Blockchain x credit insurance = ?
The purpose of this paper is to provide an overview of the main drivers for blockchain adoption in trade credit insurance and the related opportunities. This analysis is solely Alpine Style’s opinion, and does not aim at reflecting the views of our customers and partners in any way.
Our industry is changing under the combined effect of new technologies and business models. In particular, platform business models, APIs, data, machine learning, venture capital approach and — obviously — blockchain.
Blockchain is essentially data, stored in a decentralised and logical manner (the distributed ledger), with robust integrity (anti-tampering protection with various cryptographic techniques). It can be used in wide range of cases, most of which have nothing to do with cryptocurrencies. The later evolutions of blockchains add interesting features like automated triggers (smart contracts) and scalability for real-world use cases.
Credit insurance is a really (really) great product, when you know about it: too bad most companies have no clue about credit insurance :-). And indeed our biggest problem is the size of our industry (volume of insured B2B transactions) compared to the total addressable market (whole B2B trade). Current trade credit insurance is a very relevant solution, but for a very small number of companies.
We know the weaknesses of credit insurance: to name a few, the average credit insurance user experience is weak, our policies are complex and it takes at least 5 meetings to sell most of our products, our underwriting decisions can lack transparency at times. But all this is nothing compared to our need to get more data. To get 1000x more data.
Because we protect such a small portion of the whole B2B trade, and because of how credit insurance was designed and operated so far (e.g relying on structured data), we have access to a tiny fraction of the world’s B2B data. We are like analog camera craftsmen, in an Instagram world: we enjoy unparalleled expertise and strong barriers to entry, we are the masters of a profitable market but small and not growing, somehow comfortable away from the digital reality. A few of us may continue building great analog cameras, because why not? And yet, why not also try to capture the bigger opportunities (and grow like one of those tech and data unicorns)?
The bigger opportunity (services aiming at the whole B2B trade) requires to master the rules of data — and a little bit of technology too. What does it have to do with blockchain? Blockchain is a key, if not the key, to unlocking access to 1000x more data.
Have you ever dreamt of underwriting based on all the data of all the corporates? While keeping every corporate’s data strictly private? This dream is now a possible reality, and it is powered with blockchain. Let’s see / discuss how.
Practical use cases of blockchain in trade credit insurance
Trade credit insurance users are looking for our expertise and risk mitigation capabilities. We can agree that they are not early adopters of new technologies, nor ambassadors of a decentralised future.
Are credit insurance customers looking for blockchain? Absolutely not: for most of our users, blockchain is still a buzzword and a complicated technology, that they do not experience in their everyday life yet. They hear about it in the media, and sometimes they get confused with the few illegal use cases of certain cryptocurrencies. Our most innovative users may have tried to play with blockchain by earning a fraction of a bitcoin on Earn, and even these users probably got frustrated with the quite overwhelming technicalities of MetaMask, Trust …
It is funny because meanwhile, we see all these glorious press releases about companies, banks or insurers advertising their use of blockchain. For me, it means that we are still at the stage of “blockchain washing”, where opportunists try to get attention by using the big words of the crypto world. Real use cases ( i.e adopted by real users) will come when we will no longer speak about blockchain or tokens: they will be embedded, hidden even, in our technology stack and in our business model, serving an unparalleled user experience, without any jargon or technicalities.
So let’s look for real blockchain use cases in our industry, that our users may want and helping us protect a greater fraction of B2B trade. The 3 following use cases are presented in growing order of business potential size.
Use case 1: allowing credit insurance customers to check the integrity of policies
One of the most immediate candidate use cases for blockchain in credit insurance is to facilitate the verification of the contracts’ signatures and clauses integrity. You know how it works: after a number of iterations between the client, the broker, the commercial team and the underwriting team, the terms of the policy finally get approved and signed. Sometimes, they even get signed digitally (great!).
Can blockchain be embedded in this process, and help in any way in the policy signature process? Yes it can, and by doing so it can help users trust that the clauses that will be applied are really the ones that were agreed. It basically removes the possibility for the insurer, the broker and the client to argue on different versions of the policy after the signature.
In practical terms, this requires to create an immutable record of the agreed policy. Blockchains are great at creating immutable records, by design. The process is as follows ( bear in mind, the user will never need or want to be aware of these technicalities):
- Generate a one-way cryptographic hash of the signed contract (a unique string of letters and numbers, like a digital fingerprint of the policy)
- Record it in your blockchain of choice. Unless the blockchain is hacked (which is practically impossible for public blockchains), this record is effectively immutable.
- Anyone with a copy of the signed contract can check it against the blockchain-stored evidence to verify the copy’s integrity, by generating the hash again and comparing both hashes.
The user interface is purely on the web or in a mobile app, and does not require any technical or crypto know-how. Of course, the content of the credit insurance policy is never written to the blockchain or exposed publicly. The one-way cryptographic hash cannot be used to reverse engineer the original policy.
Illustrations of this use case outside of trade credit insurance include Docusign, BlockSign (I love BlockSign), OpenSig (although deprecated and without user-friendly interface), Signatura, and with improvements on the legal enforceability: Clause, OpenLaw… We can also be inspired by the experiments and solutions helping factors to avoid double-financing (for example Innopay and TradeFinanceMarket’s InvoiceCheck).
Use case 2: documents and trade data distribution / blockchain-based trade finance consortiums and standalone initiatives
The immutability of the blockchain and its distributed nature opens up to the quite obvious, but nevertheless interesting, use case of trade document and contracts management across supply chains. Our rulebooks, processes and technologies failed to properly track and manage the trade-related documents and their changes across the multiple parties involved in open account transactions, letters of credit, … The better version, with blockchain inside, holds the promise to have everyone always work on the same version of the docs, have access to (only) the data they need and subsequently, to generate significant gains in efficiency while improving trust.
Fundamentally, as credit insurance is a fix for trust and risk, this is not particularly great news for credit insurers: but let’s face it, we have not been particularly good at fixing the data gap across supply chains in the past, and the blockchain revolution of trade finance is unstoppable anyway, as there are a number of global consortiums already ( Marco Polo, we.trade, Batavia (which apparently just merged with we.trade), Voltron, HKTFP, …), and an even greater number of standalone initiatives pairing innovators and banks.
As far as I know, the first application of credit insurance with blockchain was precisely with this use case: I helped Skuchain protect transactions from credit risk with the very first version of Single Invoice Cover in Q3 2015, that I had designed and built with L’Atelier. It was the same use case again with Hijro (formerly Fluent), and a number of others. From there on, it was obvious that credit insurance has a lot to gain from participating in such use cases, although it may not require the insurer to use blockchain: the key is to take a a wholistic view on the supply chain, and subsequently to leverage the end-to-end data in our underwriting.
The opportunity for us with such use cases is still wide open: apart from AIG and Euler Hermes who support some of these initiatives, the other consortiums do not seem to have credit risk protection solutions. And the standalone initiatives need our help too, while Hokodo is showing the way with its partnership with Centrifuge.
Use case 3: data marketplaces to expand the reach of credit insurance
Data marketplaces are probably the biggest opportunity for credit insurers with blockchain right now (and for our information providers too). A few players leverage blockchain to create platforms to unlock the access to data, and in particular unstructured data. These platforms make more data available to more organisations, liberating information once held under lock and key. This creates immense value in industries where access to data is critical: healthcare, mobility / IoT, consumer credit … and of course trade finance with access to credit risk data.
Blockchain-based consumer credit scores are a good proxy to understand how such platforms could work in our industry. A number of players aim at fixing the shortcomings of the current credit scoring systems. Bloom, Colendi, … are animated by the vision that the world deserves something better than the FICO score or its equivalent in other countries. The way they usually work is to build on a platform that lets users attest their identity, expose and register their data (for example current and historical debt obligations). The platform then uses this data to build a metric of the users’ creditworthiness. Scores in markets and communities with sparse data are bootstrapped by using peer-to-peer credit stakes.
In the credit insurance industry, we mostly underwrite risk based on sparse and outdated information. The largest credit insurers have databases of a few tens of millions of companies, filled with data that is frequently 2 years old or more, and with very patchy coverage in many countries. This is fine for the business we focus on, but this is also one of the reasons why we refrained from covering a broader fraction of B2B trade so far: transactions involving new or fast-growing companies, freelancers, SMEs, …
Blockchain changed that (already): now we can cover a broader fraction of B2B trade, by buying data on the related companies. Eqitii seems to be the most advanced data marketplace for our industry, both in terms of how well they built their technical platform, and in terms of the traction the get with relevant data owners for our underwriting needs. Credit risk data owners on Eqitii will include utility providers, commercial financial service companies, real estate service providers … The rationale for these data owners is to monetise their data assets, while keeping the data strictly private and secure.
On the other side of Eqitii data marketplace, credit insurers and other trade finance data buyers make offers: data owners control all data rights and have the power to accept or reject any offer. Exclusive data rights are made possible and enforceable thanks to advanced fingerprinting of the data and zero-knowledge proofs. When the offer is accepted, the data is transferred to the credit insurer in a separate secure channel.
In the credit insurance industry, our current scoring models (and the systems we use to run these models) may not easily swallow this new (big) data, especially the unstructured data. However building a new model is a rather good problem to solve (and a relatively easy one), knowing that the data increase allows our new models to vastly outperform the old models (this topic will deserve a publication on its own, but in the meantime, you can dig into the machine learning papers from before the Internet bubble, for example this one). Once this is done, we will virtually be capable of delivering a reliable score for any company, and to update our limits within seconds after any change (good or bad) in the situation of the company.
Inspirations for credit insurance 10 years from now
Beyond the 3 use cases mentioned above, if / when we overcome our fear to self-disrupt, we can truly embrace the blockchain philosophy and reinvent our business with the “10 years from now” horizon in mind, to address the needs of the big parts of the global economy that are currently beyond our reach. Here are a few illustrations taken from upcoming Alpine Style ventures:
Reinvention 1: parametric credit protection
Parametric insurance ex ante agrees to make a payment upon the occurence of a triggering event. The payment itself does not necessarily indemnify the pure loss: for example it can be a fixed amount. With parametric insurance, commercial underwriting (premium setting) and loss assessment (claims handling) are less costly than traditional insurance products.
The second generation of blockchains (Ethereum, …) introduced the notion of smart contracts, which in practical terms are programs activated by certain triggers, run on the blockchain. Smart contracts can monitor certain information sources and automatically trigger payments when certain conditions are met.
With these principles, Alpine Style is working on a parametric credit protection solution, that indemnifies companies when their customers fail to pay on time. At launch, this solution will only cover cases where the customer is declared bankrupt and the bankruptcy is published on official public sources, but we see opportunities to extend the concept and to protect companies against most cases of late payments.
Outside of credit insurance, AXA Parametric is leading the way in leveraging smart contracts for parametric insurance with Fizzy, its flight delay insurance solution.
Reinvention 2: protection for the crypto economy
In the crypto economy, entities transact and their behaviours tend to present risk patterns that echo what we cover in the general economy: credit risk, execution risk, fraud risk …
In particular, several domains of the crypto economy are currently looking for risk mitigation solutions, and most of the players do not suspect the existence of trade credit insurance and the sureties we are able to offer: however, the essence of our products can translate well in this environment.
For example, crypto-backed lending platforms ( Salt, Cred, Unchained, Nexo, Teneos, BlockFi, MakerDAO …) lend money from investors to crypto-wealthy individuals. In case these individuals default, the collateral is in crypto currencies, which the investor will want to liquidate in fiat currency. What if the crypto exchanges fail to execute this liquidation on the spot? Situations of this kind can typically be hedged with a well-designed “execution surety”.
As far as we saw from the experiments we initiated in this field with Alpine Style, this is a greenfield opportunity awaiting credit insurance trailblazers.
Reinvention 3: distributed trade protection
We saw earlier how blockchain creates the possibility for a new credit risk data paradigm: marketplaces where companies (data owners) monetise their data asset with credit insurers (data buyers) directly. If we believe in the decentralised approach can succeed long term (e.g if the 3rd generation of blockchain manages to overcome the scalability and governance challenges it is meant to solve) and if we embrace the risk to self-disrupt, we create the possibility for a deeper reinvention or our business, beyond the data pillar.
Can we imagine to augment a credit insurer (like Ironman’s exo-skeleton) with a bolder view on what the blockchain has to offer? Let’s envisage a 100% decentralised trade protection solution: indeed two of our most critical pillars, underwriting and risk transfer (balance sheet) can be redesigned in a distributed manner, assuming we already fixed the data question by joining a data marketplace (see above).
Underwriting is essentially the application of models and expertise to decide to cover a given risk and to define the relevant premium. In the asset management world, Numerai is a demonstration that the application of models and expertise can be distributed. This hedge fund sends encrypted market data to any data scientist who wants to compete to model the stock market. They combine the best model submissions into a meta-model, trade that meta-model, and pay data scientists whose models perform well.
Similarly, we can create an ecosystem of model builders who constantly augment our in-house expertise and improve our own underwriting models and expertise, with potential access to quasi infinite brain power and creativity. Suddenly, underwriting transactions in a challenging industry (diamonds …), in a shaky economic period (bubbles, pre- / post-crisis), or in a fast-growing / atypical segment of companies (freelancers, startups …) is easy to try, and offers the promise of a continuous cycle of improvements of our models. And yes, this will be yet another challenge for the regulators, who already had to adapt their framework to the AI revolution coming to our industry.
Another promising example of a such a promising protocol is Ocean. They build a marketplace with private machine learning at its core and blockchain-based (token) incentives. Private machine learning allows for training to be done on sensitive private data without revealing it. The token incentives allow the marketplace to attract the best data and models, and to make them smarter.
Risk transfer is the process of swapping (and usually pooling) the uncertainty (credit risk, …) between two entities (an insured party and a balance sheet) in exchange of a payment (the premium). The developments in Peer-to-Peer (P2P) / mutual insurance paved the way for a decentralised version of risk transfer mechanisms — and blockchain is the technology to operate it in a transparent and cost-efficient manner.
In practice, we could help companies to cover certain risks, without carrying these risk ourselves: once the premium has been set, the participating companies put their money into an escrow account. In case of a claim, the participating companies approve/reject its validity with a voting mechanism, and the amount is paid to the beneficiary. Any remaining money is re-distributed to the participating companies. Blockchain handles the payment workflows and the voting mechanisms very well. The solution could be applied to credit risks beyond our risk appetite, or to other risks (dispute, fraud, …), helping us to deliver an inclusive protection against the various risks associated with trade.
Did you find this paper useful? Give us a few claps and share it to show how much you liked it :-) You can also follow Alpine Style’s hot news and areas where we invest on our LinkedIn page or by subscribing to our newsletter.
Originally published at https://www.linkedin.com/pulse/blockchain-x-credit-insurance-christophe-spoerry/.
