Sunsetting Perp V1

From discord:

How does everyone feel about this idea:

Everyone gets their full position in newly minted PERP, and that PERP is locked+staked for 1 year. All of the remaining USDC funds are paid over 2 years as weekly staking rewards to all staked PERP.

To be clear you get 0 of your USDC back. You only get staked/locked PERP.

Existing PERP holders are happy because they get a yield increase, it incentivizes new buyers of PERP because of the increased staking yield. When PERP 1 year unlocks, the yield increase continues for another year so many will stay staked. If PERP price holds then everyone gets their funds back.
PERP price may initially increase because of this.

At least this way there is a chance that everyone is repaid, and PERP doesn’t suffer reputation damage.

2 Likes

Relative to 1-4, I think is a good option

1 Like

This could be combined with one of the “Robin Hood” proposals (options 2/3, 3 is better IMO), considering that 90% of the claim value (margin + uPNL) concentrates in the top 5% of accounts.

We could thus, e.g., use 10% of the USDC to immediately make whole the bottom 95% of the userbase. The remaining top 5% would instead get compensated with PERP in the two-year-lock staking pool, which gets fed with the other 90% of USDC.

This assumes that all HNWI are treated equally. It’s a separate question whether the very “whale” that overstrained the system past its point of rupture (Alameda or whoever they are) should get the same PERP compensation or not.

2 Likes

When can we get the official spread sheets of various scenarios? I believe even the top 5% should have the choice of whether they want to take the loss or participate in more risk. Let’s say if we still get 60% of the Margin-PNL from option 1~4, then we’re essentially taking a huge risk in participating and risk losing everything if Perp couldn’t hold the price. It really depends on how much we can get back through option 1~4 before putting such risky proposition to the vote.

4 Likes

Proposal:

  1. We’re doing calculation for each wallet : Value = margin + Upnl at oracle price at shutdown moment Upnl could be negative. Upnl = size*(oracle_price at shutdown - entry_price)
  2. Sort every wallet by Value
  3. compensate in full starting from wallet with smallest Value and going up until we run out of compensation fund (5.5 mln)

I pretty much sure, as per my calculation retail even significant ones will be covered. Alameda unfortunately not.

3 Likes

Can’t we use for example last index price to calculate every trader equity at shutdown time, then we know the bad debt, and we can take into account OI to give a fair share of debt to every trader?
That way, no matter your size, pnl or leverage, everyone take their share of the losses?
If one has 10% of the OI, he’ll have 10% of the final haircut?

4 Likes

Proposal
Option 6:

Perp team refund 100% of the funds at closing exchange prices from the perp1 funds, perp2 funds, and team’s PERP token reserves.
w\wo locks but it should be whole 100%

The reputation and reliability of all of the team’s products is at stake, so all means must be employed without any compromise.

7 Likes

I support a Robin Hood option better, but if we’re going for a proportional option, then yes make the haircut proportional to OI seems much fairer than making it proportional to the netted margin

3 Likes

See no reason to harm PERP holders and compensate Alameda who was harvesting funding and knew they couldn’t leave unharmed. If we compensate them would be completely unfair.

2 Likes

It looks to me that the community is proposing many interesting options. In contrast, I see little support expressed for any of the 4 originally proposed options (@Long-run-view has expressed support for option #4 though in a somewhat modified form; myself I’ve expressed a slight preference for option #3, though not any strongly). I’d like to see all of the community proposed options going to the vote but, at the same time, I’m afraid of having a divided vote between too many similar options which will make them lose strength even though they’d have a majority support if merged (and I don’t think Snapshot supports something like the Borda count, does it?)

Therefore, I’d like to propose that we do a multiple vote, with the elements of the compensation plan voted independently. It could be something like this, or a variation of this:

  • Question 1, should small traders (margin+uPNL < $100k) be made whole from the clearinghouse funds first? (yes/no)
    (if yes wins, small traders are already whole and thus removed from what follows)

  • Question 2, should there be a full compensation from the protocol, this is, on top of what remains on the clearinghouse? (yes/no)

  • Sub-question 2.1, in case yes wins question 2, what form should take the compensation?:
    i) fresh PERP minted and autctioned for USDC, or
    ii) remaining USDC and some new PERP, staked in a one-year vesting pool

  • Sub-question 2.2, in case no wins question 2, what should be the bail-in priority?
    i) Simple cut, as proposed by @Ryan (leave out UHNWIs like Alameda or others)
    ii) Proportional, prioritizing the less-leveraged positions / higher margin ratios. I think this is one of the iterating methods like the option 4 or Long’s proposed variation of it.
    iii) Proportional, prioritizing by twapPositionNotional

3 Likes

Alameda wasn’t the only one who had a big position on the perp. I’ve been trading there for the whole year and a half since it opened and I’ve done nothing wrong. Why should I get less than the guys with smaller positions? it’s just not fair.
If the team and participants want to punish Alameda - let them not refund only those addresses that belong to Alameda, but not all holders of large positions.

4 Likes

What’s your position was? Let’s wait for csv and check who will be out. IMO large positions say 500k+ knew what they were doing. They earned good money doing funding, and now getting compensated above that is not right.

1 Like

There may be some nuance to be made between HNWI and UHNWI, though I’d prefer to avoid the bike-shedding this can bring. Besides, is it reasonable to think that people where putting six figures in this and not using any surveillance scripts or advanced metrics? It seems reckless if they didn’t (at least that’s what my rich frens do)

I think a ranked choice voting system would solve the problem of a divided vote between too many similar options.

See this video: What is Ranked Choice Voting? - YouTube

I believe crypto governance as a whole would benefit from ranked choice voting; hopefully it catches on.

Alameda has received over $36 billion in farming income.

Source: https://twitter.com/DefiMoon/status/1526783596572422146

Every single depositor in v1is a shrimp compared to Alameda.

Alameda also is getting almost half of the remaining funds in option 1.

The team has said that they will add any option that has community support. Overwhelmingly in this thread excluding Alameda has been supported.

So add option 1 to the vote, except exclude the 5 Alameda addresses:

0x0f4ee9631f4be0a63756515141281a3e2b293bbe
0x975d2dd35726e1f4e6af7e22880c49ffeda67339
0x88301e10b0ee2f7eb2c9dc9c8b98b813946d144e
0x3013ea1baafe595d6c54419a0d7e420de524c52b
0x5dac431f0479a1413c8734634c7d9de577675d40

The reason is simple: They can afford this loss much MUCH more easily than any other addresses in v1.

Edit:

Seeing that Option 2 is winning, I am losing sleep over this loss. A lot of sleep. I don’t think Alameda is. I don’t think the $1000 account average joe would lose much sleep over a $200 loss in the case of choosing option 4. Option 4 would save a lot of users like me from tremendous pain by essentially transferring funds from Alameda to medium size accounts.

10 Likes

Updated on 19th May 2:30 UTC

Please find data for each of the 4 team proposals at the link below:
https://docs.google.com/spreadsheets/d/15hajD9CatNgNMgjGq6QdvCw4dGqndFGG0V30vof2Smo/edit#gid=0

Notes:

  1. All amounts shown are expected refund under the respective proposal. Cells with - are dust positions.
  2. Amounts shown for options 1-3 assume price impact of a full close with no other trades occurring, using original v1 K values and index price at block 22169196, and assuming that your position was the first and only position to close.
  3. Option 4 models the actual functioning of an AMM system in a scenario where all traders gradually reduce their positions to zero.
  4. Data is not available for community proposals. Due to this, the vote will include a fifth option to fully evaluate and vote on community proposals in a separate vote, in favor of the four options presented by the team.

Vote is Live

The vote to decide the method in which remaining Perp v1 funds are to be shared among remaining users of the v1 exchange following the shutdown on May 15 is now live. Holders of PERP and staked PERP on Ethereum Mainnet and the Optimism Rollup are eligible to vote.

https://snapshot.org/#/vote-perp.eth/proposal/0x846cb20e36f367dac78fbdf100a6c791d119cc5269850e060fcb87040122893e

Option 4 is closest to actual system closing, as hana said:

Option 4 models the actual functioning of an AMM system in a scenario where all traders gradually reduce their positions to zero.

Option 4 is basically the “exclude Alameda” option.

It bankrupts multiple Alameda positions, while leaving everyone else with reasonable amounts.

For medium size accounts, anything other than option 4 creates a massive life & death difference (to borrow some wording from Discord).

As discussed in the last two days, option 4 treats low margin leverage accounts that have excess collateral much more fairly.
Alameda closely monitored their positions and has already withdrawn collateral, while other unsuspecting “large” retail accounts left all their collateral in v1, in some cases with less than 1.0x leverage. All options other than option 4 devastate those accounts for this reason.

4 Likes

Option 4 is the most fair, in my opinion. It is the only OI based option.

Note option 4 is the only option that doesn’t just use net margin. Instead it takes the users actual position into account.

This addresses concerns raised earlier in this thread. Other options haircut free collateral of users that have lots of collateral but a small position. Those users could have just withdrawn collateral without making a trade if they had any idea the v1 failure was coming, like Alameda knew.

The top 5 addresses in the following image are Alameda. Look how much more option 4 gives to almost all Perp retail users. Notice that option 1 rewards Alameda the most of all options.

This is partly because Alameda has already carefully kept as much collateral as they could out of v1 before failure occurred, while unsuspecting users didn’t try to withdraw their collateral, like Alameda did.

It is also because in v1 large positions take more exit risk than small ones. Alameda took this risk. Why should mid size accounts now pay for Alameda’s risk with a new set of rules introduced with Options 1-3?

Backup image link: https://i.imgur.com/mzvJkNb.png

Option 4, which models the actual closing of the system is the most fair.

4 Likes

Hana, it’s disappointing to see that despite you requested feedback on the options, I see none of that feedback incorporated in the final vote. No word on leaving Alameda out (“Alameda” as a placeholder for UHNWIs). No word on a compensation with PERP. Essentially, zero changes on the options as stated two days ago, except for the addition of a totally unspecific fifth option, which only divides the vote even further – four options are already too many.

If I was in your place, I’d seriously consider doing a subsequent vote to treat the question of the PERP compensation. Something like this: After the current vote, and on top of whatever option wins it, choose on what to do with the amounts left unfulfilled:

  1. Mint PERP to compensate everyone, including Alameda/UHNWI
  2. Same, but excluding Alameda
  3. Do not compensate

Do you think that would be acceptable?

2 Likes