Stabilizing HBD

in #hbdlast year

stable coins balance.jpeg

I have literally been trying to write this post for a week.

Time to just get it done.


It's no secret that HBD is a pretty useless asset. It's ability to offset volatility is only slightly better than Hive itself. It has mechanics that get it back to $1, but the time it takes to get there is way too long. How can we reduce this time?


I've long spoken to how awesome the MakerDAO and stable-coin DAI are. Rather than being pegged to dollars in a bank DAI is pegged to crypto collateral. However, this latest Black Swan event has caused some serious problems, to the point of legal action even being taken against the creators of this technology. Even I recently had my CDP liquidated (although it was due to a MetaMask bug).

The main problem is liquidity.

Investors in MakerDAO were told loans that fell under 150% collateralization would be liquidated at a 13% penalty and the surplus would be given back to the owner of the loan. The problem with this is that when Black Swan events like this happen, everyone is scrambling to push transactions through on the Ethereum blockchain. Transaction fees skyrocket and the network becomes overloaded.

This causes the mechanics in place to not work as intended. Many of the loans that got liquidated ended up occurring at the last second when the collateral was much less than 150%. This caused many CDP holders to have their money liquidated and received very little Ethereum back afterwards.

Also, as I said before, this is a liquidity issue. If big CPDs get liquidated at a 13% penalty, what happens to that Ethereum (bad debt) that was bought out by some vulture capitalist? He probably sells it into a stable coin rather than risk Ethereum crashing further. This ironically causes Ethereum to crash further and liquidate more bad debt, causing a vicious cycle.

How can HBD do better?

I've been thinking about this a lot, especially in light of Justin Sun basically copying the idea and adding it to Tron recently (good idea; this is what open-source tech is all about). There are fundamental things about our blockchain that allow us to build a superior product.


The best advantage we have is 3 second blocks with zero transaction fees. During a Black Swan event we can liquidate bad debt 100 times faster than Ethereum could ever dream. This would also allow us to have a lesser penalty, not charge interest on loans, and allow HBD to be collateralized at a lower percent.


Maker is an ERC-20 token that governs this loan process. Steem does not need a governance token, because Steem itself is suitable for this purpose. We can use the same powers of voting that allow us to elect witnesses and approve SPS proposals to regulate Steem collateralized loans that create HBD. We can vote on everything to change the variables (minimum_collateral_percent, liquidation_penalty, interest_rate, liquidation_percent, etc).

Multiple techniques to stabilize.

HBD already has mechanics that help to stabilize the price to $1. This gives us way more options in terms of how we want collateralized loans to operate. With Dai, the MakerDAO must assume that bad debt is the worst thing in the world and must be eliminated at all costs. With Hive, this is not the case.

Example of bad debt being fine on Hive.

The best example of this is happening right now. HBD is already trading under $1, which means we already have bad debt, but we are doing just fine regardless.

In the case of MakerDAO, if Ethereum CDP loans fall under 100% collateral, Maker becomes the lender of last resort. Maker is created out of thin air, diluting investor value in order to keep Dai propped up at $1 and make sure there isn't bad debt floating around.

In the case of Hive...

We don't need to do any of that. We can just let the bad debt sit there and allow the HBD peg to fall. This wouldn't be any worse than what we are already doing. In addition, we see that these Black Swan events are usually just volatility. Ethereum crashed from $200 to $100 but is now already back up to $180. If MakerDAO had just allowed that bad debt to sit there everything would have probably worked out just fine. In my opinion, they focus much too hard on stabilizing DAI rather than keeping the network secure.


A whale locks up $2M worth of Hive and creates 1M HBD. The price of hive crashes by 50%. Now there is only $1M worth of collateral for the 1M HBD. This triggers a platform-wide auction to buy this bad debt. Anyone could bid to burn their HBD in order to buy some/all of this bad debt.

However, in this example, lets say that other CDP loans are getting liquidated and there is no demand to buy bad debt. Assume our market doesn't have the liquidity to handle bad debt being purchased and then dumped on the market. In this case, the value of Hive falls further, and now there is no longer any financial incentive to buy the bad debt. Now we have bad debt just sitting on the platform.

However, this situation in turn would cause HBD to break the peg and fall lower than $1. If this were to happen, and HBD fell far enough, it would now be worth it to purchase HBD off the market with the intention of buying bad debt with it. We can see that this system is potentially a lot more resistant to volatility than the MakerDAO could ever hope to achieve.


The MakerDAO must assume that everyone holding a debt position would NEVER pay off their loan if the Ethereum collateral was worth less than the loan being paid off. Is the same true for Hive? I think not.

We are a social network. We can implement reputation and even credit scores. If someone logs in and sees they have bad debt, there is a good chance they'll pay back the loan even though that's not the profitable thing to do. This is especially true if there are credit scores and accounts want to show a history of not having bad debt and supporting the network. There could be many advantages to having this credit history on the platform across multiple dapps. It could even be used as a metric for airdropped tokens and finding trustworthy community members.


The chance of this idea becoming a thing outright is pretty low, I will admit that. However, anyone could create a token with SMT (or even SteemEngine), lets call it FakeHBD. FakeHBD token would be a test to see if the system I've outlined actually works. If the FakeHBD token holds a better peg than actual HBD then the witnesses would be much more likely to adopt the CPD smart-contract into the core consensus layer. It would also be more likely because the majority of the code would already be written in that case and would just need to be ported over.

Variable to vote on.

There are several variables we could put on a sliding scale so we could change them dynamically (giving us a further advantage against Dai).

  • required_collateral_percent
    This is the amount of minimum Hive collateral required to create more HBD. MakerDAO uses 150%. If your collateral goes less than 150% it can be liquidated for the penalty as someone has bought your bad debt. However, I believe this should be a separate variable.

  • liquidation_percent
    MakerDAO doesn't have this (it is instead combined with the previous variable). This would be the percent of collateral a CDP would need to get to before someone can buy the bad debt. Again, MakerDAO uses 150%, but I believe this should be two variables to create a gap between HBD creation and liquidation of assets. Hive is such a fast network that we could go as low as 105%. This is especially true when we consider that we don't need a lender of last resort and can simply allow HBD to break the $1 peg by allowing bad debt to sit on the network (as we are already doing).

  • penalty_percent
    This is how much we punish users for having bad debt. MakerDAO uses 13%. I don't think HBD loans would need this variable. Instead, it could simply be combined with the liquidation_percent. For example, if the liquidation_percent was 105% collateral then that would be a built in penalty of 5%. An auction would be created that anyone could bid on. Say you have a outstanding loan for 100 HBD that is collateralized by $105 worth of Hive. Someone bids 100 HBD to buy your bad debt, but then another Hivian comes along and bids 105 HBD and wins the auction. 100 of that HBD would go to paying off the bad debt with the extra 5 HBD would get kicked back to the original owner of the CDP. In this case, the person who took out the loan got lucky and was penalized by 0% due to the mechanics of the auction and someone was nice enough to give them a break-even cut rather than trying to profit. Again, a move like this could increase the credit score reputation of the person who bought the bad debt and provided the break-even deal.

  • max_debt
    A lot of people use these systems to margin trade. They'd create HBD with a Hive CDP and then use the HBD to buy more Hive (and perhaps margin trade again). To prevent debt on the platform from getting out of control MakerDAO uses a cap. Last I checked it was 100M. However, Dai currently sits at a market cap of 88M, and I'm not sure how valuable having a cap is. The cap always just has to get raised when it gets hit and the value of the stable coin breaks above $1. At the same time, when there is too much margin trading going on usually the value of the stable asset will be under $1 because everyone is dumping it to buy more crypto, so maybe this variable does have value in that case.

  • interest_rate
    This is the one variable on MakerDAO that they actually do vote on, which is hilarious to me because it shouldn't even exist. No one should have to pay interest on a loan that they collateralize themselves for over 100%. The MakerDAO system is greedy, and they justify this interest rate by claiming it boosts the value of DAI up to $1 when it is trading for less than that. If they really wanted to boost the value of DAI higher than $1, then they should give it more demand than to simply margin trade with it. When it's actually being used as a stable-coin to hedge against a crash the value is $1 or higher. Again, this interest rate is predatory and does not need to exist on Hive. We have multiple mechanics that get HBD back to $1; we don't need interest rates. In fact, having no interest rate would attract world-wide attention and the value of that is priceless.


Central banks are already talking about how stable-coins are a threat and need to be regulated. However, they'll find that regulating the banks who collateralize stable coins is far easier than regulating a blockchain. HBD is in a prime position to become the world's leading stable coin if we stabilize it correctly.

Not only can CDP loans stabilize HBD by a gigantic margin, it also allows all users to margin trade and give themselves loans using their own collateral in a decentralized permissionless way. The amount of value Hive could generate by implementing something like this is massive.

On top of all these benefits, the value of everyone's upvotes would increase. How many users do you know that would powerdown staked coins in order to allocate them to a CDP contract that lets them create HBD "out of thin air"? The more smart-contracts we have that lock up tokens, the more value we generate. This applies to both USD valuation of coins and vesting itself.

Hopefully we can get a project like this up and running within the next couple years. I'm tried of having a stable-coin that is the laughing stock of all stable-coins.


This is a really good idea! This would require smart contracts though, which aren't really a part of Hive and I see no push to put them on Hive. The no interest rate idea is kinda dangerous though, it would only work with more utility for the coin which is hard this early in the tech, maybe with more tokenization and STOs paying in stablecoins the no interest-rate would work.
The reputation system could maybe decrease incoming upvotes by a %, the longer the debt isn't paid the more the % increases!

upvoting is a smart contract. so is staking and voting for witnesses and curation etc etc. code is code.

Yeah, but those were implemented since the start of the blockchain, putting a new mechanism in would add complexity to an already complex blockchain, that would probably not be very easy.

Curation is very complex compared to this. With curation, you need to track the order and weight of every vote and crank it through an algorithm for every single post.

The smart contract involved with this would just basically be: lock coins... create HBD... burn HBD; unlock coins. Way simpler and way less taxing on the nodes.

there is no practical reason to have interest on loans. why would it be dangerous? loans have more value to the platform without interest. everyone would want to use our platform.

Without interest how would you "force" the price of the token to go above $1? Maybe on Hive you can by adjusting the issuance rate and the conversion to HIVE, but on a normal blockchain without the sort of mechanisms that Hive has, not sure if it's possible.

The value of a stable coin doesn't need to be forced anywhere except to a single stable value. That value could be 50 cents or 98 cents, less volatility is better. Why would anyone ever want to "force" the value of their stable coin to a non-stable value?

The interest rate has little affect on the stability of the coin. Real demand comes from actually using the stable-coin for its intended purpose: avoiding volatility.

However, I can see what you're saying here... what's to stop people from speculating on the stable coin and pushing the value lower to margin trade instead of using it as the stable asset it was intended to be? In this case the network raises the required_collateral_percent instead of a needless interest rate. By increasing the collateral needed to create HBD, this will limit the supply and further push up the value of Hive by locking more coins.

This is exactly why it was foolish for the MakerDAO to combine the variables required_collateral_percent and liquidation_percent. It's impossible for them at the moment to raise this bar without also raising the bar of liquidations. They are stuck at 150%. We could raise that level to 200% or even 300%. And we could do it on the fly and not punish users who took advantage of a lower level because the liquidation_percent remains unchanged.

As I understand - one of the major features missing to allow HBD to be a "true" stable coin is an option to create them, right?

A whale locks up $2M worth of Hive and creates 1M HBD

Yes, we got completely fucked over when SBD spiked to $13 during the 2017 bull market, the effects of which are still being felt today. We couldn't keep up with demand.

Can this method of creation can be implemented into current HBDs? If the whale in the given example would like to receive back his $2M Hive, simple conversion would not work because $1M HBD -> $1M Hive. I wonder if we need completely new type of HBD or the generation process can be implemented with existing ones :)

@tipu curate

Upvoted 👌 (Mana: 0/4)

I remember you mentioning this or similar to save SBD, glad to see someone is still on it... if we can fix it now early the newcomers may never suffer from the bad history

Please explain this:

HBD already has mechanics that help to stabilize the price to $1. This gives us way more options in terms of how we want collateralized loans to operate. With Dai, the MakerDAO must assume that bad debt is the worst thing in the world and must be eliminated at all costs. With Hive, this is not the case.

I still haven't figured out how SBD and now HBD work.

when HBD is higher than $1 Hive prints half of all author rewards as HBD. When HBD is lower than a dollar we print less and can burn the HBD for hive.

Yey, yes, yes.

How is that SBD floats at $1 for weeks but HBD couldnt? They share same history price.

HBD only falls below $1 when the network has to haircut it.
The haircut occurs when the market cap of HBD is more than 10% the market cap of Hive.

This is theory I know very well.
Do You mean HIVE price is too low below haircut? Everything below 16 cents makes near impossible to HBD recover because witnesses signaling HBD/HIVE ratio too high. Conversion system is not efficient for prices we have now.

Yes, that is very accurate.
Sounds like you know what's up.

When Hive crashes too quickly HBD gets a haircut.

HBD is in a prime position to become the world's leading stable coin if we stabilize it correctly.

promising theory

Excellent outlook.


 last year Reveal Comment