Changing the collateral factors for the eRSDL token

Scenario A: Least aggressive, allowing more time and smaller steps to adjust the collateral factor from 80% down to 40%.

Scenario B: This is middle-of-the-road in terms of risk and aggression, with four proposed levels in order to achieve the lower collateral factor

Scenario C: Is the quickest route to change the collateral factor, and therefore the most risky for the ecosystem

You can Vote here :point_down::

uneRSDL voting - Snapshot

eRSDL voting - Snapshot

1 Like

I appreciate the hard work and data crunching that went into these scenarios. Good job.

3 Likes

Wow a 50% reduction in eRSDL CF (80% to 40%). So anyone above 50% borrow limit (the borrow limit bar which is normalized to 80% on RL for all assets) who is using eRSDL as collateral will have to pay back debt or be partially liquidated. Even someone who is at 60% borrow limit (for current CF = 80%, so borrow/supply = 48%) will have to pay about 17% of their debt back to prevent liquidation for CF =40%

3 Likes

Our analytics have shown that at the pace laid out in the scenarios, the current market volume should absorb the potential liquidations. In advance of this vote, and the rationale behind our messaging this for some time, we have seen voluntary repayment or paydown of eRSDL-backed loans. The end result will be an appropriately risk-based priced asset. I’m excited to see the platform with optimal risk-based pricing in place.

3 Likes

I believe scenario B is the perfect blend between risk and speed to implement the proposed CF changes. In my opinion, the debt volume is not high enough to justify the slow implementation of scenario A, while scenario C might be too aggressive in terms of risk for the ecosystem.

2 Likes

I like Scenario A. The majority of the adjustment is still happening in T1 through 4. and the slower impact of -0.1 for the first 2 steps allows the borrowers more forewarning. Whereas in B and C 25% reduction in borrow limit in one shot could end up forcing unnecessary liquidations, and/or encourage more borrowers to just close out their loans all together, which I don’t think is the goal. I also suspect with all of these schedules we’ll see more “voluntary paydown” to that final level right as the schedule goes into effect.

3 Likes

SemperFiLink mentioned a spread between the CF and the liquidation trigger to encourage folks to borrow up to the adjusted CF. I’ve dropped a note into the dev team to provide us options. We are a fork of Compound, and so subject to the limitations that come with that. Also, any project you are interested in should want to keep protocol tweaking to a minimum. Every time you crack the protocol to add a feature or widget, you create a new attack vector. Just wanted to update folks in this thread.

2 Likes

maybe this has already been stated, but does T = one day?

1 Like

So “T” is equal to the date the adjustments commence. We looked at the current book the day before the Forum post, so this is based on recent data. We compared possible liquidation levels assuming no borrowers prepay with observed daily volume.

All of the scenarios resulted in liquidation volume well within daily volume amounts. Not everyone had borrowed to the hilt, and in advance of this poll, folks started to take down their borrows rather than get liquidated. Personally, I’d like to see the adjustments get behind us, sooner rather than later, so that we can move forward with safely bringing in more stablecoin.

1 Like

Firstly, I agree this change is needed and have personally voted for Scenario A.

I do however feel it’s reprehensible that a proportion of users of the lending platform who have adhered to the given lending terms will be liquidated as a result.

Those who have most likely already faced turmoil from the market conditions over the past few months and now have to allocate funds that they may not even possess to save their position.

I’m surprised at the statements as being almost a matter of the fact that users will be liquidated. The assurance, seemingly only applicable to those unaffected being in the form of the “market volume” absorbing the damage. Isn’t this negligent towards the early-adopting community of users?

I feel that the proposal by SemperFiLink or some form of cushion would be the only fair way to roll out such a change and it should be done over a considerable timeframe with every effort made to raise awareness.

3 Likes

We’ll see if we can .pdf and share the data so we are on the same page there. We are not expecting anyone to get liquidated that has not already (technically) walked away from the platform. The holders we identified as likely to be affected, and it’s not a tremendous amount at the CFs we are talking about, were not compelled to pay off their loans at 30% interest rates.

This CF change is not a first measure to properly calibrate the tokens risk v yield. The high stablecoin borrow rates, which could be perceived to being unfair to borrowers also, were put in place to encourage voluntary prepayment. To a certain extent it worked.

This CF measure then gets the risk and reward aligned based on the volatility of the collateral’s value. The last step will be a critical examination of the interest rate models versus Compound, AAVE and other large platforms. We can price competitively and properly for the risk, and start leveraging institutional relationships through our SDK/API integrations.

SemperFiLink’s suggestion that liquidations occur at some level about the CF, thus creating a buffer zone, is a good one. We have put a ticket together for our developers to work on that. Unfortunately, AAVE is the only protocol we observed having that feature, but that research is part of the ticket. The native Compound fork does not have that functionality.

I hope this addresses your concerns, and I acknowledge and appreciate your thoughts.

Out of all the scenarios option A causes the least pain to “current investors” and users of the platform…(in my opinion). Im just struggling to figure out if option C causes more harm than good. definitely seems it could in the short term. Looking forward to seeing the data if you can to help make a decision.

.

1 Like

I think it’s crazy to lower the native token borrow limits to half of the current limit. I’m sure models show it has to go to .4 but there needs to be a way for certain loans to be grandfathered in to the current model. A lot of us have stuck through the drama of SSS, the huge price drop, and were protecting our ersdl collateral because of this and market conditions…I had heard a part of the reason for the limit drop was not just to solve liquidity issues, but was to ‘protect investors’ from borrowing too much against a volatile asset like ersdl. Well, some of us have been through the ringer on this, and if part of the reasoning for this rate was calculated to ‘protect me’ I’m not a fan of that. It’s one thing for a person to be force liquidated by the market, it’s another thing to be force liquidated because of a mistake in the lending model.
I am a more risky borrower but I’m always aware of the ersdl chart and its support levels, and how much money to have aside to pay my loan down on retests of those levels. I am being forced to sell off assets when the market isn’t good and ersdl is at a low price because of this. I based my risk off that .8 number. So now, if I pay off half my loan, and then ersdl retests its .10 low, or goes lower, I’m toast, along with a few others I know.
I know the limit needs to go down, but hoping it only goes to .6 or there’s a way to grandfather in loans to the old model. I know this can’t be easy for the team to decide on, and early users of the platform are open to changes, but please find a way to keep early platform users happy who have been supporting the project since the beginning.

3 Likes

But is the +1,+2, etc after T, are those weeks, or days or?

T is the day we implement the first adjustment, and the +1, +2 are “days”.

I appreciate your thoughts, and that as a borrower using eRSDL as collateral, any of these outcomes negatively impacts your experience. All of this could have been avoided had we gone out with proper collateral factors to begin with, and I am sorry this was not the case. I am focused on our attractiveness once the newcomer/stablecoin holder fear is allayed regarding the ability for folks to leverage risky assets to the hilt.

In a conversation with another holder, I provided some thoughts on the matter (including the one you mentioned about grandfathering in):

“Thank you for sharing your experience and those other folks. The three things I want to build for us, bc they don’t exist and should): 1. the ability to select supply to sell to pay off borrows, 2. a spread between collateral factor and liquidation threshold, and 3) a way to grandfather in loans made under certain conditions (but not regulatory ones)… I think suppliers should have recourse also. My two cents. Thanks for sticking with us.”

I guess that was four things and not three. Posting data next.

1 Like

I’m not sure if its possible but can the data include user numbers thats will be under massive stress or in a liquidation position at 40% CF.

Additionally did your model look at liquidations happening potentially dropping the price alot more than normal market conditions?

As an investor of the token I am very interested in our current holder amount (floating around approx 13k for a while) and how these changes may affect that numbers please.

I understand this is a difficult decisions for the the team and its keeping people happy with one hand but taking away from others with the other hand. I myself am at no risk but I know a few people who are and have pretty much accepted there going to be unfortunatly liquidated.

I’m not sure what grandfathered means sorry can someone please elaborate?

The pictures in this post are from one of our homegrown analytics tools we built to measure the impact of the collateral factor changes. This is my first forum post with pictures in it, so hopefully it renders properly. The information in this post should be considered draft, and since all the information is on chain, we would encourage folks to do their own analysis when making the decision to engage with the platform. We cannot guarantee any outcome and their are significant risks to users which should be considered.

This table is a summary of where we feel the CFs will end up. We are are not targeting an ultimate CF, and would like to see proposals in the Forum around this and other attribute changes going forward.

Here are historical daily volume (eRSDL count) figures with a variety of statistics (e.g. ave, max, and min) displayed. If we are being conservative 3.75mn to 5mn tokens trade daily.

If we look at what happens when the CF is lowered from 0.8 to .6, about 950k of eRSDL is impacted. Our conclusion is that market volume will absorb this activity. We considered that 1) these transactions will not occur simultaneously and 2) the bots may not pick up all the loans if it doesn’t think the economics make sense.

At first glance, the impacted eRSDL count of 27.9mn with an additional 4.7mn in the orange zone would appear too much for the daily volume to handle. Consider, however, that we are observing here the total count at a 40% CF. The step-down scenarios have been crafted in a way that attempts to keep liquidations well below daily market volume (count) observations.

There are risks. A steeper drop in the token price pushes more folks into liquidation. One benefit we observed was that while, in the short term, there will be selling of the tokens tied to liquidations, the post change platform is safer than the pre-change one. As I’ve spoken about in the past, lowering risk increases value and provides interesting opportunities.

We will be monitoring for those changes, and reactions can be made accordingly. From a leadership perspective, one cannot stress how much better aligned return v risk the platform will be after the changes are implemented. We do recognize that these changes will have a negative impact on some users, and we are listening to those issues as raised.

-howard

1 Like

Based on the differences between when CF = 60% and CF = 40%, it looks like a large majority of liquidations would occur during the second half of reduction from 80 to 40%, squeezing liquidations into a small time frame. Interesting that the step down scenarios have been crafted in a way that attempts to keep liquidations well below daily market volume because it would seem that option C would completely and utterly flood the uniswap liquidity pool with ersdl.

I’m not sure if looking at daily volume is useful since not only will people not want to buy when they know an enormous sell pressure exists but so much of it is from bots. Uniswap is absolutely riddled with bots. Bitmart is very unreliable. I believe the bulk of kucoin’s volume comes from bots arbitraging between it and uniswap.

So for the sake of an easier example, let’s look at option B. The numbers you’ve given are for ersdl at about 0.194. Right now, it is 14% lower. If it were 10% lower, anyone who after the adjustment is in the 90-100% LTV region is now at 100% or higher. Then for the first adjustment, about 5.6 mn tokens will be impacted. Let’s assume there will be selling and liquidating to get people back into good standing. But where is good standing anyway when you know more selling is coming. Ongoing good standing would need to be well below 100% LTV. Then there are people who will be in the range like 80-100% who are not comfortable with this. They will sell, maybe preemptively. Swing traders or investors knowing of these changes may sell to be able to buy back lower. Even if we ignore this, what happens if 5.6 mn tokens are impacted when ersdl is at .9*.194 = .175 (still higher than where we’re at now). I’ll assume half will be sold or liquidated to get borrowers to good standing or comfortable levels. And then bots will arb with buys on uni and sells on kucoin/bancor. Demand needs to exist on kucoin so I’ll assume half of this half will be bought. That’s net 1.4 mn sold on uni. I’ll assume that 1.4 mn sold is sold at an average price of .15 such that the liquidity pool goes from 9.46 mn ersdl and 533 eth to 10.86 mn and 463 eth. For eth at $3k, that puts ersdl at 0.128. Let’s say I’ve done something wrong and we’re actually at 0.14. Well now we’re heading to CF’s under 60% at 0.14 and not at 0.194 like the above numbers rely on.

Maybe I should assumed more absorption by other exchanges via bots or people manually arbing? What about people simply panic selling? How about the lag in RL where updates sometimes don’t happen for over an hour and prices can move very quickly down?

6 Likes

If the ‘T’ periods are days and not weeks the options of spreading the change to 0.4 over 2, 4 or 9 days feel a bit pointless really, as in current market conditions an extra 7 days isn’t going to give people time to free up capital to pay down their loans.

My humble opinion: need to be very careful here as if it is days and not weeks any of the three options will cause mass sell offs/liquidations and further panic selling leading to more liquidations. In fact the announcement may lead to people stabling up in anticipation to protect capital/swing trade. I don’t think your projections of the market absorbing the liquidations would play out accurately unless you were to drop extremely bullsih news at the same time.

For the record I’m not at risk personally (other than capital depreciation on my ersdl when it nukes), I just don’t want to see loyal community stalwarts liquidated and moving on when being this aggressive feels unnecessary. A much more conservative solution feels like the play to me. If the B2B products are the ‘main course’ and the retail platform was a test bed and playground for the early community, why not just take a slower scaling back approach over 8-12 weeks, leave the lending platform to slowly change course and focus on the institutional products?

4 Likes