Gauntlet Reserve Factor Optimization Framework

[klinsuain]

As planned in an earlier forum post, we have updated our dynamic parameter methodology to include recommendations for the reserve factor, as well as our usual parameters. To outline our thinking behind this addition, we split lending parameter optimization into a two step process:

• Balancing losses due to insolvency with capital efficiency
• Balancing the tradeoff between increasing supplier APY and growing reserves to cover insolvency losses

In the first stage of our modeling, we focused on optimizing capital efficiency within the community’s acceptable risk tolerance. Our primary levers in this optimization are the Collateral Factor and Liquidation Incentive, which most importantly allow us to manage the overall Value at Risk (VaR) of the protocol. We seek to balance VaR against borrow usage, which we use as a metric for the capital efficiency of the protocol. Since losses to a lending protocol occur during insolvencies - events where collateral is unable to cover the full value of a liquidated loan - controlling the frequency and severity of such cases is key to managing the overall value at risk. Within our simulation framework, the Collateral Factor drives the likelihood of a given loan triggering its liquidation threshold. The Liquidation Incentive parameter then drives the cost of liquidating the collateral and by extension the severity of insolvency loss.

A schematic view of the first stage of our modeling process is shown in the diagram above. Starting from the top left, we search through a range of lending parameters, which serve as inputs to the simulation. The simulation output then determines the level of risk and protocol efficiency for the trial set of parameters, which are evaluated against the protocol objectives. Finally, we can adjust the input parameters based on feedback to more closely meet the objectives. Now that we have reviewed the process used so far, let’s consider how the reserve factor changes things.

How does the Reserve Factor fit in?

We are now adding a layer to the existing simulation logic that allows us to reflect the second step of lending protocol risk management: building sufficient but not excessive reserves to absorb losses that do occur. Our model assumes the protocol would like to build reserves greater than insolvency losses by a small margin, and includes this as an optimization objective. As reserves are built and losses occur in the simulation, a higher setting for the reserve factor allows us to absorb more losses while maintaining reserve coverage, while a lower setting for the reserve factor would require losses to be managed more tightly.

A schematic for the new model structure is shown above, incorporating all the existing elements of the initial model and the new reserve factor layer. This illustrates where the reserve factor fits downstream of the simulation logic and how reserves are measured against losses as a model objective. As we continue to gather data on reserve factor adjustments, we expect to also incorporate the second-order effect of the reserve factor on the simulation itself, which is shown here by the dashed line. Because setting aside reserves locks up some assets that would otherwise be available to protocol users, there may be an indirect effect on user economics and thus capital efficiency from this. To refine our model here, we plan to recommend incremental adjustments to reserve factors that will gradually move reserve growth towards protocol targets while allowing us to further study indirect effects as the data becomes available.

Our Recommendation and Reasoning

Recommendations:

We recommend decreasing reserve factors for USDC, USDT, and TUSD from 7.5% to 3.75%, and decreasing all other reserve factors by 500 bps (for example, decreasing ETH reserve factor from 20% to 15%).

Reasoning:

The one-week rolling average VaR for Compound in a Black Thursday magnitude event is $0.30M.

The protocol has an existing reserve pool of $42.87M, broken down as follows:

The protocol is expected to build reserves at a $3.86M annual rate should the reserve factors remain unchanged, broken down as follows:

Since the existing pool and growth rate are more than sufficient to cover a very severe loss event, we believe that the optimal level of reserve factors is well lower than they are currently. By implementing the recommended cuts, the protocol could reduce its estimated annual reserve growth rate from $3.86M to $2.33M, reallocating $1.53M of funds to suppliers and thus incentivizing more users to lock collateral in Compound. As we continue to improve our model, we look forward to updating the community on further progress and fine-tuning our reserve factor recommendations in a future forum post.

^{By approving this proposal, you agree that any services provided by Gauntlet shall be governed by the terms of service available at gauntlet.network/tos.}

[dvf]

I’m in favor of reducing the current reserve factors.

With the total funds in Compound having been much larger in the past, the protocol has effectively already banked about six years worth of reserve (at the current protocol size) in the last year alone.

Am I correct, @klinsuain, that if the overall total borrows and overall total supply went up, say 3x, and all other things stayed equal, then we would want to raise the reserve factor again?

[klinsuain]

Thanks @dvf. Changing the reserve factor up or down would depend on the insolvency risk. Borrow and supply amount are factors that go into insolvency risk, but there are a multitude of other factors as well (e.g. user collateralization ratios, liquidity of the borrow asset in secondary markets, asset volatility, etc). Our simulations incorporate these data and provide insight on insolvency risk, and reserve factors will be adjusted accordingly.

As an illustrative example, if the supply increases 3x and the borrows increase proportionally, then there will be more total interest going to the reserves. If the insolvency risk remains unchanged, then it may be reasonable to decrease reserve factor even more. If the insolvency risk scales proportionally, our models may recommend to keep the reserve factors the same. If the insolvency risk increases at a higher rate than the reserves, then it may be prudent to increase the reserve factors.

[pauljlei]

As a heads up to the community, Gauntlet will publish an on-chain vote on Sunday, June 12, by 5 PM PT for these reserve factor changes outlined above.

[jake.aujus]

Update from the Gauntlet Team:

• Linked Proposal: Compound

• We are putting up 10 actions due to that being the limit for proposals.

• Should the vote pass we will look to proceed with the remaining 7 we had prepared, as we cannot put up two proposals at the same time.

[Sirokko]

Before i present you my opinion, let me tell you story. Imagine you are an owner of retail store. And one day manager comes to you with a presentation of his idea. It’s a very nice presentation, with graphs and maybe even some math. But the core idea of that presentation is he suggest you to cut profit margines for your store, and the money, saved from cutting down your profits, be used to pay more for your suppliers, so then suppliers could bring more supplies to your store. And then you look at very logical presentation and look at your supplies, and then you ask just a one question. Ok, but wait a minute, i look at supply of one thing (usdc) and i see it utilised by about 35% and i look at supply of another thing (dai) and see it utilized at about 45% and i as an owner would prefer it actually about 60-75. And you seriously suggest to cut our profits, in favor to bring more of the dudes i don’t even want to have? I was actually thinking how to get rid of 1/3 of them, so utilisation would go up and remaining ones would actually earn more? How long that manager would keep it’s job i wonder?

So now, when you see the picture, let’s talk why this proposal is bad for protocol and why pretty much anyone in their mind should vote against. I don’t know where Gauntlet got that idea, that protocol reserves are there only in case of some unlikely insolvency moment. It’s not. It’s earnings of the protocol, which can be used to cover insolvencies, but also it can be used to fund development, or create a paycheck for some teams doing some job for protocol, like Gauntlet itself. So irony is that they trying to chop a branch on which they are sitting themselves.

For those who might not fully understand, reserves actually work like a sort of a tax protocol puts on a revenue, it collects from Borrowers. And the Borrowers side is actually the ones who actually create a profits. Be it in speculations which they do with borrowed assets, or maybe they use Compound as a long term cheap funding for their business. That reserves rate is actually very small to be honest. If USDC, for example, have Borrowing APY about 2%, that 7% is taken from that 2% APY and the rest is distributed to suppliers. Do not be decieved by seemingly big numbers of current reserves, much of them were collected from vastly higher APY, which used to be even double digits at a times. It takes years to collect any meaningful amounts with such taxation rates. And Gauntlet servicies aren’t coming for free, not to mention there is also OpenZepellin with 1M$ every quarter. How the heck you guys imagine to be able to collect your own paychecks, especially when COMP tokens will run out, and they going to run out fast especially if markets go into prolonged bear, as the cheaper the COMP tokens become the more of them protocol would need to spend to cover costs.

It wasn’t very smart idea to rely only on COMP tokens to begin with, as they are generally limited, and it’s not in the best interest of protocol to spend it’s treasury tokens when price is very much in bear territory. But in any case they WILL run out eventually, and the costs for protocol WILL NOT disapper, they can actually go higher due to inflation. 3.86M per year of reserve growing ain’t even going to cover OpenZepellin alone not to mention anyone else.

I strongly advice to vote NO for such proposal, as it undermines protocol long term revenue, putting it in much worse position as it is currently. And to be frank, potential increase of yield for suppliers going to be as negligeble, that pretty much every individual supplier would need a microscope to notice it. Even if we take all the potential savings of 1.53M and put it exclusevely on stable coin markets, that would be somewhere around 0.05%APR increase, which of course, would’t make anybody to earn more, as if amount of suppliers increase, the APY will go down to reach pretty much similar balance point it was at before.

Suppliers are not a priority for Compound, they not earn anything, that is expenses that borrowers pay for capital. And if there is no borrowing demand, the suppliers have to go anyway, till yields can reach equilibrium at lower point.

If anything, protocol should consider increasing reserve factors on primary earning assets (stable coins) and unifying them to be in line with each other. The primary guidence here should be that rate of taxation on borrower payments should be low enough to be considered not prohibitive, which for stable coins is probably something not higher than 10%. For collateral assets like wbtc and eth that could be much higher, as those assets do not really produce any meaningful yield for supply side, and most suppliers would’t mind to not recieve those pennies at all, as it’s all in range of less than 0.1% APY. Thus reserve factor could be all the way to 50%. If some suppliers are not happy with not recieving yields they are very welcome to withdrow and go anywhere else, as for collateral markets there is negligeble borrowing demand, and if there are no borrowers, you don’t really need many Suppliers. Minor markets are negligeble for Compound as reserve assets, and thus reserve factor for small markets could (up to governance) potentially be zeroed without any significant downside for protocol.

Those people who plan to support decrease of protocol revenue, you have been warned. If we want to have a sustainable future-proof financial system, we couldn’t build it in same way as current pretty much collapsing financial architecture is. Protocol have to be able to generate enough profits to maintain it’s own development, bug bounties, risc analytics, and do that without relying on VC, or tokens issue or similar temporary measures. Think about it, and vote wisely.

[ClairvoyantLabs]

No one stood up to challenge Gauntlet. Gauntlet initiated many proposals with low utility by modifying tiny parameters, but the optimization of protocol revenue was negative, and cancelling the employment of Gauntlet was also a good proposal.
It can save a lot of money for compound protocol.

DeBank | Your DeFi wallet

By the way, this is the risk point that leads to huge bad debts in the protocol.

[rleshner]

@Sirokko makes a strong argument against Proposal 109. Currently, most markets are over-supplied, and decreasing reserve factors (to raise supply rates, and slow the growth of reserves) does not seem like the correct direction (at this moment). I plan to vote against Proposal 109, but I’m very excited and grateful that Gauntlet’s proposal is kicking off an important discussion, and the work @klinsuain is proposing is a needed improvement.

If there are more reserves than value at risk, it might benefit the safety of users to look at alternate approaches to manage excess reserves, e.g. reallocating from individual markets. A great example are the following markets:

1. BAT: $1.48m borrowed, $1.13m reserves
2. ZRX: $1.08m borrowed, $0.43m reserves
3. USDT: $487m borrowed, $3.07m reserves
4. USDC: $304m borrowed, $13.26m reserves
5. ETH: $11.73m borrowed, $0.79m reserves

Given the markets above, should the protocol consider converting/migrating reserves from BAT/ZRX into USDT, etc, which has the worst ratio of borrowing/reserves? Or to convert excess reserves from BAT/ZRX to be able to cover bad debt from decreasing price scenarios (USDC) and rising price scenarios (ETH)? Generally, the super-majority of borrowing demand (and risk) comes from borrowing stablecoins, which are the asset most likely to not be returned via bad debt.

In addition, I’m looking forward to the near-term release of Compound III which could have its own reserve needs that could be bootstrapped from the excess reserves in the current protocol.

[blck]

I was going to vote no anyway cause for the simple reason, the protocol currently pays more for services like openzeppelin and gauntlet than the reserves grow, so there is no point to reduce it further, but nice to see others have objections aswell.
The protocol will eventually run out of COMP, then we will have to use the reserves to be able pay for services.

[inkymaze]

Thanks for the constructive feedback @Sirokko @rleshner @blck. As you all have highlighted there are numerous considerations for the protocol now and in the future. What we hope to agree on is that active governance through new proposals is a prerequisite for the sustainability or bootstrapping of future markets mentioned above. This proposal makes tradeoffs, which if passed will provide additional data to glean from subsequent user behavior.

We are happy to assist or facilitate reversing these reductions and potentially increasing RFs in the short-run (weeks/months). In the long run, migrations and Comp III can and should be considered. We look forward to those discussions.

[RogerS]

@klinsuain

By implementing the recommended cuts, the protocol could reduce its estimated annual reserve growth rate from $3.86M to $2.33M

@klinsuain, appreciate your work in adding Reserve Factor to the simulations.

Question: If collateral asset prices keep falling, then borrow capacity (in raw USD) will decrease, and, as a result, protocol’s interest income will decrease too, even if the Reserve Factor stays constant. For example, if ETH falls 50% from here, will the estimated annual reserve growth rate of $3.86M still hold good? What assumptions are you using in this estimation?

Based on the current market conditions, and also because VAR doesn’t adequately account for long tail events, I agree with others that it’s best not to reduce the Reserve Factor at this time.

[klinsuain]

Short answer is yes. Like everything else in the protocol the borrow amounts and borrow APYs are constantly changing, and as a result the annual reserve growth rate of $3.86M that we quoted at the time of the analysis is also variable (currently it’s down to $2.68M). Worth noting that as borrows decrease and reserve growth rate decreases, expected insolvencies will likely decrease as well, offsetting the effects to some extent.

Every time we generate recs we do so based on the current projected annual reserves, expected insolvencies, and current reserve pool. Should expected insolvencies or reserves change materially enough to affect our recs, we’ll cancel the proposal and create a new one with updated reserve factors. These are easy levers to change as they only immediately affect supplier APYs and reserve growth, as opposed to collateral factor changes which can have an impact on whether or not accounts are liquidated.

[pauljlei]

Gauntlet is considering the community feedback above and will return to the community with an updated proposal.