mev wif me: a story about mev, solana, and a dog wif a hat.

On January 10 2024 at 22:20:39 UTC, 5qYuZ9ZLShLB1MuV83xHRcTgVA9A5pUajnQUUcPbk3bf (zer0xtrading.sol) made an extremely costly mistake that transferred millions of dollars from their hands into the hands of arb bots, Jito validators, and their stakers.

In one transaction, they swapped 86,739.1 SOL, worth approximately $8.6M USD, for 17,225,407.03 of $WIF, a popular memecoin on Solana. Unfortunately for them, liquidity was low. During their swap, they took most on-chain liquidity, moving the price up 2800% from around $0.14 to $3.99. Let’s cover what ensued next, as the Solana arbitrage bots on Solana swarmed to capture the arbitrage and make millions of dollars. 

What Happened?

In slot 241056629, zer0xtrading.sol sent a transaction that caused a large price impact on the WIF markets on Solana. A trader, 2Fast, leveraged Jito’s block engine to send a bundle which atomically captures this arbitrage. This bundle was executed on Figment’s validator. 2Fast ended up wif a $1.9M profit, tipping a total of 890.42 SOL ($89,000) to Figment’s validator and stakers. 2Fast was capital constrained and ended up not capturing the full arbitrage. In the same block, pUEsyy sends a transaction that arbitrages WIF, going SOL -> BONK  -> WIF -> SOL. He turns 6.4 SOL into 132.81 SOL. In the next several blocks, a swarm of bots rush in to capture the arbitrage.

In the 50 blocks (24 seconds) following the arbitrage for any WIF related transactions, the following took place:

• 2,286.5 SOL tipped to Jito validators via bundles
• 8.53 SOL fees to validators from 50% base fee and priority fees
• 13% of total compute units during this time were spent processing WIF-related transactions
• 54% of WIF compute units were from failed transactions
• 44% of WIF related transactions failed
• The top three earning validators were:
  • Validator.com: 978.95 SOL
  • Figment: 890.42 SOL
  • Staking Facilities: 403.64 SOL

I’m going to skip mentioning every single arbitrage transaction that took place in those 50 blocks, as there’s more than 500 transactions touching the WIF token. If you want to examine that closer, take a look at this spreadsheet generated wif this code. I highly recommend looking through the data as there’s more than 572 WIF transactions across 69 unique searchers and several searchers made more than one hundred thousand dollars. Let’s go through a few charts showing what happened at a high level, as it ended up being extremely profitable for those traders, Jito-Solana validators and their stakers.

A Closer Look

In Figure 2 chart, the number of bundles, which only land on-chain if they succeed vs. non-bundles success and failures are shown for each slot. Comparing the blue vs. red (sorry for the color blind), we can see an overwhelming amount of searchers leveraging bundles were quick to respond. There are a few gaps in here where the leader isn't running Jito-Solana, so no bundles landed. As time progressed, it became obvious that more bots had picked up on the opportunity, landing an increasing number of transactions on WIF markets.

Figure 3 contains a breakdown of each block’s compute units. It’s interesting to note the differences in compute units across the validators and the amount of compute dedicated to executing the WIF transactions as a function of all non-voting transactions. In fact, around 13% of total non-voting compute is spent on WIF related transactions in these blocks.

Here’s the juicy part. Across ~50 slots, more than 2,286.5 SOL was tipped in WIF transactions, generating validators and their stakers ~$228,650 in the span of 24 seconds. The top validators from this were:

• Validator.com: 978.95 SOL
• Figment: 890.42 SOL
• Staking Facilities: 403.64 SOL

Interestingly, Staking Facilities luckily had two leader slots during this time span: 241056636 - 241056639 and 241056660 - 241056663. During their first chunk of slots, they were able to capture 170.68 SOL. In the second chunk of slots, they captured 232.96 SOL.

Wrapping Up

I hope you enjoyed the article wif a summary. As a reminder, you can find the data used in this article here!

MEV has rapidly increased on Solana and this is the perfect example how a small number of transactions can generate an extremely large amount of revenue for MEV searchers, validators, and stakers. Situations like these can be prevented too. In the future, anyone looking to make a large trade like this should ensure to understand the price impact of their trade. If your trade results in a large price impact, spread it out over many smaller buys or sells and let arb bots handle the price imbalances that may be generated.

If you’re a Solana MEV searcher looking to get an edge, register for a free API key here and stop by Jito’s discord for help from the community of searchers on Solana. More than 600 searchers are using the system today!

If you’re a validator and not running Jito-Solana, you could be missing out on additional revenue! MEV on Solana validators has significantly increased in the last four weeks. Switch over in less than 10 minutes to join 49% of the network and 473 validators running the client today.