One of the things everyone is always looking for is a way to get more for less. This is not something entirely new to humans, as being more efficient has always been something that we have strived for in our personal, professional, and romantic lives. But what does this have to do with yield, and what the hell does it have to do with donuts?

Donut is a protocol on Base that used a modified form of mining in order to distribute tokens, build liquidity, and generate revenue. At the time of writing, it has distributed nearly 95% of the tokens prior to tail emissions and has built significant liquidity for a protocol of its size (approximately 300k market cap at the time of writing). Both of these factors are important in building the foundation of the protocol. It has largely been a success at those two specific goals. But it has struggled in multiple ways as well. In this article I want to talk about some of the ways it struggled, potential solutions, and how those solutions can also be used to expand other protocols into generating sources of leveraged yield.

But first, how does Donut work? What's the TLDR? The primary aspect is the core mining protocol. They use gradual Dutch auctions in order to see who receives the emissions for that round. These rounds last an hour, and whoever the highest bidder is receives all emissions of the token until they get “dethroned” and a new highest bidder comes in. For each winning bid, the bid gets split in a few ways. 80% of the bid goes to the previous highest bidder, with the remaining 20% being split among things like the treasury, liquidity, and frontend providers (if they charge a fee). As time goes on, emissions decrease. This is a good way to have early users rewarded with heavier emissions, as they are taking on more risk. However, I think there is one big issue here that needs to be addressed.

As with any new token when it is starting out, it is viewed as inherently worthless, and it is the team's responsibility to create value. As much as people don't like to believe this, creating value for users is extremely difficult, especially in crypto. This is not a Donut-specific thing, this just applies to almost all tokens. Donut is at the point where they now have to prove their value (and they are building out products and working towards that). But for the vast majority of tokens this is not the case, causing later investors to be hesitant or not invest at all, as there is a significant risk that in a week, month, or year they will be significantly down from their initial investment.

Tokenomics is a really hard problem to solve. There are so many different coordination factors and external things that come into play that it can be difficult to figure out exactly how to design them. Hindsight is also 20/20, and it is much easier to criticize than to be forward-thinking about every potential situation and scenario that could potentially come up.

But with a little bit of hindsight, what solution would I propose in order to help prevent some of the issues that are currently going on with tokens that are following a similar mining pattern?

I think there is one major thing that could solve a large portion of this problem. Do away with the 80% reward to the previous miner and replace it with a call option. Essentially making it so that the 20% (or whatever arbitrary percentage) that is going to the treasury, liquidity, etc., can be viewed as a premium. Giving the “miner” (or buyers off market as well) the option to sell on the open market or redeem the tokens later.

Psychologically for the miner it is essentially identical. They are still paying to receive tokens. However, for other investors who are not these miners it creates an entirely different psychological dynamic. A much better dynamic in my opinion. One issue with this is that typically when you structure tokens in this way they become psychologically tied down by the “backing” or redemption price. So how does this get solved?

I think there are two primary mechanisms to help with this. The obvious answer is to generate additional revenue that is used to generate a “market premium.” But that is not what I am here to talk about.

The way pricing worked for Donut mining was that whatever the price paid during the last auction was then doubled, and that was the new starting price for the next auction. Over the next hour or so the price would slowly decay towards zero until it reached a point that whoever the next buyer was felt it was a fair price to pay.

This works well for their purposes but not for ours, as we have redemption value to uphold. If we let the price of the emissions tend towards zero then ultimately the redemption value can do the same.

So rather than having the ending price be zero, it should be the redemption value per token. This changes the dynamic slightly from pure “mining” to more repetitive auctions. Rather than streaming emissions, they are grouped together hour by hour (or whatever else), and one user can buy all of the emissions that would have been emitted for that hour at whichever price they decide. If no one buys them at the given price, then the protocol can choose whatever happens to those tokens. They can be stored to be sold in auction later when demand is high enough, burned, or whatever else.

This helps ensure that the baseline continues to grow over time. Also an important note: if you are selling 1000 tokens in the auction and the baseline for these is currently at 1 of X per token, the end price is not actually 1000 of token X to win the auction. It would need to be higher to help cover the premium that goes to other things. If this is not the case it also brings down the redemption value, which should never be possible.

So why would people even want this token in the first place? Generating yield is often a good way to incentivize people to stick around. And while the example I will give is going to be Donut-specific, it can work for any yield-bearing token, and in many cases would work even better for those that do not require a lock-up period.

How does the yield become eventually leveraged? Well, it is technically leveraged the entire time, but at the beginning it can be considered inversely leveraged (less than 1x leverage), as only a percentage of the mining fees are going to generating yield.

For tokens with no lockups generating yield on tokens like USDC or others it is closer to 1x, but for this example let's say 10% of mining fees in Donut get staked.

This creates a leverage of 0.1. This is not exactly true, but it's easiest to think of it like this to begin with. However, over time this can increase quite quickly due to a number of factors. There are several ways to increase it and show that yield leverage is not actually that low to begin with.

For example:

• Not everyone will stake

• There will always be tokens in the liquidity pool, so it can never be as low as the percentage given here

• Redemptions

Starting with the first two points as they tie together, there will never be a time in which every single token is staked. Staying with Donut as the example, currently only about ~25% of tokens are staked. That would move up this leverage example to around 0.4.

The real accelerator in this system however is the redemptions. Over time people leave the system, that is just inherently what happens. They have two options here: sell (what most people will do) or redeem. Most of the time people will just sell as it is easier and quicker. This will drive down price if there are not enough buyers to eclipse this.

Leading to the possibility of the price being lower than the redemption price. In which case multiple things could happen. Smarter users who wish to sell will just redeem as they get a better price, or arbitrageurs will take advantage of this risk-free profit and buy on the market and redeem for a small reward.

Users can also buy and stake in order to increase their yield leverage, but that is a different topic. The key thing is that during redemption the assets that are yield generating remain untouched, making it so that for every token left its share of these assets grows. As time goes on and this process repeats it continues to grow, never decreasing.

Now this setup can work for any token or yield-bearing token, and in cases in which there is not a lock-up to earn yield even the baseline can be placed in yield-generating solutions in order to help increase that faster as well. This obviously comes with additional complexity and smart contract risk, but it is an option.

This is just one idea and small tweak on the mechanism that can be applied. I am sure there are many others as well. The Donut mining protocol can be expanded and used in a variety of different ways. I still believe that while nothing is perfect, this is a significant improvement for investors and protocols compared to a simple Pump.fun-style launch.

As time goes on I do plan to implement some of these ideas and hopefully continue to iterate even more, improving not only the launching mechanisms and options for new protocols but also the Donut ecosystem.

As a final note, tokenomics do not fix a protocol entirely and it still needs to find a way to generate revenue and handle marketing and all of the other things that make running any business (in crypto or outside of it) very difficult. However, the right structure can do a lot in helping the success of a protocol, as well as helping users and investors manage downside risk.