So, what’s in a peer-to-peer lending platform anyway? Show me the data!
ETHLend is a fully decentralised peer-to-peer lending platform running on the Ethereum network utilising smart contracts: applications that run exactly as programmed without any possibility of downtime, censorship, fraud or third-party interference.
ETHLend Omura, our alpha release v0.2.1 provides a unique insight into the types of loan terms two complete strangers are willing to agree upon: there are no credit scores, no identification documents, no conversations, no interference by external parties or middlemen; only the safety of being backed by cryptocurrency as collateral, and the numbers.
So what do the numbers show us? Who’s lending, who’s borrowing, and how?
Let me start the story with the journey of volume since ETHLend has launched Omura. Since its initial release on the 9th of February we have seen the lending volume steadily increase to its current level of 1,293 ETH (around $760,000 at current prices) and $761,121 (ETH pegged to USD), combining to equal 2,224 ETH. The split of lending between ETH only loans and USD pegged ETH is relatively equal, however the average premium (ignoring loan length) offered by USD loans is significantly higher.
Since this analysis was done, the rate of ETH borrowed has only increased further, and we have now hit over 2500 ETH borrowed, taking the total loan volume including the previous version above 6,000 ETH!
The average loan length since Omura has launched has seen a definite upwards trend, this may reflect the increased confidence of lenders since the collateral calling feature has been introduced, this feature enables lenders to claim the tokens of borrowers if the Loan-to-Value ratio drops below 100%, providing a relatively risk free lending environment. Despite this, the most common loan length is still 30 days. I imagine borrowers will be more willing to set longer loan terms once the ability to refill collateral is introduced in our next release at the end of this month.
Loan length is just one of the potential variables that a borrower can set in the system, one of the most important variables is the premium/interest rate. So far, the average annual compounded interest rate has been 82.3%, this is huge, especially when compared to traditional forms of financing. This is an ideal environment for lenders providing large returns with relatively low risk. From a borrowers perspective, it seems less than ideal, however this is perhaps reflective of the high returns experienced in cryptocurrencies in general, combined with the high volatility. It’s also important to remember that this is a peer-to-peer environment where borrowers are freely offering these rates, the only restriction in the platform being that the return must be higher than 1% for 360 days, and less than 100% for 1 day. A general downtrend in compounding interest rates can be seen, potentially signalling that the current average is above the ideal equilibrium, even for cryptocurrency.
So, who is actually borrowing and lending?
Looking at the below map, we can see that around 50% of the loan volume (by ETH) has been generated by 17 borrowers. Likewise, around 60% of the loan volume (by ETH) has been funded by 9 lenders, with one lender in particular funding 35 loans, for a total value of 355.4 ETH.
The relationship between the number of loans funded and the total value of ETH lent has a reasonably high R-squared value of 0.7204242 indicating a relatively high correlation between the number of loans funded, and the total value of ETH lent. Conversely, the R-squared of the relationship between the number of loan requests and the value of ETH borrowed is relatively low at 0.118955. This indicates that the total value of ETH borrowed may not be particularly related to the number of loans requested.
Who’s lending to who? It’s easy to imagine that people who know each other forming contractual lending agreements on the platform, underpinned by the security of smart contracts. No lawyers or court needed if things turn sour, just the smart contract executing the rules that have been programmed. It’s hard to tell if this is already happening from the below, however we can see that largest number of times a borrower has been funded by the same lender is 7. Perhaps they know each other? Maybe it’s a coincidence? Feel free to reach out if you can enlighten us!
Now, what about collateral? What is popular? What is not? Are there differences between returns for different collateral?
Our own LEND token, is the most popular to be used on the platform with 737.9 ETH of lending volume, followed by ICX (Hello ICON World), and then OMG (OmiseGO). Some other notable high volume coins used as collateral include: GVT (Genesis Vision), VEN (VeChain Foundation), WTC (Waltonchain_EN), QSP (Quantstamp), and TRX (Tron Labs).
Finally, Is there a difference between different types of collateral, and the premium offered? It seems as though there is! VEN (VeChain Foundation) has the highest positive variance from the mean, coming in 11.25% above average. LINK (ChainLink), has the highest negative variance from the mean, coming in 3.9% below average. Our own LEND (ETHLend) token, in relatively close to the mean at 0.76% above.
Well, there it is. Hopefully this gives a bit of insight into how lenders and borrowers are coming together to borrow and lend in an anonymous, trust-less environment.
If you want to look into the data further, feel free to check it out here: https://public.tableau.com/profile/scott.waddell#!/