Imagine you want to buy bitcoin (BTC) in exchange for US dollars (USD). In Bisq terminology, you are a buyer of BTC looking for a seller of BTC who will accept USD as payment. To complete such a trade using Bisq, you would follow a series of steps similar to these:

These steps can vary in a number of ways depending on whether you wish to buy or sell bitcoin, whether you are the maker or the taker of an offer, which payment methods you have access to, and so on. But in any case, the steps above are rather different than those one would follow to complete a similar trade on a centralized exchange.

Beyond the obvious difference that Bisq is a desktop application and not a browser-based web application, the first difference experienced traders will notice is that there is no automatic order matching on the Bisq exchange. Rather, Bisq users manually search for and select specific offers they wish to take. This approach enables truly peer-to-peer trade settlement, and ensures that users are in control of which counterparties they trade with.

Bisq is also unique among decentralized bitcoin exchanges in the way it coordinates out-of-band fiat payments. Bisq does not directly integrate with banks or other national currency payment systems in any way. Rather, Bisq’s trading protocol orchestrates the process of buyer and seller working together to settle fiat payments outside of the Bisq application, as demonstrated in steps 5–7 of the trading example above.

These and other differences result in a key tradeoff for Bisq users—​one in which trade settlement takes longer, but trading itself is far more secure, private and censorship-resistant.

After two years of development and testing, Bisq went into production on April 19th, 2016—​18 months ago at time of writing in October 2017. Since then, the network has processed 5,200 trades worth a total of $4.1MM USD without downtime or major incident.

Bisq is still small, but has been growing steadily. The USD volume of bitcoin exchanged through the network has doubled roughly every 3½ months since the project went live, from $36K in April 2016 to $438K in September 2017. This growth has been organic, with minimal marketing.

Bisq is designed to be funded directly by its users through trading fees. Trading fees are paid by both buyer and seller on every trade, and are received by each trade’s arbitrator in compensation for the service they provide. As of October 2017, these trading fees total around one bitcoin per month, distributed to two arbitrators who are also the project’s founders and principal developers. These funds are insufficient to cover expenses and as a result the project remains funded in part by founder savings.

Fortunately, as mentioned above, Bisq trading volumes are growing, and total monthly trading fees are growing along with them. As such, it is reasonable to expect that trading fees will soon be sufficient to cover expenses, and even to compensate additional, non-arbitrator contributors. But as these additional funds come in, a new problem arises with them. As mentioned above, Bisq’s trading fees currently compensate arbitrators in a direct, automatic, and decentralized way, but the current approach is limited in that it only compensates arbitrators. What’s needed is a mechanism that can compensate all kinds of contributors in an equally decentralized fashion.

Bisq’s technology is fully decentralized, but its governance is not. Today, the Bisq project consists of a small team of active contributors, where most maintenance, operation and administration duties are carried out by project founders, and where those founders make virtually all major decisions. While this approach has worked well enough to bring the project to its current state, it is no longer sustainable if Bisq is to continue to grow. First, because it does not scale; project founders have become bottlenecks incapable of addressing all user needs. Second, because it introduces censorship risk to have responsibility centralized in a small group of people. What’s needed is a way to decentralize responsibility and high-trust duties away from project founders and into the hands of other competent and reputable contributors.

As mentioned above, Bisq’s mission is to provide a secure, private and censorship-resistant way to exchange bitcoin for national currencies and other cryptocurrencies over the internet. Today, Bisq is delivering on the first two aspects of this mission—​indeed, we believe Bisq to be among the most secure and private exchange options currently available. Where Bisq falls short on this mission is in the third aspect of censorship resistance. Bisq’s P2P architecture, use of Tor as a transport layer, and other built-in protections give the project a significant degree of technical censorship resistance, but with regard to funding and governance, the Bisq project remains vulnerable so long as these two critical factors remain centralized.

To fully realize its mission, what Bisq needs now are four things:

In the next section, we’ll see how the Bisq DAO and BSQ token have been designed to address these needs.

We introduce a token, BSQ, designed to facilitate a transfer of value from the traders using Bisq to the contributors maintaining it.

BSQ is Bisq’s own custom implementation of the colored coin concept.^[2] Using 2.5 of the 25 bitcoin (BTC) that have been donated to the project since its inception in March 2014, we create 2.5 million BSQ tokens, such that each BSQ token is represented on the Bitcoin blockchain by 0.00000100 BTC, or 100 satoshis.

We distribute these 2.5 million BSQ to 144 opted-in past contributors to the Bisq project, as a way of rewarding these individuals for their efforts over the years. Stake is distributed according to the relative value each contributor has added to the project over time. As will be described further below, BSQ tokens are used to vote on and make decisions about the Bisq DAO itself, and by initially distributing BSQ to past contributors, we intentionally establish the Bisq DAO as a meritocracy in which those who have contributed the most value to the project in the past are those who have the most say over its future.

Today, without BSQ, traders use Bisq and pay Bisq trading fees in BTC. At the same time, contributors work to improve Bisq, and in turn make Bisq more useful to traders. This ultimately causes more traders to use Bisq more often, and a cycle of growth perpetuates. There is a limit, however, on how much growth can occur, because while trading fees paid in BTC automatically compensate arbitrators, other contributors are currently not compensated at all.

Bisq’s current funding model has worked well enough so far mainly because the Bisq team has been so small. Bisq’s founders have played most key roles in the system, including that of arbitrator, such that compensating arbitrators with trading fees has, in practice, meant compensating the project’s primary contributors as well. The problem is that this approach breaks down quickly as more contributors get involved, and as explained above, it is critical to the success of the project to grow the number of contributors and to distribute responsibilities amongst them.

What’s needed is a decentralized way to transfer value from traders using Bisq to contributors maintaining it, and there is no practical way to achieve this with Bitcoin alone. Accumulating and distributing trading fees using multisignature addresses and transactions could in theory provide part of the solution, but in practice these tools are still too primitive to achieve the fully decentralized and meritocratic approach required to successfully operate the Bisq DAO.

Furthermore, for a system of compensation to be fully decentralized and meritocratic, an equally decentralized and meritocratic system of governance is required in order to decide which contributions should be compensated and which should not; and this, too, is not practical to implement with Bitcoin alone.

It is for these reasons that BSQ is needed. As we’ll see in the sections that follow, the BSQ token has been designed to overcome the challenges described above and to provide a fully decentralized and meritocratic funding and governance model for Bisq.

The BSQ token has five uses, also known as utilities or functions, within the Bisq DAO:

Together, these interlocking functions of the BSQ token are designed to create, regulate, and perpetuate an internal economy for the Bisq network. We call this economy the Bisq DAO.

When traders spend BSQ on trading fees, they are in fact destroying those BSQ. The spent tokens are not paid to or otherwise received by any individual contributor or group of contributors; rather they are burned, or made to be unspendable by being decolored. In this way, spending BSQ on trading fees decreases the total supply of BSQ.

Likewise, when contributors submit compensation requests for BSQ, they are in fact requesting the right to issue, or create those BSQ. Each compensation request revolves around a Bitcoin transaction in the amount of satoshis required to represent the requested amount of BSQ, and when that compensation request is approved by voting, those satoshis are colored such that the BSQ network validates them as spendable BSQ. In this way, earning BSQ through compensation requests increases the total supply of BSQ.

As we’ve seen above, BSQ is destroyed when traders spend it, and created when contributors earn it. A key benefit of this approach is that spent BSQ need not—indeed, cannot—be stored, protected, or later distributed by any individual or group. By eliminating the need to hold and later distribute BSQ, we also eliminate a set of difficult problems and risks, including determining who should maintain control over these funds, and how they should be protected and disbursed.

Overall, this approach makes it possible to transfer value from traders to contributors in a decentralized way. The transfer happens indirectly through the process of destruction and creation, but this indirection is precisely what makes it decentralized. The creation side of the transfer is intermediated by voting on compensation requests, but this is not a problem given that voting itself is a decentralized and meritocratic process.

An important property of this approach is that BSQ creation and destruction need not necessarily occur at the same rate. For example, in any given month it may be the case that more BSQ are spent by traders than are earned by contributors. In this case, the net supply of BSQ would decrease during that month. Likewise, given a month in which fewer BSQ are spent by traders than are earned by contributors, the net supply of BSQ would increase. What emerges is a form of monetary policy for the Bisq DAO in which:

Of these three roles, the third is the most important from a policy perspective, as it allows stakeholders a direct means of controlling inflation (i.e. growth in the supply of BSQ). It may be prudent, especially in the early days of the Bisq DAO to operate at a certain rate of inflation in order to fund the development of features that will later result in increased trading volumes (i.e. growth in the demand for BSQ). In the long run, however, we believe the ideal steady state for the Bisq DAO will be one in which the amount of BSQ earned in any given month matches or falls just below the amount of BSQ spent, resulting in a stable or slightly deflationary supply of BSQ over time.

Given current trends, it is important to state explicitly that BSQ is not associated with an ICO (initial coin offering), nor will there be any kind of crowdsale or other crowdfunding event. BSQ is a utility token being introduced into an already functioning system to make it function even better, and raising a large amount of capital up front is neither required nor desired.

It is one thing to design a token-based economy like the Bisq DAO; it is another thing to successfully operationalize it. There are a number of risks inherent to simply "going live" with BSQ and the Bisq DAO, including but not limited to the following:

Any of these risks could be existential. What’s needed is a conservative and incremental approach to bootstrapping the Bisq DAO that accounts for and mitigates these risks. In the next part, we’ll see how the Phase Zero plan is designed to do that.

In July 2017 we put out a call inviting past contributors to opt in to the initial BSQ distribution. We defined a past contributor to be anyone who had spent at least four hours of their time on improving or supporting the Bisq project; we also considered market makers to be past contributors, including in that group anyone with a history of more than 20 trades.

Status: complete with 144 past contributors opted-in.

On October 6th, 2017, we distributed 25 testnet bitcoins to 144 past contributors across 180 BSQ addresses in the initial tBSQ genesis distribution. This action marked the inception of Phase Zero.

Status: complete.

In addition to writing and publishing this document, we have also been engaging in outreach via talks, podcasts and other media to ensure that potential contributors are aware of the Bisq DAO and Phase Zero plan. See this [Twitter thread](https://twitter.com/bisq_network/status/946723541298360320) for a list of these resources.

Status: in progress.

A bounty is a well-defined unit of work that potentially any contributor—​especially new ones—​can work on with as little prior context as possible. We post bounties in the form of issues labeled '$BSQ bounty' in repositories throughout the bisq-network GitHub organization.

Initially, we post these bounties with a set price in BSQ. As quickly thereafter as possible, we move to an "open price" bounty model in which contributors specify a requested amount of BSQ when submitting compensation requests. This approach is consistent with our overall decentralization goals, and more importantly, creates a desirable set of incentives for contributors to determine what they believe to be a fair price for work, requiring them to ask around, look through prior contribution requests, and to generally "do their homework" as to what work is worth when contributing to Bisq. Ultimately, the open price bounty model is intended to foster a marketplace among contributors doing work and stakeholders voting on that work such that a market price for Bisq work is discovered.

Completed work for a bounty is, whenever appropriate, submitted as a GitHub pull request. We follow a C4-based process in which each repository in the bisq-network GitHub organization has one or more designated maintainers responsible for merging or rejecting pull requests.^[3] Merged pull requests do not imply or guarantee in any way that the contributor will be compensated in BSQ for their work; compensation requests are managed separately from pull requests.

Status: pending. See https://github.com/bisq-network/dao/issues/11 for details.

We enumerate and define the roles necessary to operate, maintain and administrate the Bisq project, Bisq network and Bisq DAO, such that responsibilities can be transferred from founders to other reputable contributors. Because each of these roles requires a degree of trust, filling the role requires putting up a BSQ bond. Initially, founders will use their BSQ stake to bond into these roles, and will then transfer those roles to contributors who (a) wish to take the role over and (b) have earned sufficient BSQ to do so.

Status: in progress. Each role has been enumerated in the form of a GitHub issue in the bisq-network/roles repository, and we continue to add to and modify these roles as appropriate. Bonding is not yet in place for any role.

Given that BSQ trading is prohibited during Phase Zero, it is not possible for normal market forces to discover a price for the BSQ token. Still, an estimated price for BSQ is required in order to compensate contributors, and we call this estimated price the fair market value of BSQ, borrowing a term from traditional valuation processes.

Presently, we are working with a valuation of 1 BSQ == 1 USD. As Phase Zero progresses, we expect this estimate and the model behind it to evolve, with the goal of arriving at a maximally credible fair market value for BSQ by the completion of Phase Zero. As mentioned in the risk management section above, this process should significantly mitigate valuation risk when BSQ goes live on Bitcoin mainnet.

Status: ongoing.

Status: ongoing. As of October 2017, all contributions should be captured in one or more compensation requests. See contributor use cases for details.

On the first three days of each month, stakeholders vote on contribution requests and other proposals that have been submitted throughout the previous month.

Status: ongoing.

Initially, founders use their BSQ stake to assume all bonded contributor roles, reflecting the fact that they do already "have all the keys" and do already carry out most of these duties. As contributors earn sufficient BSQ for bonding, they can approach founders and request to take these roles over. Contributors have a twofold incentive to do so: (1) the BSQ they earn by carrying out the duties of the role, and (2) the interest they earn on their BSQ bond over time.

Status: in progress. Founders are distributing responsibility to contributors by having them take on bonded contributor roles, but bonding itself is not yet in place, and may not be until BSQ goes live. We are still sorting out exactly how to handle this during phase zero.

We keep marketing to a minimum during the Phase Zero period, and observe whether our current growth trend continues to play out on its own. Rather than focusing on growing our user base, we now focus on growing our contributor base. The goal is to foster continued organic growth by making Bisq ever more useful and relevant. In any case, we will track the growth trend and adjust course as necessary should it change.

Status: in progress.

We prohibit buying, selling and spending BSQ tokens during Phase Zero in three ways:

As a result, BSQ utility is limited to earning, voting and bonding during Phase Zero. See the risk mitigation section for rationale.

At the beginning of Phase Zero, we use simplistic, high-trust systems such as spreadsheets to track BSQ stake, voting and bonding, and we issue BSQ on Bitcoin testnet vs mainnet. These are pragmatic decisions designed to allow our small team to move quickly and iterate toward trust-minimized, production-ready solutions. Note that the on-chain systems that will replace these spreadsheets have already been designed, and to a certain degree, already implemented. We will migrate to them as they become ready; in the meantime, spreadsheets are easy for everyone to understand and easy to change as we learn through practice what it really means to operate the Bisq DAO.

See vote on compensation requests for instructions.

Eventually, compensation requests will be modeled around Bitcoin transactions. As Phase Zero begins, however, they are modeled as GitHub issues in the bisq-network/compensation repository.

See submit a compensation request for instructions.

Any conceivable change to the Bisq DAO can be submitted as a proposal, and any proposal can be put to a vote. With that said, the goal should be to vote as infrequently as possible. Where good proposals are submitted, and rough consensus is achieved, no vote is necessary.^[4] If rough consensus has not been achieved, then further work is necessary to either improve the proposal or to change minds about it. Voting on a contentious proposal should be seen as a last resort, after all attempts to achieve consensus have failed.

With the exception of compensation requests (which must be voted on in order for BSQ to be issued), voting should be regarded as a crude tool to be used only when absolutely necessary.

The Bisq DAO is founded as a meritocracy, and we believe it is critical to the long-term success of the project that it remain one. Whenever submitting or voting on a proposal to change the way the DAO works, look closely for unintended consequences that may negatively impact this meritocratic foundation.

It is not required to work on an existing bounty issue to contribute to Bisq, and no one is here to tell you what to do. Contributors who have their own ideas are free to work in their own forks on whatever they wish, however they wish, and without any permission from Bisq stakeholders.

With that said, it’s a good idea to consult with stakeholders via the Bisq forum, mailing list, or other communication channels before setting out on any serious contribution effort. Do this in order to ensure your contribution is:

Remember: every contributor is free to work on what they want, including maintainers who may or may not want to review and merge your pull request if they don’t have any prior context for it, or reason to believe it’s worth spending their time on.

Phase Zero began on October 6th, 2017 with the distribution of 25 testnet bitcoins to 144 past contributors in the initial tBSQ genesis distribution.

Phase Zero is expected to last at least six months. We estimate this to be the minimum amount of time necessary for contributors to earn enough BSQ to fill all bonded roles.

When the exit criteria for Phase Zero have been met, a vote will be put to stakeholders whether to complete the phase by (a) issuing the official genesis distribution of BSQ on Bitcoin mainnet; and (b) releasing a version of Bisq that enables BSQ trading, BSQ trading fee payment, and all other DAO use cases.

The repositories in the bisq-network GitHub organization are where changes to Bisq software and documentation are integrated and staged for release. The only contributors who have write access to these repositories are their respective maintainers. Work on bug fixes, features and other improvements is done in contributors' personal forks of these repositories and changes are then submitted via pull request. Our approach to collaboration is based closely on the C4 process.^[3]

issues labeled $BSQ bounty in the bisq-network GitHub organization

https://github.com/bisq-network/compensation

https://github.com/bisq-network/roles

See the Roles document for full details.

https://github.com/bisq-network/proposals

https://github.com/bisq-network/docs

BSQ stake and vote tracking (Google Sheet)

Feel free to discuss the Bisq DAO, BSQ token and Phase Zero plan in the #dao channel of the Bisq Slack workspace.

Testnet BSQ transactions can be explored at https://explorer.bisq.network/testnet.

The following GitHub Issues search query will display all open issues in the bisq-network organization with the label $BSQ bounty:

is:open org:bisq-network label:'$BSQ bounty'

Contributors can browse this list to find open bounties they are interested in working on. Note that the list is currently empty. See the Define bounty issues section above.

Once you have found a bounty issue you would like to work on, it’s a good idea to let others know that you plan to work on it.

Pull requests must be well-formed. Follow the guidelines in CONTRIBUTING.md.

See how to submit a compensation request.

See how to submit a proposal.

During Phase Zero, we do not formally vote on any proposals other than compensation requests. Generic proposal voting will come online with the first production software release of the Bisq DAO, i.e. when we exit Phase Zero.