Insurance and Risk on Defi

DeFi is a new line of business and by DeFinition a new risk sector with particular risks differing from the traditional financial sector and covering intangible risks. It will require modelling and correlation like other sectors but will have the advantage of real time granular data reducing the dependency of assumptions made in risk modelling. The following diagram shows the risk landscape as we perceive it traditionally pre DeFi evolution.

This is a centralized model where intermediaries and third parties perform risk analysis to match the assets with liabilities. Much of DeFi has similar categories, including puts and swaps, but the risk shifts to the technology as in the smart contracts and less to a human counterparty risk. The risk model moves to a decentralized ecosystem model as shown below. Like all risk assessment, frequency and severity of events drive the protection landscape and the parameters of the modelling.

The risk here is threefold – to the ecosystem, to the protocols and to the investor. Regulatory risk looms as regulators are starting to look at the sector with some exchanges under investigation or banned in some jurisdictions. There is a need to be compliant. As DeFi ecosystems need to interoperate, the composability technology (sidechains, parachains, bridges, crosschains) that links protocols, this presents a risk as these standards are still developing and each protocol has a different software code base. Jurisdictions need to be crossed and recent CBDC developments add to the risk. Event risk can occur on any financial market such as the Wall Street crash or Global Financial Crisis.

Technical or operational failures categorise most of the risk as DeFi is a digitized and automated investment sector based on software protocols. Smart contract and oracle failure needs to be mitigated by extensive audits, standards, testing and open source principles. Data integrity and provenance should be a warranty for the insurance but data accuracy is in the hands of the protocols and the investors, so the latter need to manage their personal risks. Updates and configurations to protocols needs to be applied on a regular basis to avoid any accidental forking of protocols^[xvi] which can cause double spending or exploitation loopholes. That said, blockchain is known for trust and security, and technology is quick to fix. Custodial risk (security of keys) occurs where protocol developers hold “admin keys” that are used to push upgrades to the protocol and a risk exists from bad actors. Security of admin keys are operational security that can be compromised due to human error.

Regular audits should be carried out, some in real time and recorded incidents to date have been caused by smart contract bugs, oracle manipulation, whitelisting impostors, theft of digital assets, minting of unlimited tokens and hacks on flash loans, wallets, random number generators and voting. There are multiple attack surfaces and a good security posture is essential. Many attacks are carried out by white hat hackers who often become employed by the organizations they hacked. As DeFi uses economic incentives that are used for users to perform self-regulation and governance, failure of these tenets could lead to a bad outcome.

The investors that purchase insurance or risk tokens buy and take a percentage of risk in the protocol. They are rewarded with a higher rate of return on their principal purchase for taking the risk. Each insurance contract token has a short-term duration and a premium. They are deploying high-risk strategies to seek out yield involving lending, borrowing, liquidity mining or any other means. In the digital asset market with no intermediaries, investors take and shoulder risk. The market risk of having a negative result in investment based on asset or market price volatility is present in DeFi, just as in conventional markets, as crypto is universally known as a volatile asset class open to hacking, arbitrage and market manipulation.

Legal risks question the enforceability of smart contract liability, data privacy laws, intellectual property (IP) by ensuring that dApps being developed do not infringe others intellectual property rights (as managed by IPwe), building arbitration and dispute resolution into smart contracts and consumer protection in general. Audit issues around tax are also pertinent as when data becomes a tangible asset it is liable for indirect tax.

There is a need to get risk adjusted premium income for insurers and liquidity providers which requires stochastic modelling in order to price the risk charge. However, this requires access to prediction markets as well as historical event data. The correlation between the risk run in investing and the performance of the investments is known as the risk-return trade off (higher the risk, higher the reward). Running predictive scenarios prior to smart contract development will assess the worst case scenario so we can mitigate in the crypto world that assets always exceed liabilities. DeFi should not be bucketed into general operational risk.


DEFI INSURANCE

Technology aside, today’s finance sector is based on a centralized authority. Decentralized finance is the antithesis, meaning that a network can make its own decisions based on a consensus of participants. This moves the trust back to the community from insurers while maintaining the integrity of the contracts and mitigating any form of risk. This autonomous shift means that what a third party accomplishes now is done using mathematical techniques and tokens working for the mutual benefit of each community member participant with the help of specific incentives or cost reductions. It requires adaption and understanding.

Decentralized insurance allows for public trades leading to issues around privacy and the autonomous nature leads to a perception that users cannot change their information or interact directly with their data. This has led to disputes and fraudulent cases in the industry as immutability is a crucial part of the blockchain technology. Data is not stored on the blockchain layer but on distributed ledgers and legacy data stores. If ownership of data allows, then access to data is permissible and the blockchain keeps an immutable audit trail of events by hash key. The “right to be forgotten” of privacy laws is not violated. With $2+ trillion of digital assets in circulation and $100 billion of DeFi intellectual property extant, the values at risk has greatly increased making access to (re)insurance capacity a critical need.

The traditional insurance market has been wary about underwriting risks relating to the DeFi space especially where the loss is denominated in crypto. Institutional investors are now entering the crypto world so as the emphasis shifts from early adopters to more risk savvy  investors so insurance becomes the key barrier for entry of their involvement.

DeFi insurance and alternative risk coverage platforms therefore have the potential to fill the crypto-protection gap and to facilitate risk exposures faced by businesses operating in a decentralised economy. As smart contracts mimic parametric insurance structures already utilised in catastrophe management, it makes perfect sense for insurers to bootstrap these into decentralized systems using the same methodology. The insurance sector is also undergoing a major digital transformation with regulated digital insurers emerging and trading online.

There are 4 insurance models developing for digital asset risk:

• Self-insurance, beyond captives, handled by the ecosystem to protect protocols
• Protection of members via a digital mutual/cooperative approach
• Parametric insurance solutions
• Traditional insurers offering risk capacity in the intangible space

There are many intangible covers emerging but base cover is about smart contract protection that might result in loss of funds or identity from a technology failure. The income to the liquidity markets and the sum insured is the TVL amount created by yield farming investors. These protocol owners are cost sensitive about insurance and are searching for protection covers embedded in the ecosystems, essentially insurance tokens. This is a self-governance approach where participants take a percentage of the risk by staking more tokens for higher yield and they in effect take the risk on technology failure as they have access to the open-source audits of the protocols and likely feel comfortable with the mitigation. Conversely, on the risk adverse side investors can purchase claim tokens which reduces their yield in return for a level of cover. So, in effect investors who want to stay hedged against exploit risk buy claim tokens while others who believe the underlying protocol is secure buy premium tokens. When there is an event the claims process will be managed by the participants (token holders) on the network where they vote on a claim and the majority decides which claims are to be paid and the settlement amount. The losses are not directly aligned to the indemnity but to events aligning with a parametric approach. A reserving approach is taken when insurance tokens are purchased, the majority of the premium is returned to the pool to cover the risk and locked in. The participants themselves become the risk assessors and claims handlers.

Efficient on chain governance of claims processes are required to protect users and protocols to quantify smart contract risk to help determine claim pay outs. Oracles activate smart contracts enabling them to access real time data off chain that is related to real world events that trigger claims. The blockchain can be used to mitigate fraud and provide forensic evidence. Once a smart contract is triggered then the assets in the pool should be frozen to allow claims investigation and assessment. If the claim is successful it is paid and the funds in the pool released. This mechanism raises questions of claims reserving and will no doubt get regulator attention, but it is clear how this approach could be applied in an ecosystem.   

An alternative is to utilise a regulated mutual company operating as a DAO where cover is purchased by opt in of members’ whose contributions flow into a capital pool which improves the funding position of the mutual. The cover is priced to generate a long-term surplus which is then mutually shared between the membership base. This means the mutual grows as its capital resources steadily increase over time. A good example is Nexus Mutual

Custody cover will protect users who put funds into an organisation responsible for the safekeeping of cryptocurrency assets private keys on behalf of users. Users are covered if the custodian gets hacked and the user loses more than a certain percentage of their funds, or cannot withdrawal funds from the protocol for a certain number of days. In general, cold storage of cryptocurrencies (held offline) and hot storage (wallets online) are subject to insurance as long as the process to get from cold to hot storage and vice versa is secure.
 

REINSURANCE

Reinsurance will be a natural consequence of the insurance development in DeFi but will be done in a more automated fashion, flattening out hierarchies, and moving automatically to access the various layers in the reinsurance tower.

DeFi tokens can also be utilised to support a sale of an insurance-linked security (ILS) and any capital market participation will need to be on a regulated basis through well-documented SPV structures that offer investors transparency.  It is unlikely that cryptocurrency and digital asset risk will become a significant peril within the ILS market in the short term, but will form a new marketplace for DeFi insurance contracts to raise money from capital markets. If the structures and mechanisms for risk transfer and ultimately securitization are in place, and in the right jurisdiction, then risk could still be transferred using ILS products to capital market investors for whom the returns from an insurance-related digital asset linked to cryptocurrency risk could be attractive.

DeFi attracts funding into pools by giving incentives, similar to coupons, in return for staking crypto money for a period of time to cover the risk. These coupons are tokenised and consist of smart contracts containing tokens managed by the protocol. ILS investors earn interest on staked collateralized crypto assets locked within the smart contract without selling the crypto assets. This is very similar to a traditional sidecar based structure. This would be backed by parametric structures using a self-governance claims mechanism described earlier. The digitized insurance market is starting to develop alongside DeFi to provide capacity. 

Collateral protection insurance brings together leading crypto-backed lenders to share risks and remove bottlenecks. CBDC developments will also be a catalyst to reinsurance once they become established as government backed issuances.

B3i RE^[xix] is an ecosystem company of insurers and reinsurers. It is a blockchain based digital ledger reinsurance placement platform that is capable of structuring and placing reinsurance contracts which are transacted in the market. This gives the ability to automate the reinsurance process by layer using smart contracts. As there is always a single version of the truth, contract certainty is an inherent part of the reinsurance process and B3i ecosystem is available to interoperate with the DeFi networks as the need arises. 

NYDIG are cryptocurrency asset managers and insurance professionals and are new entrants, capitalised with BTC and are an insurance industry world first.

Insurers find it difficult to use traditional insurance covers to insure against potential DeFi and Crypto “Black Swan” events so parametric risk management products are feasible alternatives which would protect specific digital assets backed by Stablecoins using a combination of smart contracts and data feeds. 

Smart contracts and protocols can be linked together to form insurance pools in a placement marketplace where underwriters and risk managers/buyers seamlessly interact. Market players decide on which insurance pool is to be used to hedge the crypto risk.  A smart contract is mathematically measurable, has formal security and can be embedded in policy wordings. Parameterization in trigger design sets the range of values and conditions for which a policy needs to meet the pre agreed sum insured before the trigger is alerted and a quick payment is made, removing the need for lengthy claim investigations, subrogation, claim reserving  and expenses. When DeFi is the line of business, claims investigations also include elimination of network “gas” fees based on time spent in gaining consensus. Parameterization can be complex to DeFine the events and result in nested smart contracts. Procedures would need to be put in place for a delay period if any human interaction was required to approve the triggers to pay the claim and address basis risk.

Since blockchains cannot make external calls due to built-in functionality to preserve security, the claims process requires an additional piece of secure middleware known as a blockchain oracle to bring data on-chain. Oracles retrieve external data on behalf of the smart contract, validate it via unique cryptographic signatures, and broadcast it on the blockchain to be ingested by the smart contract proving the origin of data supplied.