Marketmaker incentives / MM Pool

#1

As a reaction to the recent work on MM and incentives. It would be good to see some more interaction with market-making in practice. My goal is to make the business of MM more widely available to anyone. The hurdles to understanding the actual work and the implementation is very high. I’ve been fairly active in the discord to discuss the implementation and strategies.

The goal of MM as a strategy is mostly similar to any trading strategy - high risk adjusted returns (sharpe ratio).

In my view the issue of MM relates to trading on 0x versus central exchange. 0x is an alternative to centralised trading, it doesn’t exist in isolation, but in a big marketplace with ~2B$/day traded.

The calculation for Profit function is roughly: risk adjusted profit = current volume * potential spread captured. Current volume is ~100k$ and capturing 0.20% is feasible. So that’s 200$/day in revenue. Given the higher integration cost its really maybe 50$/day in profit which is at best 20’000$ p.a. At current rates market-making for professionals is not feasible. It might be ok for some people to earn some money on the side with this strategy. With MM pool I am working on enabling more people staking smaller amounts to earn returns.

It would help to bring to more awareness the advantages on trading on 0x versus central exchange, ie. lower custodial risk. The more those are available to the public, the more will switch from central to decentral. Most users will compare the alternatives that are available to them.

1 Like
#2

Hi to be clear, we do not expect the taker fee and maker rebate program to be a near-term fix for this issue. At best, it will allow makers to capture a slightly larger percentage say 0.25% instead of 0.2%, while still quoting the same bid-ask spread. To really tackle this, we will need more direct and aggressive interventions. We are exploring a variety of options to do this. For example, hummingbot would be one way of putting idle capital of non-professionals to work.

2 Likes
#3

@0x_peter agreed, to lower spread is only one part. There could be work on taker-incentives, so potentially if someone has a website which enables trading he could capture a spread based on that traffic.

“Do you think folks are going to want to see historical profit results before joining the pool”?

generally yes, but one question is how to audit results. stakers def. assume some risks there. I’m assuming there is some MM reward which pool has access to, to cover costs, similar to MM program

in some schemes there is the ability to stake tokens which someone already has. in that case its like loaning out the tokens

1 Like
#4

Yes actually all the MM has to do is sign an agreement to give x% of rewards to people who stake him.

It is possible for the MM
a) first, sign agreement
b) second, perform trades
Then,
c) the stakers observe these trades and loan/rent/delegate stake to allow the MM to claim rebates,
The stakers are already certain to get a reward at that point. There is some risk that the MM will stop trading after they receive stake and the stakers will need to respond by movimg funds to another pool. There will be a 2-4 week latency associated with moving funds, so the staker would lose out on some income. That’s the extent of the risk.
In our experience, the set of addresses that do most MM’ing has been very consistent over time. I expect the risk for stakers to be minimal in practice,

We will build a user interface for this which should make things very intuitive for both stakers and makers. The user interface will show the MMs trading history.

The MM doesn’t have to do anything except sign the agreement. All the other work is done by stakers.

The payouts are all done through smart contracts. It is not possible for anyone to cheat, The incentive just offsets part of the cost of using the blockchain. It is not possible to cheat by wash trading because the reward would be smaller than the mining fees the wash trader pays to submit ETH transactions.

Finally, we do have some systems that allow websites to drive taker traffic and earn fees from doing this, Prices probably will need to get better before these systems become widely used.

#5

I see. MM pool would be similar, except it would be cross-exchange. So there needs to be some service which can calculate Profit-and-Loss statement for many transactions across many exchanges, which is not easy. It might also require third party audit of some kind. The great benefit of that would be this can serve the entire crypto market, so the addressable market is larger. I think doing all of this in just smart contracts is very hard. There is in any case trust required in terms of the execution of the MM operator.