Cauldron Arbitrage Statistics

Technical
#1

I have been gathering some statistics regarding the arbitrage that happens on the Bitcoin Cash network for extracting value from unbalanced Cauldron pools.
This is a static page that I might be updating going forwards:
https://robchank.github.io/arbstats/

Some notes:

  1. The patterns suggest several bots are doing arbitrage and competing for the opportunities. This results in mempool de-synchronization and will get worse the more bots that are deployed.
  2. Miners are sleeping on this. This is MEV left on the table for user bots to take profit. Miners can also do it more cheaply since there is no need to create an output for collecting the profits.
  3. Wallets need to be smart and use as many pools as economically feasible to keep the number of arbitrage opportunities low.
4 Likes
#2

Hmmmmmm very very interesting.

We’re going to need to try be on the front foot with MEV as much as we can. It’s ultimately a sign of success, a growing ecosystem will create more opportunities for MEV because we’re growing, but that also raises the bar for us to address it and minimise it as much as we can.

1 Like
#3

Wallets need to be smart and use as many pools as economically feasible to keep the number of arbitrage opportunities low.

This is an interesting dilemma. It is better for the health of the network to use fewer UTXOs to minimize contention over any single pool. It is better for the user to use the set of UTXOs that results in the best trade for the user, which is possibly not the minimum number of pools. And it is better for the health of the market to use as many UTXOs as possible to minimize arbitration.

Intuitively I would believe that getting the user the best price would also be the best for minimizing arbitration, and that network health is reliant on a well-saturated collection of pool UTXOs.

1 Like
#4

I’ve always had the belief that MEV on BCH is actually a good thing because it incentivizes miners to execute contracts, given that they cannot be executed autonomously. As I understand, it is more problematic on account-based chains due to frontrunning/sandwich attacks, whereas it’s basically a non-issue for UTXO chains due to their transactions’ atomic nature. Do you know if this is accurate?

2 Likes
#5

Sandwich attacks are not relevant to UTXOs because they’re not ordered, and that’s always the main example used for MEV (I should actually research what other kinds of MEV there are, I’m sure there must be some), but it’s possible there is an equivalent for UTXO chains. We are actually going to be on the forefront of finding out, because we have the most advanced DeFi ecosystem of any UTXO chain afaik.

MEV appearing overall is a good thing definitely, it shows that we have an active ecosystem with lots of opportunities and of course it will boost transaction volume and mining rewards paid for by the inefficiencies of the ecosystem itself.

1 Like
#6

I copied this exact comment into Grok, here is the most relevant part of its answer:

Where MEV still exists and can matter on BCH

It is not entirely a non-issue:

  • Stale-state arbitrage is possible. An adversary can try to double-spend / compete against an unconfirmed chain to interact with an older, more favorable price state on a DEX or oracle-dependent contract, and pay a higher fee (“miner bribe”) to get their version preferred. This is different from sandwiching but still a form of MEV.

  • There are real, enumerated MEV opportunities today (see the “catalogue of Miner Extractable Value on Bitcoin Cash” threads on bitcoincashresearch.org).

  • If DeFi volume grows significantly, specialized block construction, competition over high-value settlements, and pressure on first-seen conventions could increase. Some designs (e.g., Jedex-style joint/batched execution) exist specifically to mitigate this.

  • Impact is mostly isolated to the DeFi UTXOs involved rather than the broader network, thanks to the UTXO model’s parallelism — ordinary P2PKH payments are largely unaffected.

In short: the architectural claim is solid. Classic user-hostile sandwiching and easy frontrunning of sequential state changes are much harder (and often impossible) on BCH’s UTXO model compared with account-based chains. The forms of MEV that do exist are frequently intentional incentives for contract execution or limited forms of arbitrage/competition, rather than the systemic “invisible tax” seen on Ethereum-style DEXes. The severity and character of the problem are different.

1 Like
#7

The problem with MEV at scale in this scenario from my perspective is unconfirmed chains of swaps would be broken leading to bad UX. In the case of bots competing for profits one could theoretically deploy multiple nodes to detect races and make a statistical guess on the most probable tip to be mined. A pragmatic miner would look at all swaps in the mempool to determine where to take most profits and invalidate any descendants.

This is partly mitigated by lowering the block times but Cauldron can not be seen as a zero conf protocol at the scale we all want to see.

2 Likes
#8

The user exceptions of unsettled trades, from cauldron or elsewhere, can be better managed with UI.

Instead of treating trades as settled, wallets and apps should convey the transactions do not satisfy zero-conf conditions and are subject to contention in mempool.

1 Like
#9

True.

But it seems to me everything will be fine. It works ok now at low volume. As scale arrives, both more DEXes will spin up to spread the load (see already GuanacoDEX) and the pressure of increased volume will drive new solutions (it’s almost guaranteed that there is some clever way with CashTokens to reserve UTXO threads & process them in an efficient and non-contentious way, particularly with AI available to consult on such designs) plus bring the resources to get them researched and built.

So 0 conf DEX performance is not a major problem yet, and by the time it is we’ll be in a strong position to solve it.