Governance Request: Fair Withdrawal Mechanism for the Illiquid Base mUSDC Market

Background

I am raising this as a separate governance issue following guidance from Moonwell moderation, who correctly noted that decisions around withdrawal treatment and remediation should be addressed through governance.

I currently have approximately $44,000 USDC supplied to the affected Base mUSDC market.

I am not asking for preferential treatment because my position is larger.

My concern is about whether the current withdrawal mechanism can unintentionally create an unfair allocation of the remaining shortfall among suppliers.

The core issue

The Base mUSDC market currently has effectively no persistent available liquidity.

At the same time, approximately $8.7M of active/performing USDC loans remain outstanding, in addition to approximately $2.5M of current bad debt.

As performing borrowers repay, liquidity can temporarily return to the market.

However, we have already observed that when relatively small amounts of liquidity become available, they can be withdrawn almost immediately.

Under the current mechanism, withdrawals therefore effectively operate on a:

first-come, first-served basis.

This creates a potentially serious fairness problem.

A supplier with a small position may be able to withdraw their entire balance during a relatively small liquidity event.

A supplier with a larger position may repeatedly be unable to exit because sufficient liquidity for the full position is rarely available at once.

Over time, this can create a path-dependent outcome:

Suppliers who happen to withdraw earlier reduce or eliminate their exposure, while the unresolved bad debt becomes increasingly concentrated among the suppliers who remain in the market.

The final allocation of losses could therefore depend not on a governance decision or a defined loss-sharing framework, but simply on who was able to click Withdraw first when liquidity appeared.

I do not believe that should be the intended outcome.

Simple example

Assume a market has:

  • $8.7M of performing loans
  • $2.5M of bad debt
  • near-zero cash

Suppose performing borrowers gradually repay several million USDC over the coming weeks.

If every incoming repayment is immediately withdrawn by whichever suppliers reach the liquidity first, those suppliers leave with 100% of their redeemed amount.

The remaining suppliers progressively become a smaller pool supporting the same unresolved shortfall.

Eventually, if the remaining bad debt has not yet been fully remediated, the last suppliers in the market could carry a disproportionately large effective exposure to that shortfall.

That is very different from an explicitly approved pro-rata loss allocation.

Why larger suppliers may be disproportionately affected

This is not an argument that larger suppliers deserve priority.

It is almost the opposite.

When liquidity arrives in small increments, larger positions can be structurally harder to exit.

For example, someone with $500 supplied may be able to exit during a $1,000 liquidity event.

Someone with $44,000 supplied cannot.

If this happens repeatedly, smaller positions can disappear from the market while larger positions remain trapped.

The result may unintentionally penalize suppliers who provided more liquidity to the protocol.

Post-incident suppliers make this issue even more important

There is an additional concern for users who supplied after the August 27 incident while deposits remained enabled.

If post-incident suppliers entered a market that already contained a known shortfall, and subsequent withdrawals are allowed purely on a first-come, first-served basis, then the eventual distribution of that pre-existing shortfall could become highly dependent on withdrawal timing.

I believe governance should explicitly clarify how this is intended to be handled.

Questions for Governance

I would appreciate a clear discussion of the following:

1. Is the current first-come, first-served withdrawal behavior considered acceptable during this remediation period?

2. Has governance assessed whether continued unrestricted withdrawals could concentrate the remaining bad debt among the last suppliers in the market?

3. Is any snapshot being considered to identify suppliers affected at a particular block/date, especially post-incident suppliers?

4. Is a pro-rata withdrawal mechanism technically or operationally possible while the market remains impaired?

For example, incoming liquidity could potentially be distributed proportionally among affected suppliers rather than being entirely captured by whoever withdraws first.

5. If pro-rata withdrawals are not practical, are alternative mechanisms being considered?

Possible approaches could include:

  • periodic pro-rata redemption windows,
  • withdrawal queues,
  • per-period redemption limits,
  • claim accounting based on a snapshot,
  • partial redemption rights,
  • or another mechanism designed by contributors/risk managers.

I am not prescribing a specific solution. I am asking governance to evaluate the problem.

6. Most importantly:

Can governance confirm that the unresolved shortfall will not simply end up being borne by whichever suppliers happen to remain in the market after others withdraw?

Why I believe this deserves a governance decision

The current market is not operating under normal liquidity conditions.

It contains known bad debt, borrow activity has been restricted, MIP-X66 has been executed, reserves are being considered for remediation, and further recovery actions are still under discussion.

In that environment, the normal withdrawal mechanism may produce outcomes that were never designed for a distressed market.

A normal first-come, first-served redemption model may be appropriate in a healthy lending market.

It may not be appropriate when there is a known and unresolved shortfall.

This is therefore not simply a liquidity question.

It is a question of:

how losses and recovery are ultimately allocated among Moonwell suppliers.

Disclosure

I have approximately $44,000 USDC currently affected by this market.

I am disclosing my position for transparency.

Again, I am not requesting priority for larger accounts.

I am requesting a mechanism that prevents withdrawal timing alone from determining who ultimately bears the remaining loss.

I would appreciate input from Moonwell contributors, Anthias Labs, delegates, and other affected suppliers on whether governance should formalize a fair withdrawal framework while remediation is ongoing.

Dr. Bahmani
Affected mUSDC supplier

Follow-up: today’s ~$250k repayment demonstrated this issue in practice

I’d like to add a real example from today that illustrates the concern raised in this thread.

A transaction shared by Moonwell moderation called Repay Borrow Behalf on the Base mUSDC Core Market.

At roughly the same time, Anthias data changed from approximately:

  • USDC bad debt: $2.554M
  • to USDC bad debt: $2.308M

That is a reduction of approximately $245.6k.

Based on the transaction and the corresponding change in Anthias data, this appears to be the protocol reserve repayment associated with MIP-X66, rather than a normal supplier deposit.

This is clearly a positive development: real capital is now being used to reduce the shortfall.

However, it also demonstrated the withdrawal problem discussed in this thread.

According to users monitoring the market, the liquidity created by this repayment was consumed within seconds.

I personally monitor the mUSDC liquidity closely, but I missed the entire event because I was simply away from my system at that moment.

This raises an important practical issue:

If meaningful liquidity can appear and disappear within seconds, access to withdrawals effectively depends on who happens to be online at that exact moment — or who is using automated bots.

Normal suppliers cannot reasonably monitor the market 24/7.

And automated users may have a significant advantage over manual users in capturing newly available liquidity.

For larger positions, this problem becomes even more significant, because a supplier may need multiple substantial liquidity events before being able to exit.

So today’s repayment reinforces the original governance question:

Should liquidity generated by remediation, protocol reserves, recoveries, or other extraordinary measures be distributed purely through the normal first-come, first-served redemption mechanism?

Or should governance consider a more equitable mechanism for remediation-related liquidity, such as:

  • pro-rata distribution,
  • a withdrawal queue,
  • snapshot-based claims,
  • scheduled redemption windows,
  • or another mechanism that does not reward transaction speed or automated bots?

I want to emphasize that the ~$245k reduction in bad debt is a positive and encouraging development.

My concern is specifically about how remediation liquidity is distributed once it reaches the market.

If governance is using protocol-controlled resources to repair an impaired market, I believe it is worth discussing whether that liquidity should also be distributed through a mechanism designed specifically for an impaired market.

Follow-up: withdrawal rules should adapt to bots and AI agents

Today’s remediation event highlighted another issue that I think governance should consider.

The current withdrawal mechanism is effectively first-come, first-served. That may have been acceptable when most users were interacting manually.

But DeFi has changed.

Bots and AI agents can now monitor liquidity continuously and submit transactions within seconds. If remediation liquidity appears only briefly, automated users can have a major advantage over normal suppliers.

That means access to recovery liquidity can become a competition based on automation speed, rather than fairness among affected suppliers.

I do not think the solution is to restrict technology.
I think the protocol rules should evolve with the technology.

For remediation-related liquidity, governance could consider mechanisms such as:

  • a 24-hour claim window
  • pro-rata distribution
  • a temporary withdrawal queue
  • or another system that prevents recovery funds from being captured only by the fastest bots

My concern is simple:

Recovery should not depend on who has the fastest automation.

If Moonwell is adapting its protocol to new technical and market risks, I believe its withdrawal and remediation framework should also adapt to the reality of bots and AI agents.

1 Like