I’m going to write a post that’s critical because that’s what I do 
but I am glad you put together a very thorough outline of your thoughts, and there is a lot I agree with, but I think it’s important to focus on criticisms because that’s where the change needs to be IMO. Most of what you said I completely agree with, it’s just the finer points around dex liquidity, how we use bonds, and how we use the treasury I disagree with, so I’m writing a lot here, but it’s because I think the topic is complex and I believe a lot needs to be said to justify my point, not because I think everything you said is completely wrong. The first five seconds of this video is one of my favorite philosophies around business: https://www.youtube.com/watch?v=GltlJO56S1g and I think what you’ve outined around business philosophy is in perfect alignment with my philosophy, and I do have great conviction that a customer focused and utility focused business philosophy is a key to success.
So I want to focus on the topic of bootstrapping liquidity, the transition from a “bootstrap” phase to a “long term” phase, and what the goals and responsibilities for our exchange product is.
On utility, I agree broadly that utility for Silk is the most important thing, but I would like to clarify that exchange liquidity is only one form of utility. Other forms of utility are things like Silk Savings, synthetics using Silk as collateral, deposits in stability pools, etc, so liquidity is not the only route to utility. Of course liquidity is important though, I just don’t want liquidity to be framed as the only option we have so that it’s clear that we are not failing at creating utility for Silk if we don’t choose to pursue liquidity at all costs.
How do we bootstrap liquidity? My recommendation is a combination of bonds and LP emissions on exclusively USDC/SILK, with a bias towards bonds over the long haul of the protocol’s existence.
On this point, the idea of bootstrapping liquidity and “using bonds over the long haul” are incongruent. Bootstrapping fundamentally is only about how you get from zero to one, and the strategies for “how we get initial liqudiity” are not the same as “how we keep liquidity”. As I brought up in the waffle vs bird thread, an exchange is a meeting point for willing buyers and sellers, it’s the job of those willing buyers and sellers (most specifically, the subset of those buyers known as market makers) to create liquidity. Building an exchange and providing liquidity are different business operations. So I do not agree with you that it is our responsibility to provide and guarantee liquidity, it is our responsibility to attract willing participants to provide that liquidity for us. The biggest reason for this is because a single entity cannot create sufficient liquidity across a number of pools to enable trading, and because market making is a an extremely specialized skillset.
Caterpillar makes construction equipment, but they don’t focus their business on buying land and building homes, they make the equipment that enables other people to do so. They don’t buy forests and have a timber agriculture wing to produce more lumber to create more construction activity. They don’t own oil refineries so that customers can purchase their equipment and their fuel from Caterpillar. Caterpillar makes industrial equipment, and that’s what they focus on.
Zim Integrated Shipping is a marine shipping company. They ship products over the ocean across continents. They don’t also invest in manufacturing in China so they have more products to ship on their shipping lines. They don’t buy factories and rent them to people so they have more end products to ship. They exist to provide a service to people who need shipping services. They invest in buying more boats to ship more products. They invest in staff to run their boats so they can ship more products.
An exchange is a central forum for buyers and sellers to meet up and exchange. Our job is not to spend our money to give people tokens to buy and sell. Our job is to make a product that allows people to come to Shade Swap to make their exchanges. I believe your argument is that us providing liquidity achieves that. To present my argument I’ll use an analogy I originally wrote in telegram: everyone tends to agree that helping homeless veterans is a net good to society, but nobody would agree to spend a million dollars to give one guy a pair of socks. That is the situation we find ourselves in. Using protocol funds to provide a meager amount of liquidity is spending a million dollars to buy a guy a pair of socks, because that amount of money is instead better spent on investments in the company to create a more capital efficient exchange pool, or a more marketing to attract more willing buyers and sellers, or more primitives to create more utility and “things to do on Secret” which creates more value, or novel finance products that allow the creation of more liquidity through leverage.
Note that 902,000 SHD have been earmarked for Y1 LP.
I will say that while this is true, it doesn’t have to be. That money can be chosen to be spent elsewhere, so the ROIC needs to be there compared to the alternative options (opportunity cost). I know you explore this in your post. I just want to make that point that we don’t have to be married to this. If we can find a better way to spend 902k SHD then we should spend it on the better things.
Once we turn off stkd-SCRT / sSCRT and focus solely on USDC/SILK with 2.1k SHD per day (800k SHD / 365) we would attract roughly x20 the amount of liquidity on the USDC/SILK pair as the stkd-SCRT / sSCRT pair
2.1k SHD per day is an enormous amount of SHD and that will create a huge amount of sell pressure so that yield will drop fast (this happens to literally every single defi token. throwing insane amounts of money at a pool has never created sustainable yield. not once, ever.) I also don’t really understand where the 20x number is coming from. 70 SHD per day (you said week, isn’t it per day?) is 1/30th of 2.1k SHD per day, so are you just assuming 30x more emissions is roughly going to make 20x more liquidity? From what I’ve observed the relationship between APR and liquidity is an S-curve. The first few points don’t attract much, the next few attract almost all of it, and at a certain point nobody really cares about 150% vs 250% apr because those high APRs are universally red flags anyway. Although since this is a stablecoin LP we might see more liquidity just in general, but i still don’t think the logic of “30x emissions = 20x liquidity” is sound. With a pool of $27m, 2.1k SHD per day at $10 per SHD is 283% APR, The closest pool to our would be the frax pool on eth which is currently 2.58% APR. Put a yield premium on the fact that we’re a small chain and a small protocol, maybe our equilibrium point is 10% APR. Bastion protocol on Aurora is probably the best stablecoin yield right now in defi if you accept some smart contract risk, and the yield is 5% as of writing this post in a pool of $6m, so I think it’s fair to say that the yield of our pool would eventually hit equilibrium between 2.58% and 5%, but we can be optimistic and say 10%.
If we offer extremely high APRs on a stablecoin pool, people will deposit funds, claim the tokens, sell them, dumping the price, reducing the APR, which also causes liquidity to evaporate. Equilibrium is achieved once the sell pressure from emissions equalizes with buy pressure. Equilibrium tends to be very far away from the price when emissions started. There is a reason why every single defi farm token price chart looks like this:
And the reason is because of the sell pressure I just described.
So if we’re going to emit 2.1k SHD per day for a whole year, and the equilibrium APR for decentralized stablecoin pools in defi right now is somewhere in the 2-10% range, and I’m being optimistic and saying we will settle at 10%. Well, if the number of SHD being emitted isn’t changing, how is the APR going to go from 283% to 10%? Through the price of SHD decreasing. 10% is 3.5% of 283%, so the price of SHD would have to drop to at least 3.5% of what it is right now, assuming no liquidity withdraws from the pool. Current price is $9.42, which sets our price target at $0.33.
Now you have to factor in the fact that as APR drops, liquidity will leave the pool, which will increase APR (the same number of SHD emissions are going to be shared among fewer deposits). This means the real price target is even lower, and is probably in the range of ten cents to a penny, IMO, but i don’t have any quant data to back that up. But it doesn’t really matter, right? 33 cents is bad enough.
Sounds extreme? It’s not. Pick a farm token and go look at the price chart, they all go to zero, and this is why.
But I think the conclusion from all of that is simple - if the equilibrium of our APR is somewhere in the 5-10% range, just set the initial emissions to target 5-10% APR. Sure, it won’t attract as much liquidity as 284% APR initially, but it doesn’t really matter because when you frame it against the ratio of how much we spend vs how much liquidity we attract. A 10% APR on a stablecoin LP is phenomenal in these market conditions, and almost all of the liquidity that would be willing to invest in a stablecoin LP will take 10% APR because those are market-leading rates and they don’t sound ridiculous. I actually feel like high APRs will turn people away because they’re a red flag.
I also think we should play with that number. Maybe we start with a 10% target, and we gradually move down toward 5% and we see how it impacts liquidity. IF we can get $10m of liquidity with 5% vs $13m of liquidity with 10%, I’d say just set the rate at 5%. Stableswaps are capital efficient and we’re getting 30% more liquidity for 100% more money - not worth it.
If we slowly issue bonds from the treasury for USDC/SILK, we are not only building reserves for the grants, but simultaneously are also building out a liquidity floor which is directly correlated to the quality of the service that Silk can provide.
I like this idea, but I just want to emphasize that the focus needs to be selling bonds to raise capital for funding biz dev/marketing/protocol dev. The fact that we can take that money and LP with it is a side effect. I think it’s important to emphasize it in this way because we don’t want to be trapped thinking we can’t withdraw that money from the pool, because that money was never supposed to be earmarked for LPing anyway.
The more the treasury permanently owns this stablecoin liquidity, the more the service is guaranteed into perpetuity. We never have to worry about liquidity walking away if we are the market maker providing this service.
I think while this is technically true it’s not feasible. Stablecoin liquidity especially wants to be in the range of billions of dollars of TVL. The protocol can’t own even a small fraction of that, not at this stage of its life. We need to attract that liquidity through market participation (making the protocol worth using, making Silk an attractive currency to hold and use, etc), not pay for it out of our own pocket. That does mean we still have a problem to solve around “how do we keep liquidity from leaving”, but, well, that’s the business of an exchange, and I think the things that will make us grow are the things that will keep liquidity around. As market participation increases and utility increases, liquidity will increase.
Because I believe we should be obsessed with the quality of service that the protocol can guarantee users
I totally agree, I just think we should be careful where we draw the line of “what services we are guaranteeing”. We can guarantee an easy to use product, we can guarantee a product that integrates nicely into the rest of your financial activity by having it all in one place, we can guarantee a product that is safe and secure, we can guarantee a product that will see active development, growth, and innovation to improve over time, and I think all of those guarantees are appropriate for an engineering organization to honor. Guaranteeing that there will be tokens on the exchange for you to sell though? We can’t guarantee that. Under this thought process, consider the other pools. Okay, if we invest heavily in USDC/SILK (which is a good idea), what about the other pools? MANA liquidity is going to be low. sOSMO liquidity will be low, etc. What about all those other tokens? Should we guarantee them too? I think the answer is no, of course, but then that extends to USDC/SILK. Why should we guarantee that? One thing I like to think about is this: if USDC/SILK is an attractive LP option for us, it’s an attractive LP option for the market, too, so the market will fill that gap. Our job is to make USDC/SILK attractive (by making Silk attractive). Since there are other ways to make Silk attractive that aren’t LPing that have lower capital requirements and better ROICs, we should focus on spending money on those things and let the market fill in the liquidity gap, IMO.
Liquidity providers are a dependency we shouldn’t want to be reliant on over the long haul.
I’ve said it a few times but I do not agree with this. A financial system is dependent on its participants. That’s just a reality. LPers are market participants. There is no way around this. The reason is because 1. we can’t secure enough liquidity to eliminate dependency on LPers and 2. with no participants there is no market, so that liquidity would be meaningless anyway.
Finally, I fundamentally believe that if you zoom out over the course of 5-10 years the protocol will generate more yield and provide more of a service with protocol owned assets than what value that is extracted from liquidity providers in return for emissions.
I just think we need to be cognizant of the balance. Being long term focused is good, but completely ignoring the short term means you don’t survive until the long term, and I think if we are not ice-cold and incredibly stingy when it comes to how we spend our seed and early stage capital, we will not survive to live out the 5-10 year vision.
To be clear I agree with you that emissions are unsustainable - they’re a bootstrapping mechanism. We need ways to make liquidity providing attractive, and there are ways of doing that.
-
Building capital efficient pools. This actually isn’t hard, we already have two proven models (Uniswap v3 and curve v2) that work. We can literally just do no thinking at all and copy them, and we will improve usability. In this regard, just consider spending some of that emissiosn and bond money on funding the development of a concentrated liquidity AMM, and you will get a better return on invested capital and have something really attractive and marketable. Capital efficient pools provide better usability (much deeper depth for a given amount of liquidity) and better yield for LPers, so they accomplish all the same things as emissions and POL without the protocol taking on the risk of LPing (which is again, in my opinion, outside of its responsibilities.)
-
Leveraged LPing. Again, something that already exists, we just need to build it. If I remember correctly, you can take a 3x levereaged position in a volatile/stablecoin pair and you won’t be liquidated unless there are a 97% drawdown in the volatile token. Obviously 3x leverage means a higher beta position, so there is some additional risk, but not really risk of liquidiation. In the case of a stablecoin pair, that “safe” leverage number can be 10x or 20x, depending on LTV requirements. This lets smaller players create more liquidity, earn more yield, and also leveraged yield farming allows you to take different market positions (e.g. delta neutral positions on multiple leveraged yield farms on different pairs), and when there are more market conditions that are favorable to LPing because there are more ways to structure an LP position beyond the simple “no-drift” setup, there are more LP participants.
-
General Silk utility. Between #2 and this point, I think there is a compelling argument for all shadeswap pairs to be paired with Silk (pairing everything with Silk increases its utility and emphasizes its position as a currency, much like equity trades on brokerages are denominated in local fiat e.g. USD on robinhood). As Silk utility increases, volume of Silk activity increases. Silk minting and burning activity increases, which generates revenue in the form of fees, which can be earmarked for revenue sharing with LPers. If we spend our money on making Silk a better currency, we will naturally create more Silk liquidity because participating in the Silk ecosystem gets you more trading fees.
In conclusion I think the simplest way to describe my point is that we should view liquidity on the exchange as a side effect. It’s something that happens when we do our job well. When we make and do things that make Silk valuable, we will attract more liquidity.
- Privacy
- Multiple DeFi primitives interacting in permissioned ways
- ShadeDAO asset utilisation permissions
- Unified UI/UX of multiple DeFi primitives
- Sustainable revenue based on providing a service that users care about
- Community
- Integrations
- Liquidity
Despite my extremely long post expressing criticism, like I said in my first paragraph I agree almost entirely with everything you’ve laid out, and I even agree that liquidity is our moat. I just don’t agree with the specific mechanism of attracting that liquidity that you laid out. I don’t want to come off as too adversarial 
I also think the way this is framed as a holistic strategy is important, and I don’t think any of the alternative suggestions I’ve made in this post conflict with the holistic strategy.
In summary my suggestions are:
- Bootstrap liquidity with targeted emissions on only a USDC/SILK Stableswap pair (as you’ve said)
- Target 10% APR emissions with SHD
- Play with the emissions and see how low we can go without losing too much liquidity, maybe targetting a floor of 5%. IMO the target ratio should be 1.5:1 (e.g. if we spend 100% more to get 75% more liquidity, its worth it. anything less, it’s not worth it.)
- Pair all shadeswap LPs against Silk, making Silk feel more like a currency that you can bring to our exchange and purchase other things with.
That is the bootstrap phase. To transition beyond this phase and reduce reliance on emissions:
- Focus on creating capital efficiency for non-stable pairs by using curve V2 or concentrated liquidity pools
- Build leveraged yield farming - building the product, educating users, focusing on usability as the user stories for leveraged yield farming are complex.
- Explore other novel uses of Silk. I’ve brought up integrations with tokenized real estate to you and to others in the community before, and i won’t use this post to shill my idea further but just in general, “ways of doing things in web3 that you cant do anywhere else” bring a new demographic of user, improve the legitimacy of the protocol and of the space, and create real value. That value can be in the form of Silk utility (e.g. using Silk to purchase assets), it can be in novel revenue sources (e.g. low risk assets appropriate for a protocol treasury to speculate on, such as tokenized US bonds, if such a thing were to exist), or it could even just be in exposure to a new demographic.
And to emphasize what I think we should not do:
- Do not sell bonds to acquire funds with the sole purpose of LPing.
- Do not set extremely high emissions thinking that it will attract more liquidity.
- Do not rush into development of new things. We have a lot going on right now and not a lot has shipped, so I think even if we agree that “building other things” is the right move, let’s be deliberate about how we kickstart that development.