# Some thoughts on MVI

***tl;dr**: while minimum viable issuance is, by definition, a good idea, if we are to minimize the risk of unexpected negative externalities we need to start viewing viability through a more multidimensional lens. It’s not enough to simply look at it through the lens of economic security.*

***Thanks** to Vasiliy Shapovalov, Sam Kozin, Barnabé Monnot, Justin Drake, Mike Neuder, Jon Charbonneau, Myles O'Neil, Vishal Talasani, Sam Hart, Hasu, Arjun Balaji, and Jasper the friendly ghost for reading drafts of this. and in particular to Vasiliy and Sam for the thought-provoking conversations that gave birth to it*


The main arguments Anders [puts forward](https://twitter.com/weboftrees/status/1710704461750944190) in his mammoth thread on minimal viable issuance are solid if you agree with the underlying axioms -- issuance should be reduced as much as possible, ETH (not stETH) should be the money of Ethereum, and stakers should be a minority to allow for frictionless forking of bad actors. But it feels like there is a lack of adequate exploration of the tradeoffs at hand, which stems from an overemphasis on economic security as the guiding metric.

Focusing too much on economic security in this way increases the risk of negative externalities that could make the outcome worse than the status quo. In particular, one can sense an underlying (perhaps subconscious) [power struggle](https://gwart.substack.com/p/the-real-fork) around the moneyness of ETH vs stETH, which is muddying the waters somewhat.

In a response to Jasper's [concerns](https://twitter.com/drjasper_eth/status/1710712801633739011)

> I have some serious concerns about the idea that this will reduce the amount of LSTs Vs unstaked ETH.
>
> “In the event that we reduce the yield, SSPs would hopefully need to increase fees to properly cover and amortize costs. The delegating staker’s costs are variable, including the PAP and fees. They can walk away from increased fees with very little friction.”
>
> I see no reason for this to be true. Why would fees go up if the yield goes down? If they are profit maximizing they will beat out the competition and slowly extract higher fees.
>
> Further, the section on restaking seems misguided. It may boost solo stakers but most AVS services are likely geared towards professional stakers and further LSTs can be used as collateral. I don’t think AVS can save solo stakers, especially if the overall incentive drops.
>
>Simply put, there is no reason to hold ETH over an LST. People suck at evaluating tail risk, and likely will pile into one provider leading to slashing loss, but then the stake will just shuffle.
>
> There isn’t really any reason that ETH would be more useful as an asset. It has more liquidity on CEX but that’s not going to last forever if a dominant LST emerges and eventually consumes the ETH volume.

Anders [writes](https://twitter.com/weboftrees/status/1710761999951352163) (in one passage of his answer):

> "Higher issuance leads to the money function becoming a stratum for cartelization making it much more attractive to first drive out competitors and then increase fees once you have a monopoly. If we were to moderate issuance so that the deposit ratio stops growing, LSTs cannot compete with ETH on the money function and may find non-monopolizing strategies more profitable. If an SSPs proportion of all ETH is not too high, the risk of a moral hazard is reduced because they cannot become too big to fail. **My hope is that, at margin, security and value-alignment will then be more important properties for delegators when they pick their SSP [staking service provider]."**

With all due respect, leaning on these sorts of unrealistic user ideals feels like the same category of mistake which led to exchanges dominating before Lido launched, and to the mistaken expectations some researchers held regarding the potential of home stakers to take significant share away from Lido.


Practically speaking, under MVI, lower rewards will mean greater (if not exclusive) competition on margins. Economies of scale on cost will, in all likelihood, end up winning out, and bigger players will be in a position to either price out or outlast smaller ones.

To unpack this, I believe that the following three things will almost surely be true:

(1) smaller staking services (including solo staking) will become less attractive[^1](/qyaKX-JBSpOc70t8_sOEFA)

(2) staking providers will be forced to compete on fees and yield

Centralized entities are better suited to compete on fees and yield due to fewer operational costs, vertical integration, and non-staking revenue sources (for example, CEXes make most of their revenue outside of staking). It's certainly not a stretch to imagine that such an issuance change could play into the hands of a few centralized entities or a [collective](https://twitter.com/ViktorBunin/status/1633213347255681024) that can afford to be loss making on the staking front in exchange for greater consensus power.

Additionally, restaking and other forms of leverage will likely become more attractive relative to just validating on Ethereum since the rewards that flow from restaking will become more significant as a percentage of total rewards. This will cause the restaking market to grow faster, and the associated leverage risks to grow with it too.

Last but not least, from a [memetic perspective](https://twitter.com/iamDCinvestor/status/1621192574538063873), staking yield has proved a great way to attract people to Ethereum today -- staking sounds a lot more attractive than holding to many people even if rationally speaking it’s not all that different; folks simply enjoy feeling like they are [generating a return]((https://twitter.com/jp_koning/status/1711714224664186943)) while [contributing to something important](https://twitter.com/udiWertheimer/status/1410373397989793793).


Anyone who's spent significant time at the application layer will tell you that onboarding new people to crypto with simple and easy products is one of the [hardest things to do](https://twitter.com/js_horne/status/1712186770744361166). I expect it will be much harder for [ETH to compete with USD](https://twitter.com/ASvanevik/status/1700330111156011168) as the heart of DeFi without a simple earn product -- like [staking](https://twitter.com/iamDCinvestor/status/1646495115156832257).

Lido's referral program, for example, has brought in ∼[86k](https://dune.com/queries/3113489?category=abstraction) new users (with ∼67% still holding stETH 12 months later). It is doubtful whether it would have been even half as effective without meaningful staking rewards. In no small part because [individual investors](https://www.sciencedirect.com/science/article/abs/pii/S0148619507000392 ) are -- somewhat irrationally -- more willing to [pay fees](https://www.etftrends.com/model-portfolio-channel/dividends-traditional-vs-behavioral-finance/?utm_source=Yahoo&utm_medium=referral&utm_campaign=ReadMore) on earnings than [on principal](https://kanebridgenews.sg/a-dollar-is-a-dollar-is-a-dollar-except-in-our-minds/) (e.g. see [Thaler on mental accounting](https://onlinelibrary.wiley.com/doi/abs/10.1002/%28SICI%291099-0771%28199909%2912%3A3%3C183%3A%3AAID-BDM318%3E3.0.CO%3B2-F)).


Ethereum staking can be thought of as the second largest DeFi app in all of the industry (the largest being USDT), both by TVL and value generated. It acts as a sort of fertile soil which supports the growth of a plurality of actors building on top: from LSTs to money markets to stablecoins to user channel providers (be they wallets, neobanks, CEXes, hedge funds, or ETFs).

Concretely, staking rewards provide a great stream of value to:
1. offer a good product to less sophisticated people who are eth-curious (where serious != sophisticated)
2. support an industry that can package, explain and deliver this product to people (wallets, financial products like ETFs, etc)

Reasonable people can disagree over whether or not this is worth the printed issuance, but it's important to keep in mind that there aren't a million other ways to support this industry.

While it's very hard to disentangle correlation from causation, in Lido's case there are two important metrics that stand out for me: ∼40% of total stETH supply is further utilized in DeFi (predominantly in lending and liquidity markets), and you're 100x more likely to be using Aave and Maker if you hold stETH vs just ETH.[^2](/Sg022TS1Q9CMBrjLtc_anw)


In sum, the memetic argument can be summarized as follows:

1. users are attracted to crypto in no small part because of the potential for rewards (e.g. [coinbase earn](https://www.coinbase.com/earn) has brought in > [6.5M](https://twitter.com/zosegal/status/1442550078237929477) users).
2. if we reduce staking rewards, then staking will become less attractive to the average user -- Thaler's [mental accounting concept](https://www.cairn.info/revue-d-economie-politique-2008-4-page-475.htm) helps explains why individuals tend to prefer rewards to capital gains (even if it's not economically rational to do so).
3. making staking less attractive will have important downstream consequences on the rest of Ethereum DeFi -- in particular it's likely to play in the favour of USD centric DeFi (via [simpler USD earn](https://twitter.com/Pledditor/status/1687494000662441984) apps); as a result, it may be harder for ETH to compete with USD as money.


To conclude, while MVI is a noble and worthy goal, overindexing on *economic* security may well [cool down DeFi](https://twitter.com/MacroMate8/status/1689006327878840320), [slow down onramping](https://twitter.com/poordart/status/1572914685783183360), hurt solo stakers, push LST providers to vertically integrate, reduce monetization opportunities for channel providers, and make wallets more likely to seek different monetization schemes (like MEV). If we are to minimize the risk of unexpected negative externalities we need to start viewing viability through a more multidimensional lens. It’s not enough to simply look at it through the lens of economic security.