So I still want to create my own cryptocurrency...
I think about it every so often. Been thinking about it a bit more lately. The concepts I came up with were pretty sound, but actually making it decentralized is quite a tall order. Creating a database is one thing, but how do you actually incentivize other people to spend money to boot up a node and maintain a shared database? No easy feat, to say the least.
Not only that, but even if the backend protocol is decentralized... if it only has one frontend then the centralized bottleneck is still completely intact. Doesn't matter what the database actually says if I simply show users a different set of information. If regulators were breathing down my neck and I just shut down my node, would someone else really keep it going? Decentralization is pretty hard when you actually get into the technicalities.
The easiest way to play it is to just keep the thing centralized until you have a big enough userbase and the proper incentives to decentralize the protocol later. We've all seen what a slippery slope this is, as founders will often never actually decentralize their service because they are making money pretending to be decentralized while actually in control of everything. Pretty sure I wouldn't have this problem, but ignoring the issue entirely is also foolish.
Hm, well how does Hive do it?
I'm obviously a fan of DPOS and how Hive does business, so I could just copy that, yeah? So then I'd have to allocate 5%-10% of network inflation to a certain number of consensus witnesses. Would I need, or even want, 20 consensus witnesses? Probably not... that would be so much for such a small and new thing. Even 5 would probably be fine to start. A 3/5 or 4/5 consensus vote would be adequate.
At the same time I want most variables within my ecosystem to be dynamic. I want the network to be able to increase or decrease the number of block producers while also being able to modify the amount of inflation allocated to them. But this is perhaps getting way ahead of myself.
Because what even is a block producer within this context?
This is why I originally thought I could get away with not having any block producers at all: because if the thing is built on Hive then it's not even creating its own blocks, but rather each block on the second layer network is just a derivative of some other block on Hive that filters out all information except for custom JSONs tagged with 'magitek'.
If blocks are already basically confirmed by Hive witnesses... do they really need to be confirmed again by second-layer block producers? Maybe... or maybe there is some shortcut that I haven't fully considered yet.
There is also the issue of forks...
Forking the code by anyone other than me would be challenging, as the only identifier is a tag in custom JSON that says 'magitek'. Anyone else that tried to use that tag would just be labeled a hacker and a fraud by yours truly. Anyone who changed the tag could easily create a new thing and be ignored by the original network. So that's a weird aspect in all of this as well... the way Hive handles forks and fork ID is fairly complex and I don't fully understand it.
And I haven't even gotten to any of the tokenomics stuff yet.
Now that we've experienced a full DEFI bull market accompanied by the full DEFI bear market, it's even easier to see what worked and what didn't. If we listen to Bitcoin maximalists, everything that offers yield is an automatic failure no matter how that yield is structured. "You are the yield," was a common mantra when 99% of DEFI products were falling to pieces and crashing to zero. This continued on as centralized exchanges started collapsing in droves as well. Of course this is the extremely short-sighted view of the maximalists, who aren't looking beyond the surface because they live in an echo-chamber where nothing but Bitcoin can succeed long-term.
We need self-contained ecosystems.
Most DEFI products baked their own destruction into the cake by allocating yield and inflation into areas of little to no value. What good does allocating yield to a BTC/ETH LP have, other than giving free liquidity to Bitcoin and Ethereum? Why would anyone do this? And yet, they all did. It is unclear if the reason behind these moves was incompetence, because they wanted to farm/dump their own token risk-free, or because this strategy actually worked for the first projects when everything was new and these kinds of pools hooked in new users who otherwise wouldn't participate.
Rule #1 of DEFI tokenomics: Conservative yields.
Every single yield allocation in DEFI needs to be deeply scrutinized. The question we should be asking is a simple one, but the answer is not always simple: Does this yield allocation bring more value into the network than it extracts through token dilution? For certain projects it's very easy to see that they messed up big time on this front and allocated inflation into places that were permanently hemorrhaging money and value away from the ecosystem. The only way to scale is to capture more value than you print away.
Rule #2 of DEFI tokenomics: Yield is the killer dapp.
Intrinsically linked to rule #1, if the DEFI token can allocate inflation only to places that generate more value then the cost of the yield, this token is guaranteed to experience exponential growth on a mathematical level. This doesn't even include all the free value that can be generated with code that doesn't require an inflation allocation. For example if I create liquidity pools, that has value but they need yield to be incentivized. However, if I created an NFT game on top of the network the NFTs don't need inflation allocations and simply are icing on the cake. The NFTs have their own value because they have the utility of being used in the game, and that value would be linked back to the governance token for even more free synergy.
Rule #3 of DEFI tokenomics: Elasticity is crucial.
Just like Rule #2 is linked to Rule #1, Rule #3 is linked to Rule #2. Financial elasticity is impossible to achieve without dynamic changes to the inflation/debt rates across the board. The goal here is to manipulate the economic model in such a way that it benefits everyone. No one can control demand of currency or debt, as that is determined collectively by everyone within the economy simultaneously. However, if we can lower money supply when demand drops and increase money supply when demand increases, this is highly beneficial to the next rule and ultimate goal of DEFI.
Rule #4 of DEFI tokenomics: Stability is the ultimate goal.
All money must have a certain level of MoE (medium of exchange), SoV (store of value), and UoA (unit of account). Intrinsically, crypto is an incredible medium of exchange that is worldwide, borderless, basically instant, and totally censorship resistant (governments can't stop the transfer of wealth across their own borders). This will only become more impressive in the future as we creep toward mainstream adoption, so MoE is something that crypto will basically never have to worry about and will essentially be grandfathered in for life as a superior option.
Store of Value is a bit more tricky, and only some networks have proven that they are able to achieve it to various degrees. However, the volatility of crypto is such that every bear market all skeptics end up making the claim that crypto is dead and has zero store of value. This is only true on shortsighted timelines when it comes to networks like Bitcoin and Ethereum. Many other networks have yet to prove themselves as we zoom out, but Hive is beginning to raise the floor from previous levels, so that's exciting.
Unit of Account is the one thing that no crypto has yet except for stablecoins, which basically don't count because they aren't creating their own stability, and are instead piggyback on USD: which is exactly the opposite of what we are trying to do long-term. By way of elasticity with a focus on stability, DEFI can create a product that has all three functions (MoE, SoV, UoA), which economists have previously thought to be impossible. I can see that it's not only possible, but will eventually be the guaranteed outcome of DEFI once the proper templates are in place and are shown to have success so they can be copied by others.
So rather than trying to get the price of the token to moon, the real goal of DEFI is for the price of the token to remain perfectly flat, while only the yields fluctuate depending on current demand. Ironically, this means lowering yield and inflation during the good times and increasing yield and inflation during the bad times to incentivize users to prop up the current price.
The reason why this is so confusing and outright absurd-sounding to most people is that theoretically increasing yield should increase the supply, but it doesn't. It has the opposite effect, especially on the short term. If I double APR and everyone sees that yield is x2 what it was the day before, no extra tokens have been created (yet). But the demand to buy tokens off the market and lock them up skyrockets immediately, taking them out of circulation at the given price point, which was the goal the entire time.
This is exactly what happens when the FED increases interest rates. An interest rate is denoted in APR or APY. We would do well to remember what the 'A' stands for: it stands for "ANNUAL"; meaning yearly. Thus when an APR gets modified the actual affect of that change does not fully happen until an entire year goes by and the ANNUAL rate has had a full year to do its business. That's just how percentages and interest rates work on an intrinsic level.
On the flip side, increasing supply (and thus lowering price) is as easy as reducing the incentive to hold the asset, which means doing the opposite and lowering inflation, yield, and APR on the LPs. Again, this not how people think it works, but it is exactly how it works. If we take away the incentive to do something, less people are going to do it. It's quite simple to prove even though it sounds wrong on a common sense level.
The problem with all this economic theory is that we have to not only guess right, but also guess right potentially an entire year in advance. If we guess that inflation needs to be increased to suck liquidity off of the market and we're wrong, a year goes by without the intended effect and now tokens have officially flooded the market. If the expected demand does not appear after a year's time, the economy will be in trouble. Of course this assumes that the developers didn't create anything of value within the last year, which... would be embarrassing in crypto as developments happen so quickly within this environment.
But the ultimate problem is lowering token price on purpose during the bull market and having the common sense to create the required buffer to survive the inevitable bear market. Everyone gets greedy and no one wants to see the price come back down to Earth. It would really take someone with diamond hands to create that synthetic bear market on purpose in order to lower the price back to the implied targets. Perhaps the future of these tokenomics are algorithms and AI that will do it automatically rather than depend on ridiculous degens to follow through and do the thing.
lol I haven't even gotten to negative interest rates...
This is actually the thing that I've been thinking about this entire time that sparked me to write this in the first place. My my how long it takes me to get where I need to go.
So originally I envisioned Magitek with 4 main tokens.
The theme is Wizard RPG.
- Fire (A derivative of Hive)
- Lightning (A derivative of HBD)
- Ice (A derivative the 2nd-layer network governs)
- Mana (The governance token)
And so I had to ask myself... have I already theory-crafted a bloated system? If I employ my own rules of DEFI onto this system... does it work? Are MANA/FIRE and MANA/LIT LPs worthy of having inflation allocated to them? Or should I just scrap the idea and just try to create liquidity pools directly connected to Hive and HBD rather than a derivative?
After careful consideration I believe that the derivatives do in fact have value. One of the big reasons for this is that, even though it is risky business, I'd very much like to create a collateralized lending protocol that allows people to mint the derivatives on demand. Being able to mint FIRE with LIT/ICE/MANA collateral would allow users to essentially short Hive during irrational bull markets, while allowing users to mint LIT/ICE on demand would allow them to go long (by going short on stable coins) during the bear market. Just the idea of being able to give yourself a loan is pretty amazing and I'd really like to see Hive be able to capitalize on that in the way that Ethereum has. Protocols like MAKERDAO have proven invaluable and there are dozens of improvements that could be made.
Positive/Neutral/Negative interest rates
The concept of being able to manipulate interest rates across a full spectrum intrigues me. Neutral and negative interest rates do not make sense within the context of a debt-based ecosystem, but they can make sense within a collateral based ecosystem, which is exactly how crypto operates outside of the fractional reserve system. Not only is every loan fully collateralized in crypto, it is overcollateralized to shield against volatility and possible insolvency.
Within this context it makes very little sense that you would be charged a positive interest rate on tokens you created using collateral that was worth x2 more than the mint. And yet that's exactly how MakerDAO does it. Give yourself a loan with your own money and they want to charge you for it. This shows a complete lack of competition in the space as there's absolutely no reason it needs to work like this. Well perhaps for MakerDAO it does make sense because their MAKER token is deflationary and they have zero ability to incentivize anything with inflation due to their own flawed archaic model.
There are many situations where an asset would be overblown in value and we should want to increase supply of that token to match the demand of the FOMO market. Negative interest rates could be a great way to accomplish this without affecting other parts of the economy.
For example, say we were trying to stabilize the governance token. The price has spiked out of control due to high demand and we need to increase supply. There are many ways to accomplish this:
- Reduce yield to single-staking pool (Cache). This would push MANA out of the Cache and into the LPs, creating more liquidity and lowering the price of MANA, making it more available to the masses. Very good first option if there is still yield left to cut.
- Reduce yield of LPs This would lower liquidity across the board and potentially incentivize users to exit the platform entirely. Could backfire if users exit the LPs and stack the Cache during peak FOMO. Not a great option but could work in a pinch.
- Reduce global yield Less governance tokens minted per day would lower the incentive to hold the token across the board. Good option in the middle of a rampaging bull market so there is buffer to increase later during the bear.
- Create a negative interest rate on collateralized loans. Value in the Cache can be used to mint any of the four main tokens. By attributing a negative interest rate to MANA debt, users could earn yield while they short MANA during bull market season. They put their Cache at risk while they do this (if MANA keeps spiking they could get liquidated) but with a high enough collateral ratio and aggressive enough pressure on downward price action the risk should be relatively low and achieve the desired result.
Why are you so focused on LOWERING the price of crypto?
The goal here is balance and elasticity. If we lower the price during the bull market, we can increase the price during the bear market and close the gap, creating a system that is much more stable and trustworthy. It's like surfing a wave: if you're at the very top of the wave or the very bottom, you're doing it wrong. Balance is key, and we can reverse any of all of these tactics to bring the price back up when it is needed, without destabilizing the network through hyperinflation. A stable network is a healthy one.
Risks of collateralized lending.
There's a reason why not a lot of platforms do this... and if you want any examples just ask FTX or Celsius why it's not always the best idea. The ultimate downside to collateralized lending, especially in an environment of extreme volatility: are the liquidations. If the collateral becomes worth less than the loans, the entire system breaks and there is no way to maintain pegs or value within the system as it becomes flooded with bad debt.
Black Hat
Hackers can also move in and attempt to manipulate the value of collateral in order make it artificially appear to be worth less than it is. Once this goal is achieved, they can essentially liquidate user collateral and scoop it into their own pocket after paying back the "bad debt" for cheaper than the collateral is worth.
Hive avoids this situation with HBD conversions to Hive by using a 3.5 day average on price. This is a great way to avoid the attack vector because manipulating the price for three days straight would be way too expensive and never profitable for this kind of attack.
The problem with applying this solution to collateralized lending is that it opens up the network to legitimate Black Swan systemic failure. We can prevent the network's oracle price feed from being manipulated on a short term basis to avoid wrongful liquidations, but in turn, given a legitimate Black Swan price crash, collateral can become worth less than the loans that were taken quite quickly, leaving the system underwater with no incentive to pay off bad debt. Certainly there is a balance here that would lead to the safest way to do business. What that equation is unfortunately is unknown to us until we test it in the field or do some very accurate theory-crafting.
Conclusion
Alright well this thing is over 3000 words so I need to cut myself off even though I still have a lot more to say. Long story short I've been thinking more about my own token within the context of collateralized lending and simple NFT games with actual utility. This bear market has been a bit depressing and I find it impossibly difficult to jump back into the dev game, but hopefully one of these days I'll find the energy to dive back into it head first. Until then it's all just theory-crafting and conjecture.
Return from Tokenomics Ramblings: DEFI & Magitek to edicted's Web3 Blog
