The business model allows investors to participate in the process and contribute money toward the capital reserves for the various insurance pools. Once the required period of time has elapsed, and after all claims have been paid, the investors and the carrier firm divide the outstanding balance remaining in the premium pool.”.

  4. 4.

    Teambrella uses this model (Paperno et al. 2016).

  5. 5.

    EIOPA (2019), p. 29 reported that such a technical service provider might leverage Blockchain and smart contracts and facilitate users coming together and creating their own “pools”.

  6. 6.

    For instance, FriendInsurance, Inspeer, PeerCover, Guevara and Axieme follow this model.

  7. 7.

    The IIF outlined that most P2P insurance schemes act as brokers and aim to lower the cost of insurance for their consumers by pooling policyholders together online and leveraging their buying power. (IIF 2016).

  8. 8.

    EIOPA (2019), p.26 outlined that it is a matter of evaluating concrete business models and the outcome can be that a platform is operating under insurance regulation, or that it is outside of regulation, e.g. in the context of payments services.

  9. 9.

    Lima Rego and Campos Carvalho (2019) reached the conclusion that, to a large extent, the leading characters in P2P insurance appear to play the same roles traditionally ascribed to insurers and insurance intermediaries in ordinary B2C contexts.

  10. 10.

    From the website: “Wir verbinden Sie mit anderen Versicherten zu einer Gruppe von durchschnittlich 10 Personen. Die Gruppe macht Ihren Bonus überhaupt erst möglich. Ohne Gruppe geht es nicht”, i.e., we connect you with other insured persons to a group of on average 10 persons. The group makes your bonus possible in the first place. It cannot be done without a group. Axieme connects people to other people with the same rating that bought the same contracts. Lemonade places people that select the same charity for the cashback in the same group.

  11. 11.

    Commission Implementing Regulation (EU) 2017/1469 of 11 August 2017 laying down a standardised presentation format for the insurance product information document details both the format and contents of the IPID, which are harmonised across the member States.

  12. 12.

    See Commission Delegated Regulation (EU) 2017/653 of 8 March 2017 laying down regulatory technical standards with regard to the presentation, content, review and revision of key information documents and the conditions for fulfilling the requirement to provide such documents.

  13. 13.

    Accordingly, the co-manufacturing regulation would be extended to include the following case: (i) an insurer agrees to supply an insurance product, and (ii) an insurance intermediary assembles this product with a product other than insurance, and (iii) the intermediary presents the assembled product to the customer as if it provided a full insurance guarantee.

  14. 14.

    See article 5, para. 2, of Directive 2005/29/EC.

  15. 15.

    The Product Disclosure Statement of FreindInsurance Cashback Bike Insurance Product outlines that: “Connections include You and each other person You are connected to through MyPlace for entire Calculation Months”.

  16. 16.

    Different rules are defined in practice. For instance, Lemonade gives back the money left over to specific causes the policyholder cares about (subject to board discretion and the company meeting certain financial standards).

  17. 17.

    Alternatively, the claims could be capped in this case. However, as stressed by EIOPA (2019), this solution introduces a potential inequality between first claims (entirely indemnified) and claims paid at the end of the year (potentially not indemnified at all).

  18. 18.

    For instance, \({r}_{g}\) is usually up to 10 in P2P groups (see and

  19. 19.

    These assumptions can be violated in some specific contexts (for instance, the independence between number and cost in case of cat claims) and specific models have been developed for treating these issues (see, for instance, (Garrido et al. 2016)). For our purposes, we are neither interested in verifying if these assumptions are satisfied (or not) nor in identifying which is the best distribution for the claim size distribution.

  20. 20.

    In some contexts it is also defined as risk premium (see, for instance, (Daykin et al. 1994)). Actuarially fair insurance has an expected net pay-off of zero. From a consumer's point of view, an insurance contract is actuarially fair if the premiums paid are equal to the expected value of the compensation received. This expected value is, in turn, defined as the probability of the insured-against event occurring multiplied by the compensation received in the event of a loss.

  21. 21.

    Notice that in Braun et al. (2015), premiums charged by a stock insurer are defined as the difference between discounted expected aggregate loss and the so-called default put option of the stock insurer. In practice as well as in the actuarial literature, it is not so common to include the default option of the stock insurer (see, for instance, (Daykin et al. 1994), (Werner and Modlin 2016)). Capital allocation can obviously be introduced in pricing contracts (see, for instance, (Gründl and Schmeiser 2007)).

  22. 22.

    This criterion is typically applied by Lemonade, FriendInsurance and Axieme.

  23. 23.

    This criterion is typically applied by Guevara and Inspeer.

  24. 24.

    For instance, Axieme outlines on its website that the total premium is given by the premium paid to the third-party, the fees of the broker and the cashback.

  25. 25.

    For instance, Lemonade takes a flat 20 percent fee off the premium to cover salaries, running costs, and technology replacement.


  1. Banhemann, D. 2015. Distribution for Actuaries. CAS Monographs, 2.

  2. Biener, C., and M. Eling. 2012. Organization and Efficiency in the International Insurance Industry: A Cross-Frontier Analysis. European Journal of Operational Research 221 (2): 454–468.

    Google Scholar 

  3. Braun, A., H. Schmeiser, and P. Rymaszewski. 2015. Stock vs. mutual insurers: Who should and who does charge more? European Journal of Operational Research 242 (3): 875–889.

    Google Scholar 

  4. Braun, A., and F. Schreiber. 2017. The current InsurTech landscape: business models and disruptive potential Study Report. St. Gallen: Institute of Insurance Economics.

    Google Scholar 

  5. Bühlmann, H. 1980. An economic premium principle. ASTIN. Bulletin 11 (1): 52–60.

    Google Scholar 

  6. Cummins, J.D., and R.A. Derrig. 1988. Classical Insurance Solvency Theory Huebner International Series on Risk Insurance and Economic Security. Norwell, MA: Kluwer Academic Publishers.

    Google Scholar 

  7. Cummins, J.D., M. Rubio-Misas, and H. Zi. 2004. The Effect of Organizational Structure on Efficiency: Evidence from the Spanish Insurance Industry. Journal of Banking & Finance 28 (12): 3113–3150.

    Google Scholar 

  8. Cummins, J.D., M.A. Weiss, and H. Zi. 1999. Organizational Form and Efficiency: The Coexistence of Stock and Mutual Property-Liability Insurers. Management Science 45 (9): 1254–1269.

    Google Scholar 

  9. Delbaen, F. 2000. Coherent Risk Measures. Pisa: Cattedra Galileiana.

    Google Scholar 

  10. Daykin, C.D., T. Pentikäinen, and M. Pesonen. 1994. Practical Risk Theory for Actuaries. London: Chapman & Hall.

    Google Scholar 

  11. EIOPA. 2019. Report on best practices on licencing requirements, peer-to-peer insurance and the principle of proportionality in an insurtech context. Retrieved from March 2019

  12. Garrido, J., C. Genest, and J. Shultz. 2016. Generalized linear models for dependent frequency and severity of insurance claims. Insurance: Mathematics and Economics 70: 205–215.

    Google Scholar 

  13. Gomber, P., J.-A. Koch, and M. Siering. 2017. Digital finance and FinTech: current research and future research directions. Journal of Business Economics 87: 537–580.

    Google Scholar 

  14. Gründl, H., and H. Schmeiser. 2007. Capital Allocation for Insurance Companies. What Good is It? Journal of Risk and Insurance 74 (2): 301–317.

    Google Scholar 

  15. Institute of International Finance. 2016. Innovation in Insurance: How technology is changing the industry. Retrieved form

  16. Kahane, Y. 1979. The Theory of Insurance Risk Premiums—A Re-examination in the Light of Recent Developments in Capital Market Theory. Astin Bullettin 10: 223–239.

    Google Scholar 

  17. IAIS. 2017. FinTech Developments in the Insurance Industry

  18. Lima Rego, M., and J. Campos Carvalho. 2019. Insurance in today’s sharing economy: new challenges ahead or a return to the origins of insurance? In InsurTech: A Legal and Regulatory View, ed. P. Marano and K. Noussia, 27–47. Berlin: Springer.

    Google Scholar 

  19. MacMinn, R., and Y. Ren. 2011. Mutual Versus Stock Insurers: A Synthesis of the Theoretical and Empirical Research. Journal of Insurance Issues 34 (2): 101–111.

    Google Scholar 

  20. Marano, P. 2017. Sources and Tools of the Insurance Regulation in the European Union. In Insurance Regulation in the European Union, ed. P. Marano and M. Siri, 5–29. Springer, Berlin: Solvency II and Beyond.

    Google Scholar 

  21. Marano, P. 2019. Navigating InsurTech: The digital intermediaries of insurance products and customer protection in the EU. Maastricht Journal of European and Comparative Law.

    Google Scholar 

  22. Mayers, D., and C.W. Smith. 2013. On the choice of organizational form: theory and evidence from the insurance industry. In Handbook of Insurance, ed. G. Dionne, 669–687. Geneva: The Geneva Association.

    Google Scholar 

  23. Möhlmann, M. 2016. Digital Trust and Peer-to-Peer Collaborative Consumption Platforms: A Mediation Analysis. Retrieved from SSRN

  24. National Association of Insurance Commissioners (NAIC). 2017. Peer-to-Peer (p2p) Insurance. Retrieved from

  25. OECD. 2017. Technology and innovation in the insurance sector. Paris: OECD.

    Google Scholar 

  26. Oryzak, D., and A. Verna. 2015. Insurance 2.0: Insuring the Sharing Economy & Sharing the Insurance Economy, Casualty Actuarial Society E-Forum, Summer 2015.

  27. Paperno, A., V. Kravchuk and E. Porubaev. 2016. Teambrella: A Peer to Peer Insurance System (Whitepaper). Retrieved from

  28. Savelli, N., and G.P. Clemente. 2011. Hierarchical Structures in the Aggregation of Premium Risk for insurance underwriting. Scandinavian Actuarial Journal 2011 (3): 193–213.

    Google Scholar 

  29. Scollnick, D.P.M. 2007. On composite lognormal-Pareto models. Scandinavian Actuarial Journal 2007 (1): 20–33.

    Google Scholar 

  30. Scollnick, D.P.M., and C. Sun. 2012. Modeling with Weibull-Pareto Models. North American Actuarial Journal 16 (2): 260–272.

    Google Scholar 

  31. Werner, G., and C. Modlin. 2016. Basic Ratemaking. Arlington: Casualty Actuarial Society.

    Google Scholar 

  32. Wüthrich, M. V. 2019. Non-Life Insurance: Mathematics & Statistics. Retrieved from SSRN: or

Download references

Author information



Corresponding author

Correspondence to Gian Paolo Clemente.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendix A: Details on the calculation of P2P premiums in different scenarios

Appendix A: Details on the calculation of P2P premiums in different scenarios

In formula (4), we defined the total fair premiums of the P2P portfolio as:


where \({P}_{i,g}^{F}\) is the premiums’ portion kept in the mutual pool, \({P}_{i,g}^{TPI}\) is the fair premium paid to third-party insurance companies and \({P}_{i,g}^{TPI2}\) is used to provide an insurance coverage to cover minor claims in case the mutual pool is empty.

We denote with \({S}_{g}^{L}\) and \({S}_{g}^{S}\) the total aggregate claims amount, for the group \(g\), over and below the franchise limit \(t\), respectively. These two r.v.s are defined as:

$${S}_{g}^{L}=\left(\left(\sum_{i=1}^{{N}_{g}}{X}_{i,g}\right)|{X}_{i,g}>t\right)\, and\, {S}_{g}^{S}=\left(\left(\sum_{i=1}^{{N}_{g}}{X}_{i,g}\right)|{X}_{i,g}\le t\right).$$

Obviously, we have \(S_{g} = S_{g}^{L} + S_{g}^{S}\).

The total expected cost, considering both claims and possible cashback, can be defined as:

$$P^{P2P} = E\left( {\mathop \sum \limits_{g = 1}^{G} S_{g}^{L} } \right) + \mathop \sum \limits_{g = 1}^{G} E\left( {{\max}\left( {S_{g}^{S} - \mathop \sum \limits_{i = 1}^{{r_{g} }} P_{i,g}^{F} ,0} \right)} \right) + \mathop \sum \limits_{g = 1}^{G} E\left( {{\min}\left( {S_{g}^{S} ,\mathop \sum \limits_{i = 1}^{{r_{g} }} P_{i,g}^{F} } \right)} \right) + E\left( {CB} \right).$$

In formula (A.1), we have the following terms:

  • The expected value of aggregate amount that regards only claims larger than the limit. This amount is equal to the sum of fair premiums charged by the third-party insurance \({P}_{i,g}^{TPI}\) to each policyholder \(i\). In other words, it represents the total premium of a portfolio of homogeneous insurance contracts with a franchise equal to \(t\);

  • The expected value of the total amount considering only small claims (i.e. lower than the limit) and covering only the case of an empty pool. It represents the fair premium of the specific coverage \({P}_{i,g}^{TPI2}\), aggregated at portfolio level (i.e., the sum of premiums for each group);

  • The expected values of claims covered by the pool. For each group \(g\), this amount is upper bounded by the initial amount placed in the pool \(\sum_{i=1}^{{r}_{g}}{P}_{i,g}^{F}\);

  • The expected cashback. This amount is obviously related to the cashback rule defined by the entity managing the P2P insurance and it will be explored subsequently.

In formula (A.1), we can focus on the second and third terms


Given the linearity of the mean, we can rewrite formula (1) as:

$${P}^{P2P}=E\left(\sum_{g=1}^{G}{S}_{g}^{L}\right)+\sum\limits_{g=1}^{G}E\left({S}_{g}^{S}\right)+E\left(CB\right)=E\left(S\right)+E\left(CB\right)= P+E\left(CB\right),$$

where in the last step we have assumed that the expected aggregate amount of claims \(E\left(S\right)\) is not affected by the model. In other words policyholders file the same claims in either a P2P model or a traditional stock insurer model.

Different criteria can be set up in order to assess the expected value of the cashback. We provide two different alternatives:

  1. 1.

    Any positive amount remaining in the mutual pool is paid back to the policyholders of the group (or to a charity). In this case, we have:

  2. 2.

    To each policyholder in the group (or to a charity) is paid the difference, if positive, between the Potential Maximum Cashback (\(PMCB_{i,g}\)), defined by the entity managing the P2P insurance, and the total claim payments related to claims made by policyholders in the group

    $$E\left( {CB} \right) = \mathop \sum \limits_{g = 1}^{G} \left( {\mathop \sum \limits_{i = 1}^{{r_{g} }} PMCB_{i,g} - E\left( {min\left( {S_{g}^{{}} ,\mathop \sum \limits_{i = 1}^{{r_{g} }} P_{i,g}^{F} } \right)} \right)} \right).$$

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Clemente, G.P., Marano, P. The broker model for peer-to-peer insurance: an analysis of its value. Geneva Pap Risk Insur Issues Pract 45, 457–481 (2020).

Download citation


  • InsurTech
  • Distribution of insurance products
  • Peer-to-peer insurance
  • Premium rating
  • Insurance broker
  • Insurance regulation