Loss Refunds [TESTNET]

Since the Omni Liquidity Provider is the counterparty to all trades, Omni is able to refund a portion of user losses. Loss refunds operate via a lottery system, with increases in odds based on reward tier and expected vs. actual refund frequency.

Loss refunds are funded by a portion of OLP's revenue, which is directed to a "Loss Refund Pool." Each time a user realizes negative PnL, they have a chance to immediately receive a complete refund of the realized loss or 20% of the Loss Refund Pool, whichever is lesser.

Tier
Loss Refund Odds Boost
True Base Loss Refund Odds

No Tier

1x

1%

Bronze

1.25x

1.25%

Silver

1.5x

1.5%

Gold

2x

2%

Platinum

3x

3%

Diamond

4x

4%

Grandmaster

5x

5%

Mechanism

  1. 1/6th of OLP's spread revenue is directed to the Loss Refund Pool.

  2. Each time a user realizes negative PnL, they have a chance to receive a refund.

    1. Base Refund Chance: 1%

    2. Reward Tier Multiplier: 1-5x

    3. Luckiness Multiplier: Expected Refunds ÷ Actual Refunds (if undefined, set to 2).

    4. Refund Chance = Base Refund Chance (1%) × Reward Tier Multiplier × Luckiness Multiplier

  3. If the user is awarded a loss refund, their payout is either 100% of the realized loss or 20% of the Loss Refund Pool, whichever is lesser.

Example

Example 1: Single RPnL event, Platinum tier.

  1. Alice's account is "Platinum" tier, since she's done >$100m volume over the past 30 days.

  2. Alice opens a $10,000 BTC long.

  3. Alice's trade moves against her, so she closes her position (in one clip) with a RPnL of -$1,000.

  4. Alice's base odds to receive a loss refund are 1% * 3 (Platinum multiplier) = 3%.

  5. It's also been a while since Alice's last refund; she's realized enough losses that her "expected refunds" are 4, but she has only received 1 so far. Alice's Luckiness Multiplier is 4 (Expected) / 1 (Actual) = 4.

  6. Alice's final loss refund odds for the trade are 1% * 3 (Platinum Multiplier) * 4 (Luckiness Multiplier) = 12%.

  7. Alice gets lucky! Her loss refund hits. Since there is currently $10,000 in the Loss Refund Pool, she receives either 100% of her loss ($1,000) or 20% of the Loss Refund Pool ($2,000) whichever is lesser.

  8. Since Alice's loss does not exceed 20% of the Loss Refund Pool, Alice receives a complete refund of $1,000 to her account.

Example 2: Multiple RPnL events, Gold tier.

  1. Bob's account is "Gold" tier, since he's done >$25m volume over the past 30 days.

  2. Bob opens a $10,000 BTC long.

  3. Bob's trade moves against him, so he starts closing out his position in $1,000 chunks. He ends up realizing $900 in losses across 9 RPnL events, with an RPnL of -$100 per event.

  4. Bob's base odds to receive a loss refund (for each RPnL event) is 1% * 2 (Gold Multiplier) = 2%.

  5. Bob is still pretty new to Omni, and has only gotten one loss refund before. However, he has realized enough losses that he was statistically expected to have gotten two by now. Bob's Luckiness Multiplier is 2 (Expected) / 1 (Actual) = 2.

  6. Bob's final loss refund odds for each RPnL event are 1% * 2 (Gold Multiplier) * 2 (Luckiness Multiplier) = 4%.

    1. Note that the luckiness multiplier updates after each of Bob's RPnL events. If, for example, 4 negative RPnL events into closing his trade, Bob's "Expected Refunds" went from 2 to 3, then his Luckiness Multiplier for the remaining 5 RPnL events would be 3 (Expected) / 1 (Actual) = 3.

  7. Bob gets lucky! One of his RPnL events gets a loss refund. Since there is currently $10,000 in the Loss Refund Pool, he receives either 100% of his loss ($100, since his $900 loss was split up across multiple RPnL events) or 20% of the Loss Refund Pool ($2,000), whichever is lesser.

  8. Since Bob's refunded loss ($100) does not exceed 20% of the Loss Refund Pool, Bob receives a refund of $100 to his account.

Last updated