THORChain Lending — Unveiling Terra LUNA’s Shadow
Introduction
Upon deep examination of the lending module introduced by Thorchain on August 22nd, we have identified a shadow of Terra LUNA. The similarity with LUNA is chiefly manifested when users’ collateral is converted into RUNE. Essentially, it’s the fluctuation of the RUNE-collateral exchange rate that determines RUNE’s inflation or deflation. In this process, RUNE absorbs the volatility of its exchange rate with collateral, similar to how LUNA absorbs the volatility of UST. However, the manner in which they operate (RUNE being part of the lending process, being destroyed and minted when loans are opened and closed, and LUNA being anchored to a stablecoin, being destroyed and minted by arbitrageurs when UST is off-peg) and the underlying risk (LUNA having unlimited minting, whereas RUNE has an inflation/deflation cap and only 50% of the synthetic asset’s collateral is in RUNE) are different. Additionally, strict risk control and isolation measures are in place for the lending protocol, ensuring overall minimal risk. This means it won’t produce systemic risks akin to Terra LUNA, and even if a negative spiral occurs, it won’t affect other Thorchain functions.
I. Understanding Thorchain Lending Mechanism
A notable feature of Thorchain lending is the absence of interest, liquidation risk, and a time limit (initial phase has a minimum loan period of 30 days). For users, it essentially means collateralizing assets like BTC/ETH to borrow against USD; for the protocol, it’s essentially shorting BTC/ETH and going long on USD. The debt is priced in TOR (Thorchain’s USD equivalent), hence, for users, it’s akin to purchasing out-of-the-money call options on BTC in a gold standard system, with the protocol/RUNE holders as the counterparty.
Initiating a new loan results in a deflationary effect on $RUNE assets, while closing a loan leads to an inflationary impact. BTC collateral is first exchanged for RUNE, then destroyed, and finally, RUNE is minted to exchange the required assets. In this process, the net value of RUNE destroyed is the difference between the collateral value and the debt, excluding transaction fees.
If collateral value rises at the time of repayment and the price of RUNE remains constant, more RUNE needs to be minted to exchange the required assets, leading to inflation. If the price of RUNE increases, it’s ideal as fewer RUNEs need to be minted. If RUNE’s price drops, the inflation will be more severe. If collateral value decreases at repayment and the price of RUNE remains the same, users might opt not to repay (no minting required).
If, between loan initiation and closure, the value of RUNE relative to $BTC remains constant, there won’t be a net inflationary effect on $RUNE (destroyed amount will be equal to minted amount minus transaction fees). However, if the value of the collateral relative to RUNE increases between loan initiation and closure, a net inflation will occur in $RUNE supply.
To address inflation, lending control measures are in place — if minting results in a total supply surpassing 5 million RUNE, a circuit breaker is designed. In this scenario, reserves will intervene to redeem the loan (instead of further minting), the entire lending scheme will cease and be withdrawn, but other aspects of THORChain will continue as usual.
Thus, the lending process has a significant impact on RUNE’s inflation and deflation, but both inflation and deflation have caps in a scenario where the overall lending cap is low. When the RUNE-collateral exchange rate rises indefinitely, the maximum deflation is 15mln*0.33 (0.33 is the lending lever, which might change), which equals 4.95mln (and might increase in the future). If the RUNE-collateral exchange rate falls indefinitely, inflation is limited to 5mln by the circuit breaker.
Specifically, if a user over-collateralizes by 200% and borrows 50% of the required asset, the other 50% at redemption is minted based on the RUNE-collateral exchange rate. This step is fundamentally similar to LUNA, but under Thorchain’s lending mechanism, since only 50% is backed by Rune and the product’s capacity is relatively small, the overall risk is minimized. It won’t generate systemic risks like Terra LUNA, and the risks are isolated, meaning a negative spiral won’t affect Thorchain’s other functions.
1. Understanding how the lending design is akin to a deep out-of-the-money, resettable strike price call option for users:
When Alice collateralizes 1 BTC, she also obtains 50% in cash (assuming a 200% CR) and the opportunity to buy 1 BTC with that cash. If BTC appreciates by the time she repays (let’s say a month later), Alice repays the debt (equivalent to 50% of BTC’s value a month ago) and buys that BTC at the price from a month ago. If it has depreciated significantly, by more than 50%, Alice might choose not to repay, so the protocol won’t mint RUNE, leading to inflation (for Alice, her bullish position failed).
2. Understanding the Absence of Borrowing Interest
It can be perceived as users paying multiple swap fees instead of an interest rate, essentially, it’s a CDP (Collateralized Debt Position) product. Charging an interest rate would make this product less attractive to users.
The entire lending process is as follows:
Users deposit native assets as collateral (BTC, ETH, BNB, ATOM, AVAX, LTC, BCH, DOGE). Initially, only BTC and ETH are accepted as collateral. How much collateral each debt vault can accept is determined by the hard cap (15 million), lending lever, and pool depth coefficient. Over-collateralization creates debt, and the proportion of debt that can be obtained is determined by the CR (Collateral Ratio).
Borrowing: Alice deposits 1 BTC. This BTC is first swapped for RUNE in the BTC-RUNE swap pool. These RUNEs enter a V BTC pool where they’re destroyed and converted into a derivative asset, Thor.BTC. The collateral for this synthetic asset is constant product liquidity, always 50% of the asset, with the remaining 50% being RUNE. The Thor.BTC is then transferred to an Internal module where a dynamic CR determines how much loan can be obtained, and a Thor.Tor token (similar to USD) is generated for accounting purposes. All these steps are for internal accounting. Subsequently, a USDT loan is generated and given to Alice for her use.
Loan Repayment: When Alice repays, she sends all USDT or other Thorchain-supported assets to the protocol, which are converted into RUNE. RUNE creates Tor, and the protocol checks if the user has repaid all loans priced in Tor. If fully repaid, collateral is released, converted into derived collateral (Thor.BTC), which is then recast into RUNE, and finally swapped back to L1 BTC. RUNE is minted during this process.
It’s important to note that these swap and convert processes generate fees (at least four swap fees for a single loan). Even though there’s no interest, this fee accumulation can be considered an alternative to interest. Despite significant wear, the RUNE fees are destroyed, which leads to genuine deflation.
3. Understanding the Absence of Liquidation and No Repayment Time Limit
Since the debt priced in TOR stablecoin is fixed, although borrowers can choose any asset for repayment, they’ll eventually convert it to RUNE through the market. Liquidity providers and depositors don’t directly lend their assets to borrowers. The pool merely acts as a medium for swaps between collateral and debt. This process is essentially a wager, hence the absence of liquidation. The protocol requires sufficient RUNE to repay enough TOR (complete repayment) to help users retrieve their collateral. If the collateral price drops significantly and the user chooses not to repay, this portion of RUNE won’t be reminted, resulting in net destruction. The protocol, in fact, prefers users not to repay, as repaying when collateral prices rise and RUNE prices drop would lead to inflation.
4. Understanding the Deflation and Inflation of RUNE as a Trading Medium
Firstly, the total cap of all lending pools is determined by the RUNE Burnt segment in the following graph multiplied by the Lending lever. The 15 million RUNE Burnt resulted from the protocol burning non-upgraded BEP2/ERC20 RUNE. Therefore, the protocol currently has 15 million headroom for inflation from the maximum supply of 500 million RUNE.
RUNE’s role in the entire borrowing process was previously discussed. Initiating a new loan creates a deflationary effect on RUNE, while closing a loan inflates RUNE. If the collateral appreciates at repayment and RUNE’s price remains unchanged, more RUNE needs to be minted to exchange the required assets, leading to inflation. Ideally, if RUNE’s price rises, not as much RUNE will need to be minted. If RUNE’s price falls, inflation becomes more severe. If the value of collateral drops at repayment and RUNE’s price remains the same, users might opt not to repay (no minting).
If, between loan initiation and closure, RUNE’s value relative to BTC remains unchanged, then RUNE won’t experience net inflation (the number destroyed equals the number minted minus swap fees). However, if the collateral’s value relative to RUNE increases between loan opening and closure, RUNE supply will undergo net inflation.
To address the inflation problem, lending controls are in place — if minting causes the total supply to exceed 500 million RUNE, there’s a circuit breaker in the design. In this scenario, reserves will intervene to redeem loans (rather than further minting), halting the entire lending design, though other aspects of THORChain will continue to operate normally.
Based on the parameters in the graph, currently, all the debt vault pools combined only amount to 4.95 million RUNE. This means all debt vaults can only accept collateral equivalent to 4.95 million RUNE.
Total available Rune for protocol = Lending lever * Rune Burnt（4.95mln=15mln*33%）
If Rune price=1.5, network can accept up to 7.4mln(1.5*4.95mln) of collateral for lending
The entire Reserve’s RUNE Burnt serves as the buffer for all debt vaults and is the last resort in the face of inflation. The total amount of RUNE Burnt in the Reserve multiplied by the Lending lever (currently 4.95mln) is distributed according to the depth of each debt vault pool. The deeper the vault, the more Reserve buffer it receives. For instance, if the depth of the BTC Lending pool is twice that of the ETH Lending pool, the value of Rune Burnt * Lending lever * depth coefficient determines the maximum collateral limit this lending pool can handle. Therefore, as the price of RUNE rises, the more collateral this pool can accommodate. It’s evident that both the Lending lever and the price of RUNE determine the maximum collateral a lending pool can hold.
THORChain protocol and all RUNE holders act as counterparties to each loan. The burn/mint mechanism of RUNE means that RUNE condenses/dilutes when a debt is opened and closed among all RUNE holders. When the RUNE-collateral exchange rate falls, it leads to inflation, and conversely, deflation.
5. Is the CDP protocol a good on-chain savings model?
For the Lending launched by Thorchain, it acts as an indirect saving method, making RUNE an indispensable medium in the lending and repayment process, expanding the burn and mint scenarios.
So, is this savings model advantageous? Let’s compare it to other savings models.
CEXs (Centralized Exchanges) are the most obvious beneficiaries of savings models. As custodians, these funds often yield more returns (though these returns have decreased after requiring reserves to be made public). Protecting users’ custodial funds is a regulatory focus, with regulators typically expecting exchanges to fully reserve.
The on-chain scenario is entirely different.
For DEXs (Decentralized Exchanges), high incentives must be given to LPs (Liquidity Providers) after savings, so the goal is to increase liquidity. The funds provided by LPs aren’t directly used for profits; instead, they form a moat of liquidity through large reserves.
Pure Lending platforms like Aave or Compound have to bear interest costs for savings. The entire model is similar to traditional lending, requiring active management of loan positions and having repayment time limits.
In contrast, the CDP (Collateralized Debt Position) savings model is healthier. Due to the high volatility of collateral assets, most over-collateralized CDPs on the market involve users over-collateralizing certain assets to obtain stablecoins/other assets. In this process, the CDP protocol effectively gains more “deposits” and doesn’t need to pay interest on them.
Thorchain also falls under this CDP model. So where is the collateral held? In fact, collaterals are swapped for RUNE through liquidity pools. Hence, no one “stores” the collateral. As long as the THORChain pool operates normally, any deposited collateral will be exchanged for RUNE, and arbitrageurs will rebalance the pool as usual. This can be viewed as collateral settling in Thorchain’s RUNE pairs with other currencies. Because assets like BTC are in the circulating market rather than held in the protocol, the difference in value between the collateral and the debt is determined by RUNE’s value, casting a shadow similar to Terra LUNA.
Capital Sink might be one of the goals Thorchain lending aims to achieve, using user collateral to settle as liquidity in the swap pool. As long as users don’t close the loan and the price of RUNE doesn’t plummet, the protocol retains assets, RUNE undergoes deflation, forming a positive cycle. Of course, the opposite would result in a negative spiral.
6. Risk
BTC and other collaterals enter the open market instead of being held within the protocol. Thus, even though the generated debt is 100% collateralized, the difference between the value of the collateral and the debt is determined by the price of RUNE. This mechanism shadows similarities with Terra LUNA. Since the RUNE burned when initiating a loan and the RUNE minted when closing a loan might not be exactly equal, both deflationary and inflationary scenarios can arise. It can be interpreted that when the price of RUNE increases at repayment, deflation occurs, and vice versa for inflation. If the RUNE price drops below the lending lever multiple of the price when opening a position, a circuit breaker will be triggered. Throughout the lending process, the price of RUNE plays a decisive role in inflation and deflation. When the RUNE price declines and many users choose to close their loans, there’s a high risk of inflation. However, the protocol has stringent risk control and risk isolation measures, hence the overall risk is relatively small. Systemic risks like that of Terra LUNA won’t occur, and even if a negative spiral happens, it won’t affect other functionalities of Thorchain.
Lending lever, CR, and the decision to open different collateral debt vaults form the three pillars of Thorchain lending’s risk control.
Moreover, Thorchain has a history of being hacked, and its code is complex. Thorchain Lending might also have vulnerabilities that need pausing or fixing.
II. Conclusion
The introduction of Thorchain Lending creates network synergies, additional trading volume, and higher pool capital efficiency, bringing genuine profits to the system and increasing the total bonded amount. This allows Thorchain to gain potential upward space by reducing the circulation (especially when the RUNE-collateral exchange rate rises).
The Capital sink (or the goal Thorchain lending might want to achieve) uses users’ collateral to provide liquidity in the swap pool. As long as users don’t close their loans and the price of RUNE doesn’t drop significantly, the protocol retains assets, and RUNE undergoes deflation, forming a positive cycle.
However, the opposite market trend causing inflation and a negative spiral can very well happen. To control the risk, the use of Thorchain lending is limited with a small capacity. The effects of inflation and deflation on the current capped volume won’t fundamentally affect the price of RUNE (at most affecting 5 million RUNE).
For users, Thorchain’s capital efficiency is also not high, with CR ranging between 200%-500%, likely fluctuating between 300%-400%. Purely from a leverage perspective, it’s not the best product. And while there’s no borrowing fee, the multiple internal transaction fees are not user-friendly.
Evaluating just the lending product doesn’t represent the overall development of Thorchain’s defi product matrix. Subsequent analyses of other Thorchain products will follow.
LD Capital is a leading crypto fund who is active in primary and secondary markets, whose sub-funds include dedicated eco fund, FoF, hedge fund and Meta Fund.
LD Capital has a professional global team with deep industrial resources, and focus on develivering superior post-investment services to enhance project value growth, and specializes in long-term value and ecosystem investment.
LD Capital has successively discovered and invested more than 300 companies in Infra/Protocol/Dapp/Privacy/Metaverse/Layer2/DeFi/DAO/GameFi fields since 2016.
website: ldcap.com
twitter: twitter.com/ld_capital
mail: BP@ldcap.com
medium：ld-capital.medium.com
