Mango Markets
Searchโ€ฆ
Liquidity Mining

Overview

Mango Perps are the first perpetual futures contracts that are traded on a fully on chain order book. Providing liquidity on an order book is more of a skilled activity than providing on an AMM contract. In order to bootstrap the liquidity on Mango Perps, the Mango Protocol will disburse MNGO automatically to market makers according to an on-chain algorithm. This document describes the mechanism. But broadly:
  • Quoting closer to the top of book is rewarded much more generously
  • You do not have to be filled to get the rewards (you get points every time you cancel OR get filled)
  • There are open source market making bots developed by and used by Mango devs

Formula

The liquidity mining formula rewards the top N contracts on the bids and asks weighted quadratically to favor those closer to top of book. Your depth on the book is assessed when you place an order on the book and when you get filled/cancel the order and the worse of the two is used in calculating your liquidity mining points.
Here's the abridged formula:
1
// Definitions:
2
// max_depth - the max number of contracts rewarded in this PerpMarket
3
// depth - how many contracts are ahead your order on the bids or asks
4
// quantity - quantity of the order or the
5
โ€‹
6
let depth = max(depth_initial, depth_final)
7
let depth_factor = max_depth - depth
8
โ€‹
9
if depth_factor <= 0 {
10
return;
11
}
12
โ€‹
13
// time the order was on the book in seconds
14
let time_factor = time_final - time_initial
15
โ€‹
16
// limit to rewarding the amount of your order that remains in the window
17
let quantity_factor = min(quantity, depth_factor)
18
โ€‹
19
let points = depth_factor.pow(2) * time_factor * quantity_factor
20
โ€‹
21
// Convert the points into MNGO; adjust the rate of disbursement
22
return convert_points(points)
23
โ€‹
Copied!

Example

Suppose max_depth = 20000, mngo_per_hour=250 and the bids look like this:
1
1000 MNGO @ 0.30
2
5000 MNGO @ 0.28
3
10000 MNGO @ 0.26
Copied!
Suppose I place a bid for 8,000 MNGO at price 0.27 at time t=100. Then I get filled at time t=110. Then:
1
depth_initial = 6000
2
depth_final = 0 // since there are no bids ahead of me if I got filled
3
time_intial = 100
4
time_final = 110
5
quantity = 8000
6
โ€‹
7
depth = max(6000, 0) = 6000
8
depth_factor = 20000 - 6000 = 14000
9
time_factor = 110 - 100 = 10
10
quantity_factor = min(8000, 14000) = 8000
11
โ€‹
12
points = 14,000 ^ 2 * 10 * 8000 = 1.568e13
13
โ€‹
Copied!
The points earned are 1.568e13 which is a totally meaningless number until it is normalized using a rate parameter and adjusted constantly such that a total of 250 MNGO per hour is distributed.

Rate Adjustments

This mechanism was inspired by the Bitcoin block rewards and difficulty mining adjustments that happen every 2016 blocks. In Bitcoin the target for 2016 blocks is 2 weeks. If it took longer than 2 weeks, then difficulty is adjusted down proportional to the time. If it took less than 2 weeks, difficulty is adjusted up. Similarly, we have a target time of 1 hour with a reward of 1,000 MNGO. When 1000 MNGO is distributed in a period, the time it took is compared against the target time of 1 hour. The rate is adjusted down if it took less than 1 hour and it's adjusted up if it took more than 1 hour.

Motivation

This formula comes out of the intersection of our goals and technical limitations. Here are the goals:
  • trustless and decentralized - should work even if initial devs disappear
  • open and transparent - no special market maker agreements
  • liquidity - tight spread, thick book that rivals centralized exchanges, even if volume is low
  • easy to access - deploying the mango markets reference market maker should be profitable

Liquidity Mining Specs

Max Depth is the amount of liquidity on *each* side of the book that is rewarded.
Perp Contract
Max Depth
MNGO per hour
MNGO
500,000 MNGO
250
BTC
8.3 BTC
1000
ETH
62.5 ETH
500
SOL
2,500 SOL
1000
SRM
18,000 SRM
250
RAY
12,500 RAY
250
FTT
2,100 FTT
250
ADA
125,000 ADA
500
โ€‹
Last modified 1mo ago