Product Update Issue 3: Risk Management

How to manage risk on the lending and protection platform

Published in

UNN Finance Updates & Ideas

7 min readJan 25, 2021

In our previous product update, we explored C-OP’s economic advantages by the numbers. We revealed that C-OP augments borrowing capacity during market upswings, protects borrowing capacity during adverse price movements, and mitigates downside liquidation risk significantly for borrowers on lending platforms during volatile market conditions. As we demonstrated in the previous product issue:

“C-OP has non-linear cost savings to liquidations: the bigger the gap down in asset price, the more the C-OP saves the borrower.”

Now that we’ve introduced C-OP and covered its design mechanics, we wrap up the “under the hood” of the UNION products with risk management.

TLDR:
Solvency in the protection pool and lending pool benefits all stakeholders.
For the C-OP pool, a put option writer, we mitigate downward gap exposure with full repricing to calculate MCR (minimum capital requirement) and are exploring approaches to delta/gamma hedge. For the lending platform, using two parameters MPC (maximum protection composition) and CFPL (capacity to free post liquidation), we limit derivative concentration and increase stability in downside gap markets.
Skip to end for teaser screenshot.

C-OP Protection Pool Risk Management

At a high level, C-OP protection liquidity is sourced by UNION capital pools that underwrite protection by optimizing the leverage of protection writers. C-OP liquidity pools are constant writers of put options, which means liquidity providers are also put writers. Rewards from the pool are shared across the liquidity providers, as are the risks of a claim. On the other side of the supply-demand equation are buyers of C-OP, who are put option buyers.

The buyer demand for C-OP is contingent on C-OP’s use as an instrument for mitigating liquidation risk and preserving borrowing capacity on lending platforms.

As a result, maintaining the solvency of the protection pool is critical. Should the pool’s capital fall short of covering the protection being written, the pool can become insolvent. The capital model is the foundation for ensuring that capital locked in the protection pool (i.e., underwriter capital) is sufficient to meet the solvency requirements of the protection being underwritten (C-OP) by the pool.

Protecting Solvency with MCR/SCR

UNION’s protection pool relies on venerated models from TradFi common in the insurance industry. In particular, Minimum Capital Requirement (MCR) and Solvency Capital Requirement (SCR), paired with economic incentives via excess capital supply that is allocated for community development and bolstered reward incentives for UNN token holders.

In TradFi, MCR and SCR are the amounts of capital insurance companies must hold to cover underwritten risk under the EOIPA’s Solvency II Framework. In the case of C-OP protection pools, this equates to a specific ratio of capital in the protection pool relative to the amount of C-OP protection purchased. The risk model underlying MCR and SCR is Value at Risk (VAR), which equates to thresholds of 85% for MCR and 99.5% for SCR.

Using MCR, we can gauge the capital pool’s health relative to its underwritten risk — deploying funds as necessary. For example, suppose the capital pool is in excess of the MCR. In that case, the surplus capital is deployed for low-risk yield opportunities — creating additional revenue for pUNN holders (pool underwriters). However, if the pool capital drops below the MCR, then funds are withdrawn from the surplus pool until the MCR is met.

To capture the non-linearity of option risk, UNION is experimenting with two well-tested financial approaches: delta-gamma approximation and full repricing. Both approaches rely on first shocking the underlying collateral to its VaR price, and then calculating the resulting market value shock of the portfolio.

Delta-gamma approximation uses Taylor Expansion to estimate option prices with a quadratic equation consisting of the Delta and Gamma of the option. The benefits are simple and quick calculations. However, with the large realized volatilities of cryptos, which consistently exceed 100% annualized, and the large stresses imposed by MCR/SCR, the delta-gamma estimation error becomes substantial.

The second approach reprices each position with the shocked underlying asset price. As a reminder, UNION utilizes the binomial-tree model for pricing. The benefit is much higher accuracy at the cost of longer compute times. However, because each C-OP pool will have exposure to only one asset, the trade-off is manageable. At the time of this writing, UNION is likely to proceed with full repricing to calculate VaR.

MCR calculations will dynamically adjust based on market conditions, with MCR parameters occasionally subject to governance votes. The MCR calculations for the C-OP pool will begin conservatively and subsequently be optimized to be capital efficient without sacrificing stringent solvency risk standards.

Solvency beyond MCR/SCR

Since the C-OP pool is writing put options, the pool’s risk exposure is unbalanced towards downside price swings. Particularly in crypto markets, where the underlying collateral is often volatile (and sometimes illiquid) crypto-assets, hedging non-linear downside price risk is crucial. Instances where crypto markets gap down swiftly and with momentum, such as March 13th, 2020, often cause massive, adverse ripple effects across DeFi. For example, cascading liquidations, over-collateralized stablecoin peg deviations, and zero-bid auctions of lending platform collateral have all occurred due to non-linear downside risk exposure.

Interestingly, in normal crypto market conditions, puts are generally underpriced (until a bear market is established). Crypto market participants tend to view crypto assets through a bullish lens, which is exhibited by persistently positive perp swap funding rates. Add a recent bull run of crypto-asset prices, and going short crypto is better left to professional trading desks. With options, DeFi users’ preference to buy calls over puts manifests in the higher implied volatilities of strikes versus puts for at-the-money strikes.

Therefore, we feel there are good approaches to investigate which protect the solvency of the pool, helping writers earn more rewards by mitigating assignment risk. For example, we can periodically offset the pool’s risk by buying puts from other DeFi platforms, such as Hegic.

Solvency begets Efficiency

As C-OP’s function is risk mitigation with the upside benefit of improved capital efficiency, it serves as a dynamic tool for lending platforms and their users in DeFi to wield at their necessity. But capital efficiency isn’t explicitly confined to the C-OP instrument itself. It’s also vital to maintain capital-efficient underwriting pools of the C-OP product to attract liquidity providers to the pool, further augmenting the ability of C-OP to function effectively.

This idea also extends to the lending platforms holding borrower collateral — defend against non-linear risk outcomes while maximizing capital efficiency.

Lending Platform Risk Management

Protecting lending platforms is a byproduct of two risk parameters imbued into C-OP:

Maximum Protection Composition (MPC)
Capacity to Free Post Liquidation (CFPL)

We covered precisely how MPC and CFPL work in our previous product issue, but will touch briefly on their advantages for lending protocols to complete our overview of risk management.

MPC allows a lending platform to calibrate their risk tolerance to derivatives in each particular loan. A higher MPC means the lending platform is willing to accept a higher composition of derivatives relative to crypto assets. Putting aside the market risk, there are non-zero, albeit very small, operational risks, such as the C-OP protection pool not being solvent or active at time of exercise. At this point, the lending pool may potentially be left with losses. This risk is not specific to UNION or DeFi. It is a risk of TradFi as well.

CFPL protects a lending platform’s risk to rapidly falling markets, which may trigger yet another liquidation in quick succession. In these markets, after liquidation, the protocol may still be left with losses to the account.

For example, in the case of a liquidation of Alice’s collateral, meaning her underlying collateral is experiencing downside volatility or she is over-leveraged, the market could continue selling off. If a liquidator frees Alice’s assets right up to flat account liquidity (i.e., CFPL), markets may move rapidly again to force another liquidation. With C-OP, relative to the lending platform’s overall risk, the fine-tuning of MPC enables the proper threshold adjustment that protects the lending protocol’s overall liquidity in the case of volatile conditions.

The C-OP protection pool does not affect the MPC or CFPL adjustments or execution by the lending platform. These are independent products.

To conclude, UNION’s capital pool for C-OP is dynamically adjusted to meet stringent MCR and SCR solvency standards while effectively allowing for maximal capital efficiency based on the overall health of the pool. Lending protocol risk for C-OP is managed via a dynamic adjustment of MPC to meet the liquidation threshold requirements of the platforms, which they can adjust internally — the C-OP protection pool does not execute protection on the lending platform.

This wraps up the issues explaining the math and economics — important, but admittedly, not the easiest reading. The next issue, we move into lighter topics, such as showing progress on the application!