MakerDAO Governance & Risk Call: Stability Fee hike post-mortem

This free preview of The Block Genesis is offered to our loyal readers as a representation of the valuable journalism & research our team produces and Genesis members receive daily. Want to see more? Join today.

The following is the first edition of a weekly series summarizing and analyzing the MakerDAO Governance & Risk calls.

This week’s MakerDAO Governance & Risk call came with a greater sense of urgency than previous congregations.

Over the past few weeks Dai, the Ethereum-based, overcollateralized stablecoin, has consistently traded below its intended $1 peg, leading MakerDAO stakeholders to raise the Stability Fee by 2% APR to 3.5% APR, the largest single fee hike in the decentralized credit facility’s brief, five-vote history.


The call commenced with a discussion of the governance process. MakerDAO is ostensibly structured as a Decentralized Autonomous Organization, a kind of digital co-operative, where proposed changes to risk parameters must be voted on, and passed, by a wide set of community participants.

MakerDAO’s governance process can be divided into four distinct processes:

1. A motion is proposed on a weekly call and discussed and debated among the call’s participants. For the most part, motions are tabled by MakerDAO’s selected Risk Management team.

2. If there is a broad sense of consensus or interest around the motion, the discussion moves to social media channels — predominantly Reddit — where it can be dissected and debated further.


3. The motion then moves to Polling, a portal through which MKR holders can signal their preferences by staking their tokens in favour of the motion versus a default option of maintaining the status quo. The Polling process is currently configured to last three days, but can be adjusted through a vote.

4. If the motion passes through Polling, it moves to the final stage — Executive Vote. Again, MKR holders can vote to pass the new proposal by staking their tokens in its favour. Executive Voting is a continuous voting process: in order for the new proposal to pass, it must simply receive more staked MKR in its favour than the previously passed proposal.


The proposal to raise the Stability Fee by 2% APR passed smoothly, with 98,169.27 MKR, valued at approximately $65.77 million, staking in its favour after just 10 hours and 27 minutes. The ability to coordinate on, and proceedingly pass, such a significant parameter change in such a short period of time should ostensibly be considered an indication of MakerDAO’s triumph as a Decentralized Autonomous Organization. As the Maker economy continues to grow and as Dai’s stability becomes subject to an increasing number of exogenous factors, the ability to act swiftly will be crucial to the system’s success.  

However, on closer inspection, the governance process still leaves a lot to be desired. The chart below reveals that just 37 addresses out of 9,668 MKR holders voted in favour of the hike and a single address contributed 57% of the stake. It should be noted that the MakerDAO development fund has committed not to participate in voting and the second largest MKR holder, a16z, is yet to adjust its custody solution in a way that would support voting from a technical perspective.

Source: Richard Brown, MakerDAO Head of Community Development

The low turnout can likely be explained by a combination of factors:

1. Perhaps most significantly, there seems to be a sense of apathy among smaller token voters. Indeed, apathy is generally considered the primary reason for low turnout across national elections.

Whether a small MKR holder is in favour or against the proposal is largely irrelevant: if they feel that their stake is not large enough to have any significant influence over the outcome of the vote they are unlikely to participate.

I imagine this sense of apathy is particularly acute in the MakerDAO ecosystem, where the top 100 accounts collectively own 91.52% of tokens.


2. Participating in voting requires staking one’s tokens in the MakerDAO Governance Portal. Tokens cannot be transferred or sold while staked, so voters lose liquidity by participating in the governance process. Additionally, there is presumably some non-trivial degree of platform risk associated with moving one’s tokens to the Governance Portal, especially compared with an alternative of keeping tokens in cold storage.

3. Setting monetary policy for a decentralized stablecoin is not a particularly straightforward process, so it should not be surprising if many MKR holders do not consider themselves qualified to be involved in the decision making process.

4. The binary nature of the vote — raise the Stability Fee by 2% APR or maintain a rate of 1.5% APR — may have been unappealing to certain stakeholders who supported a smaller or larger hike.

As noted on the call, MakerDAO is still in the early stages of its life cycle and the governance process is likely to be adapted in the future to acclimatize to these factors. Indeed, one underappreciated aspect of MakerDAO is its DAO-like structure: the remarkable growth and success of Dai to date has led many to forget that the project is the first large scale system to operate in a reasonably decentralized, one-stake-one-vote fashion.

The factors listed above — apathy, difficulty, insufficient information, and displeasure at a restricted set of options — afflict all formal elections, but are especially pertinent in one-stake-one-vote systems. As the MakerDAO governance process continues to adjust, blockchain projects implementing formal on-chain governance should take careful note of MakerDAO’s difficulties and the effectiveness of its ensuing remedies.

Participants on the call proposed four reasonable alterations that could reasonably engender increased turnout:

1. Introduce an economic incentive to participate in voting. This could take the form of either an inflationary reward for voters, or a programmatic slashing for those that do not vote: the end result is practically identical.

2. Build an application that notifies MKR holders of impending votes and regularly updates holders of the latest outcome.

3. Provide empirically-driven justifications under each proposal.

4. Allow MKR holders to vote without staking their tokens, thereby avoiding gas costs and reducing exposure to platform risk.

A fifth alteration, not mentioned on the call, would be to extend the Governance & Risk conversations beyond an hour per week to allow for further in-depth analysis of the issues at hand. That most people currently taking part on the calls are either directly or indirectly involved with MakerDAO suggests that time constraints should not be a particular concern. Of course, a two-hour phone call becomes less sustainable as the ecosystem continues to grow, although the growth of stakeholders raises its own logistical questions: how can, say, 5,000 people, or 50,000 people, concurrently meaningfully participate in a phone call?

However, as MakerDAO Founder Rune Christensen deftly noted, percentage participation isn’t necessarily an accurate proxy for assessing the success of the system’s governance process as it is impossible to identify what proportion of voters acted maliciously. A proposal could theoretically pass with a large majority and yet ultimately be to the detriment of Dai’s stability.

Rather, the most important governance mechanism is the Emergency Shutdown, which, when triggered, collapses the entire system and refunds Dai holders in Ether. Emergency Shutdown must also be voted on by MKR holders – although currently it can only be activated by a select few individuals — and the necessary threshold is expected to be relatively low. As a result, it might be appropriate to shift emphasis away from low Executive Vote turnout towards making sure that there is always a sufficient number of stake ready to defend against ‘bad’ governance.


The conversation proceeded to evaluate the effectiveness of Saturday’s Stability Fee hike, with MakerDAO’s Head of Risk Management, Cyrus Younessi, leading the discussion.

Before continuing, it’s worth remembering that the intended effect of the Stability Fee hike was to concurrently increase demand for Dai and decrease outstanding supply, as the added cost of debt encourages Collateralized Debt Position holders to buy Dai on the market and use it to close out their loans. If successful, an expansion of demand and contraction of supply would drive the price of Dai back up to its $1 peg.

Dai supply has increased steadily since the interest rate increase, up approximately 0.82% in the past six days. For context, Dai supply is up 30.07% in the past two months, with a week-on-week growth rate of 3.57%.   


Meanwhile, aggregate data from eth2dai and the 0x network of relayers, together considered the most active lit liquidity pools (aside from Uniswap), indicates that the price of Dai has steadily trended upwards since March 3rd’s local lows of $0.94.

Source: Vishesh Choudhry

Unfortunately, due to the presence of multiple exogenous factors, it is difficult to quantitatively measure the effectiveness of the stability fee hike alone. That outstanding Dai supply has increased suggests that raise has not had direct impact on Dai’s stability, although it could be that speculators and CDP holders bid up Dai in anticipation of an impending return to the peg.

There are two other likely explanations:

1. Dai price is primarily led by Ether activity, and Ether’s slight dip over the past week has seen a resultant increase in demand for Dai. This argument is convincing when you consider the overwhelming use case for CDPs right now is to leverage long Ether: if Ether falls, those who have levered long will sell their Ether for Dai and conversely, when Ether rises, speculators will sell their Dai for Ether. The chart below corroborates this assertion, revealing an inverse correlation between ETH/USD and DAI/USD price activity.


2. Choudhry’s Volume Weighted Average Price chart shows that volumes have declined over the past week, meaning that comparatively smaller exchange activity will have a disproportionate effect on Dai price. As such, MakerDAO stakeholders should avoid drawing conclusions from recent price activity.

[related id=1] As previously discussed inDai falls below dollar peg, kicking off a monetary policy quandary, there are two other metrics that may provide a more accurate picture of the Stability Fee hike’s effect:

1. Perhaps the most accurate metric is to look at market maker and OTC desk Dai inventory. Dai inventory relative to Ether is high when the market is net selling more Dai, and conversely, Dai inventory is low when the market is net buying more Dai. Representatives from the MakerDAO Foundation’s trading desk and Wyre Capital reported that inventory has remained largely the same since the hike.

As discussed on the call, it might benefit MakerDAO to set up a zero-knowledge method through which market makers can share precise data without revealing their exact positions.

2. A second valuable metric is Dai borrow rates across alternative lending platforms and to form an aggregate reference rate, which I have provisionally titled the Decentralized Inter-Protocol Offered Rate (DIPOR). Borrow rates shift with supply, so an increase in borrow rates should indicate that supply has fallen and the Stability Fee has had its desired effect. Conversely, a low DIPOR would suggest that there remains a glut of Dai supply, and so the Stability Fee needs to be adjusted further.

Several call participants suggested that the data presented warrants another stability fee hike. I disagree with this assessment.

Interest rate increases should only be made based on significant volume and it is still too early since the last hike to determine what its effect has been. It seems misguided to expect a hike to have an instantaneous effect on Dai demand: if you consider the median CDP loan size to be 500 DAI, increasing the stability fee by 2% APR only leads to an extra 0.19 DAI fee per week. It is reasonable to expect CDP holders to take some time to feel the effects of any interest rate.

Source: Placeholder Capital

Of course, any stability fee hike has to go through the polling and executive vote process, and, in theory, those MKR holders that disagree can simply stake their tokens in favour of the status quo. In practice, new proposals act as a Schelling point, with their very presence seemingly legitimizing their validity. As Christensen noted on the call, the voting process is actually a technical one rather than a decision making process, so voters who do not have strong opinions but are still willing to participate will likely gravitate towards the most recent proposal.

For now, no new polls have been initiated. Conversation is likely to continue on the MKRGov subreddit over the coming week. For those who feel strongly about shaping the course of MakerDAO’s monetary policy, I encourage you to share your thoughts and feedback there.


Reach out at: [email protected]

MakerDAO Governance & Risk Call: Stability Fee hike post-mortem

written by Matteo Leibowitz @ March 15, 2019 Matteo Leibowitz

Comments are closed.