EIP-2: A more elegant way to deal with Coupon and Dept

EIP-2: A more elegant way to deal with Coupon and Dept

Authors: Reyer Chu

This document proposed a more elegant way to deal with Coupon and Dept, and introduce some parameters to adjust the distribution of inflation rewards, which will attract more mechanism participants and also make the mechanism more stable.

- Problem:

  1. At the beginning of the transition from inflation to deflation, because the Coupon is designed to expire after 90 epochs (to mitigate downward spiral events), the risk is high, which will make Coupon with low Premium lack of buyers, and the expected participants (speculators) will wait until Premium has risen enough to hedge risks. During this period, the number of participants will drop sharply, the community will become less active, and even affect ESD liquidity, which will increase the uncertainty of TWAP’s continued downward fluctuations.
  2. At the beginning of the transition from deflation to inflation, the inflation rewards will be given priority to Coupon holders, bonded ESD holders and ESD LPs may leave since they will not receive inflation rewards for a period of time. This instability of bonded ESD holders and ESD LPs has even introduced additional uncertain selling pressure in the market, which in turn affects the stability of the inflation cycle.

- Solution concept:

  1. At the beginning of the transition from inflation to deflation, active participants who are willing to buy Coupon with low Premium should have more rewards, or at least, should not be subject to the largest penalty for the earliest expiration of Coupon.
  2. At the beginning of the transition from deflation to inflation, ESP holders who are willing to stay in the bonded and LP pools should get more rewards.

- Solution brief:

  1. Deal with Coupon and Dept in a more elegant (non-violent) way, introduce an idea similar to CDO tranche, and design a Coupon stack queue, so that more people (not only speculators) are willing to participate in burning ESD to buy Coupon, and become more active. This will incentivize participants during the transition from inflation to deflation.
  2. Adjust the distribution of inflation rewards so that bonded ESD holders and ESD LPs will also be rewarded once entering inflation from deflation, while will increase their stability. Real inflation can then be controlled through parameter adjustments.

- Solution detail:

  1. Get rid of the Coupon expiration. That is, there will be no such a condition that the ecosystem is stable and alive, but the value of Coupon is forced to zero. Coupon and Dept will gradually be reduced during the inflation period.

  2. Modify the action of resetting Dept to zero in EIP-1, replaced with setting Premium (instead of Dept) to zero in inflation period. Zero-ed Premium means that there is no benefit to buying Coupon, so no one will burn ESD to exchange it. The non-zero Dept will be gradually reduced through the mechanism described in following the 4th item. If the system enters deflation before Dept is eliminated, Dept will be added up. The Premium calculated by Dept Ratio will not start from 0 to grow. Higher initial Premium will quickly encourage participants to burn ESD and exchange Coupon, reducing the uncertainty risk of a long deflation period.

  3. Design a Coupon stake queue (for auto-redeem), Coupon holder can stake his/her Coupon by itself, and the stacked Coupon will be automatically converted into redeemable ESD in order (via First-Stack-First-Service). And then the holder can redeem new ESD by burning the redeemable ESD at any time.

  4. When starting a new epoch (driven by Advance()), if any Coupon or Dept exists, divide the ESD rewards of this epoch into (p, q, r) shares with the initial value (0.6, 0.2, 0.2), where
    p is allocated to staked Coupon holders, and the system will automatically replace the stacked Coupon with redeemable ESD in order. If the stacked Coupon is emptied, the balance goes to q.
    q is allocated to Coupon holders. This part for Coupon handling is the same as it is now. This quota will be reserved for one epoch, and all holders will be allowed to grab redeem opportunities. The maximum amount of this part is the amount of all Coupon. If there is no Coupon, it will be used to offset Dept. If the balance of q in the previous epoch is not used up (there is redeem quota, but Coupon holders did not actively redeem), it will be used to offset Dept. If Dept is 0, the balance goes to r.
    r is allocated to bonded ESD holders and ESD LPs.

  5. Staked Coupon can be un-staked at any time. If stacked Coupon holders find that there exists some redeem quota q, they can immediately un-stake to grab the redeem opportunity. But if they don’t grab it and stake back, their stacked Coupon will be ranked last of stake queue.

  6. Redeemable ESD can be redeemed into ESD at any time.

  7. Summarize the order of inflation rewards allocation:
    p = 60%: stacked Coupon holder -> q
    q = 20%: un-stacked Coupon holder -> stacked Coupon holder -> Dept -> r
    r = 20%: bonded ESD holder and ESD LP

- Role:
Compared with other rebase mechanisms, ESD distinguishes different roles, rather than forcing all token holders to experience inflation and deflation, making the holding amount unstable. ESD separates the role of a common user. Following this idea, we classify roles into

  1. [Eu] Pure ESD user, general user: Simply treat ESD as a stablecoin pegged to USD for free use.
  2. [Ea] ESD price fluctuation arbitrager: They are also Eu, but based on the stable price fluctuation of ESP, they use the method of buying low and selling high to earn the spread. For example, set trading-grid in 0.97 ~ 1.03 (set buy ESD orders at 0.98 and 0.97, set sell ESD orders at 1.02 and 1.03), when the price is normally stable within the range fluctuation, Ea can use this strategy to obtain stable spread profit, the system will be more stable and close to 1 because of their participation. Compared with Eu, what they sacrificed is the liquidity of ESD for free use.
  3. [Elp] ESD LP (Liquidity Provider): Hope to be rewarded by providing transaction liquidity. What they sacrificed is the liquidity of ESD for free use, and IL (impermanent loss) of general LP is the risk they should take.
  4. [Eb] staged/bonded ESD holder: Hope to get interest rewards by depositing ESD. What they sacrificed is the liquidity of ESD for free use, and the inconvenience is higher than Elp, because it takes two epochs before it can be used freely. Compared to Elp, Eb does not have to take IL risk.
  5. [Cs] staked Coupon holder, active participant: Hope to get a small amount of stable income by purchasing Coupon with low Premium (discount), which can then be automatically replaced with redeemable ESD, and then with ESD, by staking Coupon into stake queue. What they sacrifice is the liquidity of ESD for free use, and the uncertainty of liquidity is higher than Eb because they are not sure when they can redeem ESD successfully for free use. Also, they should take the reinvestment risk after swap back to ESD, because it is uncertain whether a new Coupon is available for purchase.
  6. [Ch] un-staked Coupon holder, speculator: Hope to get a higher return by buying Coupon with high Premium (discount) and earning a profit as soon as possible by grabbing the redeem opportunity. They take the biggest risk. In addition to the risk of Cs, there is also the risk of not being able to grab the redeem opportunity for a longer time.

- Expected result:
Imagine that ESD becomes a stablecoin after the system mechanism is stable. The price fluctuates from 0.97 to 1.03 most of the time. The expected benefits of each role are analyzed as follows:

  1. Assuming that TWAP fluctuates from 1.03 to 0.97, from inflation to deflation, the active participant Cs will first enter the market to buy Coupon with low Premium, and immediately stake it to the stake queue. This makes Dept to be reduced earlier and the system cost for Premium is less, which will encourage TWAP to return to 1 to maintain stability.
  2. If the buying momentum of Coupon with low Premium is still insufficient, deflation continues and Dept increases, then Premium will gradually increase and will begin to attract speculators Ch to enter the market. Speculators look forward to the early profitability of Coupon with high Premium. The share of q provides the exit so that they are not limited to the stake queue. Note that q is a parameter. We can increase q to attract more speculators and mitigate downward spiral events.
  3. When TWAP fluctuates from 0.97 to 1.03, from deflation to inflation, Cs will have a priority quota, which will allow Cs to exit the market in a short time and get a small Premium profit. At the same time, because there is a share of r reserved for Eb and Elp, there are still inflation rewards that will make them more willing to stay in the system stably. Note that r is a parameter. We can increase the reduction rate of Coupon or Dept by lowering r to control real inflation. When the market cap is in a period of active expansion, r can be adjusted to encourage more participants to become Eb and Elp earlier, without waiting for the elimination of Coupon and Dept.
  4. Due to stable TWAP fluctuations, Ea will be attracted to participate in the mechanism. For example, they set trading-grid in 0.97 ~ 1.03. When TWAP is normally stable within the range, the spread profit will be obtained, and TWAP will be more stable and much close to 1.
  5. If the system is in a stable state, half of the time is in inflation, and half of the time is in deflation, half of the inflation reward plus the average Premium cost will be allocated to Cs and Ch, and the rest to Eb and Elp. That Cs and Ch getting a slightly higher profit is reasonable because they take a higher risk.
  6. Stability of TWAP is absolutely attractive to general users of ESD. It will increase more users into the ecosystem and bring a period of active system expansion. The 8-hour epoch and 3% inflation rewards may be adjusted in response to expansion needs, all participants will also be rewarded by market cap growth.

- Extra benefit & Future work:

  1. [Tokenized Coupon] The modified mechanism has an additional benefit: We have the opportunity to tokenize Coupon as ERC-20 token, because each Coupon is the same, and it will not be different due to different produced epochs. In this way, Coupon can have another exchange market to increase liquidity. The market efficiency will make the price refer to the total amount of Coupon, the number of stacked Coupon, the reduction rate of stake queue, the stability of the mechanism, and the inflation time to determine the price.
  2. [NFT Coupon] After the system is stable, under normal circumstances, Cs will buy out most Coupon with low Premium. We only need a few Ch speculators to participate. When the system needs to attract more Cs speculators to participate, (p, q, r) parameters can be adjusted. A more active approach is that we can also issue a special Coupon: Coupon with huge Premium with expiration time. Because there are different expiration times, we can use NFT to issue, and the system will need to increase the NFT Coupon redeem mechanism.

First of all, really great to see another community proposal! @Reyer :clap:

If coupons cannot expire, what role do they play in the system? Maybe I’m not fully understanding the mechanisms being proposed, but it seems like this makes coupons ~risk-free, which means all premiums are effectively inflation to the system. If coupons are transferrable and never expire, why wouldn’t all ESD holders just convert to coupons to capture the premium? IMO, the risk of expiry is a necessary component to ensure that coupons are purchased by those who will benefit the system by either investing more to support the peg or will ultimately reduce the supply of ESD by having their coupons expire.

Even if I can get comfortable with this overall proposal, I would be strongly opposed to eliminating the expiry on existing coupons - that is akin to a bailout of coupon holders and is not a path I think we should go down.

1 Like

@delitzer Thank you for your comment. Here I listed some viewpoints:

  1. Coupon itself is a media for dealing with Dept, similar to a bond. Since it is a debt, the borrower is, of course, looking forward to getting the repayment in the future. Just like if we take a loan with a bank, and the bank takes on debt, it is equivalent to buying Coupon (debt) from us with the bank. If we are still alive, even though we still cannot repay the money, the bank will never cancel your “Coupon” because of “expired”. This is what I mean: When the system is still alive, how can Coupon issued by the system be cancelled because of expiration? The creditor’s rights can only be cancelled when the debtor claims bankruptcy to repay the money at a discount, or dies (ie. system crash). Compared with corporate bonds, corporate bankruptcy or closed is the biggest risk for corporate bond holders. Otherwise, the holder holds it for one day and expects to get the repayment when the company regains money.

  2. Even if Coupon has no expiration, there are still risks, which is by no means equivalent to ESP. It is not real money as ESD. It has a lock period with inability to be used like money. The uncertainty of this lock period is the other risk the holder needs to take. If you are willing to take such risks and use your money (ESD) to buy Coupon, you should have Premium in return.

  3. About “If coupons cannot expire, what role do they play in the system?", we can regard Coupon holders as investors of the system. When the system enters deflation, they are willing to pay money (ESD) to help the system offset the debt (lord). They have to take the risk of not knowing when they will be able to repay the money, the time cost of their money during the lock period, and the risk of the system crashing and Coupon down to zero, which is similar to the risk of buying corporate bonds.

  4. In terms of the fairness of the current Coupon mechanism, the bonded ESP holder does not need to take the risk of the uncertainty of the lock period, and the “down to zero” risk, but it can get almost the same rewards (just no Premium) as Coupon during the inflation period. We can treat Premium as a compensation of the uncertainty of the lock period. But the profit is not large enough to bear the “down to zero” risk. Now because the mechanism has just started, many people bought Coupon without knowing the details of the mechanism. It seems that there are participants to buy Coupon. But if Coupon “down to zero” happened, after careful analysis, even if the Premium is as high as 45%, it is not enough to hedge “down-to-zero” risk, no one (even speculator) will buy Coupon anymore, and the whole mechanism will collapse. Therefore, I pay special attention to the role of stake Coupon holder. This active participant is the key role that the system can succeed, not a speculator.


I agree with @delitzer that we cannot remove the coupon expiry and we need a deflationary effect for coupons. In EIP-3 I proposed allowing coupons to “roll” into newer coupons at a premium. Perhaps that would be better relegated to this EIP-2 discussion. Rolling at a premium would have a net burn effect on coupons that roll many times, we could even factor in the epochs to expiry when calculating the rate.

More related to the proposed solution above. Stacking and ordering coupons is technically difficult and given my limited experience with Solidity, somewhat intractable in terms of gas costs. Someone more experienced can confirm or deny this.

A tokenized version of coupons in the current system already exists CPOOL. More info at https://esd.money. I think an additional measure we could take is to try to incentivize CPOOL liquidity by minting more CPOOL to give to CPOOL/ESD LPs.

I think the most interesting component of this proposal is to have some staked version of coupons get first priority redemption:

  1. This must be a fungible wrapper around tokens, priority ordering likely will not scale
  2. It must be an attractive, fair, and open alternative for the community to support inserting a new privileged form of stakeholder into the ecosystem.

I have a few more ideas relating to this “first access redemption” concept if people are open to the idea

An easy way to realize a FIFO queue mechanism:

  1. Holder stakes Coupon into stake queue => just like holder stages ESP into pool, via web now.
  2. For auto-redeem mechanism, the system doesn’t really redeem ESP to the holder’s wallet, but just change the staked Coupon’s “internal state” to become “redeemable ESP”. And then the holder can redeem new ESD by burning the redeemable ESD at any time.
1 Like