In my analysis, I will analyze failed transactions on Sushiswap-Ethereum by answering questions:

• Which user action results in the highest rate of failed transactions?
• Compare swapping, borrowing, lending, and adding/removing liquidity on Sushiswap-Ethereum.

Introduction to Sushiswap

SushiSwap is a blockchain-based decentralized exchange (DEX). It serves as an alternative to the trading of liquidity provider (LP) tokens.

SushiSwap, which uses the Automated Market Maker (AMM) protocol, is enabled by "smart contracts," allowing users to purchase and sell via Sushi DeFi and decentralized exchanges.

There are several things you can do on SushiSwap:

• Swap: Trade one cryptocurrency for another, such as turning your USD Coin (USDC) into Ethereum (ETH).
• Farm: Deposit crypto into one of SushiSwap's liquidity pools to earn rewards. Liquidity pools contain combinations of two or more cryptocurrencies, and you must deposit an equal amount of each one.
• Stake: Deposit your SushiSwap tokens to earn rewards.
• Lend: Provide crypto for others to borrow and earn interest on it.
• Borrow: Borrow crypto after providing collateral.

Methodology and data

• The source of data: The table “==ethereum.core.fact_transactions== ”, “==ethereum.sushi.ez_swaps==”, “==ethereum.sushi.ez_borrowing==”, “==ethereum.sushi.ez_lending==” and “==ethereum.core.dim_dex_liquidity_pools==” are used to conduct my SQLs.

  \

• Approach to analyze: In fact, there are actually 4 tables where the data is specialized for each type of activity on the Sushiswap platform: swap (ethereum.sushi.ez_swaps), borrow (ethereum.sushi.ez_borrowing), lend (ethereum.sushi.ez_lending), add/remove liquidity (ethereum.core.dim_dex_liquidity_pools). However, there is no available data about the status of transactions on these tables. This is the reason why I use“ethereum.core.fact_transactions”. The key field “status” is to identify the successful and failed status of transaction.

THE NUMBER OF ALL TRANSACTIONS

> (All transactions = successful transactions + failed transactions)

Looking at the charts, it can be seen that swap had the highest number of transactions (284M transactions). Its percentage in total transaction volume was 97.5%. There was a big gap in the number of all transactions between swap and the rest. Lend and borrow constituted nearly the same percentage in total transactions (~ 1.3% in each). Add/remove liquidity was the action with lowest number of transactions (65K transactions).

Let’s deep dive in the daily/weekly/monthly fluctuation in the number of all transactions over time:

As can be seen, swap had the decrease in the total number of transactions. Despite the low number of transaction when compared to swap, borrow and lend had the growth in the transaction volume while add/remove liquidity decreased significantly in its total transaction volume.

THE NUMBER OF FAILED TRANSACTIONS

Swap accounted for a very large proportion (93%) in total failed transaction volume. The gap was still huge between top 1 and the rest. However, regarding borrowing activities, when we compared this percentage with its percentage of total transactions, there was a big increase (from 1.24% to 5.51%). This is a sign that the execution of transactions for borrowing activities is not going well. When this comparison is applied to lending activities, we can see that there is no big change.

Let’s deep dive in the daily/weekly/monthly fluctuation in the number of failed transactions over time:

Swap and borrow had the climb in the number of failed transaction over 365 days. Lend had the downward trend in this number whereas its total transaction volume increased. Due to the nature of the activity, add/remove liquidity, it has no continuity in transaction execution. Its number of failed transactions rose and fell erratically. 

PERCENTAGE OF FAILED TRANSACTIONS

As a result, over the past 365 days, user action which results in the highest rate of failed transactions is borrow. Its overall percentage of failed transactions was 11.6% while this percentage of rest was lower than 10%. As can be seen from daily, weekly and monthly change, I believe that swap and lend had the good operation in executing transactions. Their failure rates were almost always less than 5%. Other actions including borrow and adding/removing liquidity need the enhancement in their operation. Although the transaction failure rate of borrowing activities has decreased recently, in general, this rate was still high compared to the beginning of the year. In term of adding/removing liquidity, the quality of performance was not stable over time.