Flamingo Lend Targeted Incentives

In this section we describe the staking contract and targeted incentives of Flamingo Lend.

Abstract

We present a possible solution to ensure the growth of synthetic assets on Flamingo Lend by implementing targeted incentives. The proposed targeted incentives consider users' risk and willingness to borrow synthetic assets, and we compared them against traditional incentives found in the Flamingo platform today. The success of FUSD is more significant than expected, thus creating more demand for FUSD, and it is the catalyst for making this possible solution.

Introduction

The Flamingo Lend protocol allows the creation of synthetic assets such as the stablecoin FUSD. Synthetic assets can be created (or minted) by borrowing against other assets used as collateral. To ensure that the market cap of the synthetic assets created in the platform grows, the right set of incentives needs to be in place; we call them targeted incentives. The targeted incentives reward users based on their willingness to a.) borrow synthetic assets to increase their total supply (minting) and b.) participate in the automated selling and buying of said assets (automated market-making). With traditional incentives for providing liquidity in automatic market making, we give little consideration to the risks involved in borrowing against collateral to add liquidity because the same rewards can be reaped by simply buying the synthetic assets in the market.

Targeted Incentives For Synthetic Assets on Flamingo Finance

If we provide the right incentives, we make the condition for steering liquidity in the desired direction. In the case of the Flamingo Lend protocol, the desired direction would be highly liquid synthetic assets. There must be a lot of synthetic assets borrowed and used in market-making to have a highly liquid synthetic asset. Therefore we can break down the incentive into two parts to achieve the desired direction: incentivized market-making and incentivized borrowing.

Incentivized market making is what is usually known as staking rewards on DeFi platforms. It rewards liquidity providers by locking assets in a liquidity pool that will be used for automated market making. Incentivized market-making works perfectly fine in conditions where the only desired goal is to make a market liquid.

We can describe the traditional staking reward algorithm on Flamingo as follows:

$$P_t = H_t - H_{t-1}$$

where the $H_t$ represents the accumulated profit of per LP token at the given index of time $t$, which represents a point in time for every settlement.

We can see that all liquidity is treated equally and so the profit for a user's account at a given time is:

$$P^u_t = (H_t - H_{t-1}) \times l^u$$

where $u$ in this article serves as the index of a user, and $l^u$ amount of liquidity the user has staked.

The traditional staking reward algorithm is simple, but is not capable of giving rewards to different types of users like we want to achieve with the targeted incentives. Therefore, we propose the following staking reward algorithm for targeted incentives:

To reward both borrowers and liquidity providers we:

1. allow users to virtually stake a percentage of their borrowed assets to a pool

2. create a bookkeeping unit called $s^u$ to weigh in with the LP token amount defined by $l^u$ in the traditional staking reward algorithm.

First, we define the totals:

$$L = \sum{l^u}$$

$$B = \sum{b^u}$$

where $u$ in this article serves as the index of a user, $L$ is the total liquidity staked, and $B$ is the total of virtually staked borrowed assets.

For a single user, we can define $b^u = xy$ where $x$ is the percentage of the virtually staked borrowed assets, and $y$ is the total amount of borrowed assets the user has. For example, if a user has borrowed 1,000 FUSD and virtually staked 50% to a liquidity pool, then $b^u$ of the user would be $500$ by doing the calculation: 50% * 1,000 FUSD = 500 FUSD.

We then define the share of rewards $s^u$ for a user as:

$$s^u = \sqrt{(b^u)^2l^u}$$

the sum of $s^u$ as:

$$S = \sum{s^u}$$

The new variable $s^u$ is then used a new reward formula for a single user in at the given index of time $t$:

$$Q^u_t = (W_t - W_{t-1}) \times s^u$$

where the $W_t$ represents the accumulated profit of per share of rewards $S$ at the given index of time $t$, which represents a point in time for every settlement.

Finally, we combine the new reward formula for $Q^u_t$ with the traditional formula for $P^u_t$, so we also can give some base reward to those with zero borrowed assets:

$$R^u_t = P^u_t + Q^u_t$$

We can then weight the rewards for a pool. For example, $Q^u_t$ gets 80% of the rewards, and $P^u_t$ gets 20% of the rewards.

With the new algorithm in place, we can steer the liquidity in the desired direction while allowing users to choose their staking strategy. With the targeted incentives, synthetic assets can grow beyond what they can with the traditional model. The growth of the synthetic assets would help the whole Flamingo Finance platform and its native token FLM succeed.