There have been a lot of conversations and questions about SBD:
In this post I want to try and explain the economics behind how SBD works, and try to clarify some of the things that are being discussed. I will also explain the two pull requests that I submitted for hardfork 20.
SBD (often referred to as "Steem Blockchain Dollars") is one of the two tokens that are currently supported by the Steem blockchain (STEEM and SBD). Similar to STEEM, SBD shares many of the benefits that the Steem blockchain has to offer such as three-second confirmation times, and zero-fees on transfers.
SBD (also like STEEM) gets it's value based on the market, i.e. what traders are willing to pay for it. If there is more supply than demand, the price goes down; if there is more demand than supply, the price goes up.
The original design behind SBD was that the price should be kept as close to $1 USD as possible. In order to achieve that, mechanisms were built into the Steem blockchain (including several tools that are made available to the witnesses), to help achieve a stable $1 USD price.
The primary mechanism that is built into the Steem blockchain in order to help keep SBD close to $1 USD is that the blockchain supports a "conversion" mechanism, which allows anyone to convert 1 SBD token into "approximately 1 USD worth of STEEM".
The way the conversion process works is that when someone initiates a conversion of SBD, it removes the SBD from the user's active/available balance, and moves it into their conversion balance. It sits in their conversion balance for 3.5 days. At the end of the 3.5 days, it takes the median price feed value from the 3.5 days (the price of STEEM supplied to the blockchain by the witnesses), and uses that price to create $1 worth of STEEM in place of the SBD token.
(Once a conversion is started, it cannot be canceled.)
Because traders can buy SBD from the market and convert it into ~$1 USD worth of STEEM, this helps to ensure that the price of SBD does not dip significantly below $1. (If it did, a trader would see an opportunity to buy it, convert it, and make a profit.)
If one SBD token is going to be worth "~$1 USD worth of STEEM", then one SBD token is going to be worth a different amount of STEEM, depending on the USD price of STEEM.
This can provide an economic benefit to STEEM holders if SBD is printed at a lower rate than it is redeemed via conversion, since the blockchain will end up producing less STEEM tokens in order to pay out the same amount of rewards.
For example, if STEEM is currently valued at $1.00, and 5,000 SBD tokens are printed today to pay users $5,000 in rewards, that is 5,000 STEEM tokens worth of rewards. If however nobody converts those SBD tokens into STEEM until much later, say when the price of STEEM is up to $10, then the blockchain would have only had to generate 500 new STEEM tokens in order to pay out those rewards.
The same mechanism can work in reverse however. If the blockchain prints a whole bunch of SBD tokens while the STEEM price is at $10, and then nobody converts them until the price goes back to $1, then it would have to print 10x as many STEEM tokens in order to pay out the same amount of rewards.
(In other words, it is a double-edged sword.)
One can make the argument though that since the general trend of cryptocurrency markets (over the long-term) is to go up, that in the long-term, this double-edged sword should work more in favor of STEEM than against it.
If you go to steemdb.com, you will see three numbers related to the supply of STEEM/SBD: current_supply, current_sbd_supply, and virtual_supply.
current_supply is the number of STEEM tokens (both liquid and powered up) that are in existence.current_sbd_supply is the number of SBD tokens that are in existence.virtual_supply represents what the current_supply of STEEM tokens would be if the entire supply of SBD tokens were instantly converted into STEEM at the current market price and added to the current_supply.There is a number that is often talked about, and that is the SBD debt ratio. The debt ratio is the number of SBD tokens in existence (current_sbd_supply) divided by the marketcap of STEEM.
The marketcap of STEEM is calculated by taking the virtual supply of STEEM (virtual_supply) and multiplying it by the current STEEM price.
For example, if there were currently 5M SBD tokens in existence, with a virtual supply of 100M STEEM tokens, and a current STEEM price of $1 USD, then the debt ratio would be 5M/(100M*1) = 5%.
A lot of times people ask why am I getting STEEM rewards instead of SBD? The reason is because of the SBD print rate rules. (The print rate rules were added as part of hardfork 14.)
The way "default" (50/50) payments work, users receive half of their rewards as Steem Power, and the other half as "liquid". Typically the liquid portion is paid out in SBD, but not always. It can also be paid out in STEEM.
The blockchain has rules for when to pay the liquid portion in SBD vs. STEEM, and they are based on the SBD debt ratio (defined above). Based on the debt ratio, the blockchain will pay out the liquid portion of rewards as:
Note: These numbers are changing as part of the planned HF20 changes, which will be explained in more detail below.
As part of hardfork 14, a rule was also added to the Steem blockchain to help ensure that a significant drop in STEEM price would not allow SBD to have a catastrophic effect on STEEM.
If for example we were at a 10% debt ratio (with no debt limit in place) and the STEEM price dropped from $10 to $1, then the virtual supply of STEEM would double - that is 100%/year inflation instead of the intended 8-9%, which would be really bad for STEEM holders. If the price continued to spiral from $1 down to $0.10, the yearly inflation for STEEM would be 1,000%. This possibility of extreme inflation from drops in price poses a serious risk to STEEM holders, which is why a protection mechanism was put in place.
The SBD debt limit that is set at 10% (often called the 10% "haircut" rule) ensures that no matter what happens to the STEEM price, the SBD debt ratio will never be allowed to exceed 10% of the STEEM marketcap.
It does this by reducing the liability of SBD conversions if/when the debt ratio exceeds 10%. If a user converts SBD to STEEM while the debt ratio is over 10%, then they will get back less than their ~$1 USD worth of STEEM. This transition happens gradually, and has a linear relationship to the distance the debt ratio is over 10%.
For example, if the debt ratio is at 10.01% the converter will get very close to ~$1 USD worth of STEEM. At 10.02%, a little less, and so on. By the time it reached a 20% debt limit, SBD conversions would result in ~$0.50 USD worth of STEEM.
Witnesses were given tools to help adjust supply and demand for SBD if the price of SBD is consistently above or below $1.00. Witnesses are instructed not to use the tools to respond to short-term market conditions, but rather focus on long-term market trends.
The views of when these tools are appropriate to use, and how to use them under different scenarios is a heavily debated topic among the witnesses.
One of the tools that witnesses have the option to use is a price feed bias. Essentially they can intentionally tell the blockchain that the price of STEEM is higher or lower than it actually is.
It should be noted that if already very few users are converting SBD into STEEM due to SBD trading above $1 USD, that decreasing the conversion output will have little effect on the supply and demand.
It should also be noted that artificially increasing the SBD production can result in higher yearly inflation numbers than estimated in the whitepaper, which could be viewed negatively by stakeholders and potential investors.
Another tool that witnesses have is "APR". APR allows witnesses to pay SBD holders interest for holding SBD, which is another way to stimulate demand.
One concern that is often brought up regarding APR however is that since a majority of SBD is stored on exchanges, it is primarily a benefit to the exchanges holding the SBD, which weakens much of the intended impact.
With the current tools that are available to the witnesses, there is limited ability for witnesses to put downward pressure on the price of SBD, if/when the market is valuing it significantly above $1 USD.
One of the ideas that has been proposed to help address this is to allow "reverse conversions", i.e. allowing someone to convert ~$1 USD worth of STEEM into 1 SBD token. For those interested, there is discussion on the idea here.
One of the changes that I submitted for hardfork 20 was to update the print rate rules to print SBD in place of STEEM for longer. Under the new rules, the print rates would change to:
TLDR: The print rate rules are being shifted from 2%-5% to 9%-10%.
The pull request for the change can be found here:
https://github.com/steemit/steem/pull/2503
There is some concern about the economic risk of increasing the print rates to allow more SBD tokens to be printed. This is a valid concern. The truth is there is some risk to the change.
If we start allowing SBD to be printed all the way up to 10%, then it will allow us to get right up to the 10% debt limit. Any reduction in STEEM price after we reach that point would result in us passing the 10% debt limit.
While there is some risk here, I believe the risk to be relatively small.
In other words, the risk is on the SBD holders, and it is their choice how much risk they want to take.
What is most beneficial about this print rate change is that it shifts the leverage equation significantly more in STEEM holders' favor:
The other change that I submitted as part of hardfork 20 was to update how beneficiaries are paid. Beneficiaries are a feature that was added in hardfork 18, which allows users to share their author rewards with other users, and for third-party user interfaces to take a small portion of the author rewards form every post/comment that is posted via their site.
Collaboration between users and application developers being able to make money by developing apps for Steem are two things that I think are mission critical for the success of Steem in the long-run. This is why I feel that improving beneficiary payouts will be a huge win for Steem.
Currently the way that beneficiary payouts work is that beneficiaries always receive their portion of the payout as 100% SP. This means that if the author selects 50/50% payout and receives half of their rewards as "liquid payout" (typically SBD), then even though the author is receiving SBD - the beneficiaries are not.
The pull request I submitted updates the logic so that beneficiaries are paid out using the same setting as the author. This means that if authors receive 50% SP and 50% SBD, then the beneficiaries will also receive 50% SP and 50% SBD.
The pull request can be found here:
https://github.com/steemit/steem/pull/2555
One of the biggest benefits of SBD was explained to me by 
Examples of things that decrease the liquid supply of STEEM include:
What is interesting is that when liquid rewards are paid in SBD (instead of STEEM), those SBD tokens essentially have the same economic benefit as a user buying STEEM and not selling it, i.e. HODLing. For as long as those SBD tokens are in existence (not converted to STEEM) that is less liquid STEEM being traded/sold. Even if a user receives a SBD token and turns around and sells it right away, that is not putting any downward pressure on STEEM.
In other words, every single SBD token that is in existence right now is reducing the supply of liquid STEEM tokens on the market, which is helping to increase the price of STEEM.