Thoughts on Decentralized Exchanges

Hundreds of projects around the world are working on Decentralized exchanges (DEXs).

An extensive list can be found here:

The most interesting DEXs in this list fall under the DEX Grade “Fully” (fully decentralized).


Benefits of a DEX

Censorship Resistance

Centralized exchanges can censor listings and trades.

Earlier this year Japanese regulators forced domestic exchanges to delist Monero, Dash, and Zcash due to money laundering concerns.

A couple of days ago the Indian government shut down Zebpay, one of the country’s biggest exchanges.

A truly decentralized exchange would provide total freedom of choice of investments.



Governments are increasingly forcing centralized exchanges to KYC users.

This personal data and trading information is often reported to tax agencies. Coinbase were forced to give the IRS data on users who did more than $20,000 in annual transactions between 2013 and 2015.

Gemini have also been sending personal data to the IRS:

Using a centralized exchange just ruined everything for me.

Gemini sent an false 1099-K on my behalf to the IRS. I put coins in, traded, then pulled them out. I did this repeatedly. Gemini counted every deposit as INCOME, as if my deposits weren’t my own. I am helpless.Jordan Kittle

The Indian government is considering a retroactive 18% “consumption tax” on previous cryptocurrency trades.

Tying personally identifying information to bitcoin transactions is arguably the riskiest thing you can possibly do, de-anonymizing yourself from the get-go on a permanent public ledger.

We have no idea what the future holds. Privacy matters.Bisq


Users Maintain Custody of Their Funds

Depositing funds to a centralized exchange creates a huge single point of failure. Numerous thefts have occurred over the years, most famously the Mt Gox debacle.

On a DEX, users keep control of their private key. This can protect coin holders from being singled out for coercion.

For example, Changelly is withholding users’ Monero until they provide KYC data, saying “Monero is the crypto that hides a sender and recipient thus making transactions untraceable”. If the data isn’t submitted Changelly will keep the funds indefinitely.

Users keeping hold of their funds also prevents the problem of billions of dollars of fractionally reserved coins at centralized exchanges inflating the money supply.


Low / No Fees

Running a compliant centralized exchange is very expensive. Disintermediating trading will drive down the cost of trading the same way crypto has for transmitting money.




Low Volume

Earlier this year all DEXs combined were reported to be a tiny 0.4% of trading volume against the top 10 centralized exchanges.

This lack of liquidity results in high spreads on trading pairs, making it inconvenient and expensive for users.

Flix1 notes, “Their low liquidity is a chicken-and-egg problem… But it really takes very little to bootstrap it. Only a handful of traders are needed in a particular currency pair to make them useful.

“I’m really amazed that traders with $millions in counterparty risk at major exchanges are not devoting a tiny effort to bootstrap decentralized exchanges… Not to mention traders in countries where centralized exchanges are censored or undeveloped. What are they waiting for?… Having a plan B is essential. Expect withdrawal blocks at multiple large exchanges.”

In the last 24 hours the leading exchange overall was Binance with over a billion dollars of volume.

The DEX with the most volume was Bitfinex’s Ethfinex with $2.6 million of volume, making it the 97th biggest exchange on coinmarketcap.

Ethfinex’s Nectar token has more than doubled in price over the last 6 weeks.


Tricky to Use

A common refrain is that DEXs are struggling for adoption because they offer a terrible user experience.

It’s true that centralized exchanges tend to be more intuitive and require less technical skills to use. Submitting orders on a DEX to a decentralized order book can be slow with costs that centralized exchanges don’t directly impose. For example, Bisq users have to pay fees to publish offers which can’t be cancelled without the fee being lost.

However, the overall experience of using a centralized exchange is hardly convenient considering all the bureaucracy and risk incurred.

A DEX designed correctly could potentially be a smoother experience because giving personal data to set up an account and waiting for account verification isn’t necessary. KYC adds significant friction that a DEX doesn’t have to deal with.


Fiat to Crypto Isn’t Viable at Scale

Andreas Antonopoulos explains here why fiat/crypto pairs aren’t a good fit for a DEX.

Despite these risks, Bisq is promoted primarily as a fiat/crypto trading platform and so far has been able to keep chargeback scams under control by not supporting PayPal, Venmo and Cash App.

To help people mitigate the risks, the Bitcoin Wiki has ranked the degree of chargeback risk for various fiat payment methods:

Bitcoin Wiki:

Despite this, receiving fiat deposits into a personal bank account from an unknown source is undoubtedly risky for the seller of crypto.

If the buyer of your crypto becomes suspected of criminal activity then your bank account can be frozen on a police order and you’ll be brought in to the investigation as a money laundering suspect. An example of this happening in Thailand is documented here.

Dealing with fiat bank accounts is a nightmare and will continue to be so. Real progress is only going to come from crypto to crypto DEXs. Well-developed closed loop crypto economies that bypass fiat entirely can’t come soon enough.


Most “Decentralized Exchanges” Today Aren’t Really Decentralized

Most so-called DEXs today are built on Ethereum. The custody of the funds on these DEXs does tend to be decentralized (users hold their private key) but most of the other aspects are still centralized. At best these are hybrid exchanges.

Bancor froze funds when their account was hacked earlier this year. This was possible because of a pause/freeze function in the smart contract that users have to interact with.

IDEX, by far the biggest Ethereum DEX by number of transactions, has also frozen funds in a similar manner.

Nic Carter points out that decentralized and freezable are mutually exclusive concepts.

Projects like these benefit from marketing themselves as a DEX when raising funds and attracting users, but what they deliver in reality is disappointing.


Warning Signs That a DEX is Actually Centralized

Listings are Centralized

The IDEX team have full control of what gets listed and delisted on their “DEX”.

IDEX have already censored several coins. The delisted coins were gambling tokens and security tokens that paid dividends. IDEX commented “We are keeping official regulations in mind with our token listing criteria.”


The Order Book is Centralized

On-chain order books are slow so today’s DEXs often skip this part and host the order books on centralized servers.


Matching/Processing the Orders is Centralized

Giuseppe Bertone explains that it’s not possible to trade without the IDEX team’s permission. “If they don’t want to process your transactions, your transactions will not be processed… you can check the smart contract and confirm that only admins can send trades to the smart contract.”

Udi Wertheimer adds that popular 0x relayers also have to sign all transactions in their system for trades to take place.


Degrees of Centralization

The centralized parts of these DEXs are often justified as being a temporary fix in the early stages of a transition phase towards decentralization.

Jill Carson says that today’s DEXs are a promising start that offer different degrees of centralization. She believes that “all aspects of the spectrum have an important role to play”.

The main benefit most offer so far is they are non-custodial. Unfortunately this tends to be ruined by having a pause/freeze function in the smart contract.

John Backus has studied the history of decentralization and observed in “Resistant protocols: How decentralization evolves” that “each successful phase of decentralization we saw represented the minimum viable decentralization necessary to stay alive…. Over applying decentralization isn’t a strategy unless your goal is obscurity; the most popular file sharing applications were never the most decentralized.”



Notable DEXs

Bisq (Formerly Bitsquare)

Intro to BISQ:

Source Code:


Bisq is a small but rapidly growing exchange (the 184th largest exchange by volume) that uses multi-sig escrow.

Last month was a record for the exchange with 358 BTC traded, a 26% increase over the previous best month.

Interestingly 98% of transactions are Monero/Bitcoin with hundreds of thousands of dollars’ worth traded every day.

Bisq appears to be benefiting from its focus on privacy (users access the site over Tor using a p2p protocol) following the centralized exchange crackdowns on Monero and the bad press Shapeshift has received over KYC.

Installing the Bisq desktop app, connecting to Tor and getting it up and running is easy.

The platform works well and the user interface looks decent enough, but it’s not immediately intuitive how to start trading crypto-to-crypto (fiat-to-crypto trading is a big focus of the platform).

It’s a bit disorienting for a beginner to figure out how to test the waters and get to grips with all the variables (the minimum amount possible to do a test trade, considering what needs to be risked/lost when factoring in all the fees/security bonds/the spread from lack of liquidity, how long funds will be tied up for, etc.)

The FAQ indicates that both the buyer and the seller need to put up a security deposit. Veteran traders will probably adjust to all this quickly, but there’s quite a few hurdles for newcomers to get past before they’ll be comfortable with the platform and the risks involved.

Bisq contributors are currently working on the Bisq DAO which should result in even more censorship resistance and alleviate the concerns over the amount of centralization in the current arbitration process.

Bisq does have a token (BSQ) but it isn’t trading on any exchanges yet and can only be only be earned as a reward for helping the project.

Overall, Bisq is an interesting, working DEX with a lot of promise, particularly with Monero volume picking up recently (I’m very bullish on Monero).


BarterDEX (Komodo)

Source Code:

The exchange claims to have already completed over 100,000 atomic swaps (Trading Stats: )

On closer inspection it looks like nobody actually trades here. Total trading volume over the last 24 hours was a grand total of $1.

Komodo’s coin trades at $1.2 with a market cap of $132 million (ranked 51st on coinmarketcap)


Atomic Wallet

Atomic Wallet has a nice user interface and claims to offer non-custodial atomic swaps with a decentralized order book.

Cross-chain atomic swaps enable coins to be exchanged without a trusted third party.

However, this is closed source software. Installing this reintroduces the trust that atomic swaps are supposed to remove.


Block DX (Blocknet)

Source Code:

Blocknet claims their block DX is a fully decentralized exchange (non-custodial, decentralized order book, matching, and coin exchange).

Blocknet’s coin trades at $4.37 with a market cap of $23 million (ranked 188th on coinmarketcap).



OpenBazaar is a decentralized platform with some zero fee crypto to crypto buying and selling going on.


Ethereum ERC-20 DEXs

* DEX tracker:

* DEX order books:

These DEXs only trade tokens that are supported by the ERC20 token standard which now enjoys a powerful network effect as projects and users converge on it. 95 of the top 100 tokens by market cap are Ethereum based.

One thought that struck me: DEXs for ERC-20 benefit from a single, open standard. When we were building integrations with exchange APIs, every single one required custom code.

Now that we’ve thrown away that code to focus on tokens, we benefit from an open system. Speed.

Richard Burton

Unfortunately Ethereum’s lack of privacy negates most of the benefits associated with a decentralized exchange. Reddit user Aminok explains:

“Everyone is still using the same zero-privacy, one-account-per-user model, of the very first Ethereum wallets.

“This is simply a no-go for any serious financial use. And blockchains, with their unparalleled immutability and censorship resistance, find their greatest usefulness in serious financial use. The inability to be fully utilized for that purpose due to a lack of privacy is a major waste of their potential.

“Most were expecting ring signatures or zk-SNARK based transactions by now, but we don’t even have something that can match Bitcoin’s UTXO model, with a new account per transaction.”

The DEX project with by far the most hype and expectations is 0x. It has an enthusiastic community and is ranked 28th on coinmarketcap with a price of $0.66 per token and a market cap of $354 million.

0x is a decentralized exchange protocol that enables people to build their own DEX/order book services on top of it. These entities are called relayers.

The relayers are non-custodial but host centralized order books to improve the user experience.

Building on the 0x protocol gives these relayers interoperability, allowing them to interact with each other to solve the severe liquidity problem of DEXs by creating a combined liquidity pool.

There are currently dozens of 0x relayers, the most famous being Paradex which was acquired by Coinbase earlier this year. KYC enforcement is a possible future for these relayers.

I bet at least one of IDEX, EtherDelta, Bancor, 0x add KYC restrictions by EoY 2019. – Udi Wertheimer

However, research from states “since the decentralized exchanges are only facilitators of orders, they are technically not required to comply with KYC/AML regulations like traditional exchanges. This means that most of the DEX allow to trade anonymously without disclosing any personal information.”

A 0x team member has stated there is “No clear regulatory guidance on DEXs in terms of state money transmitter laws, FinCEN, etc…. Working with legal counsel and regulation researchers in the space to potentially get no action letters for relayers, we’ll see…def don’t want KYC ever”.

This legal grey area may be analogous to the history of file sharing, where John Backus observed that mass adoption was based on Minimum viable decentralization — Is the technology clever enough that Hollywood can’t just kill it immediately?


Proposed/Under Development DEXs

Binance DEX

The Binance DEX is many people’s hope for the first ever consumer friendly, highly liquid DEX.

Yesterday the company confirmed that BNB Coin will be the native gas and the exchange won’t control user funds.

A public beta is targeted for the end of the year/early next year.

BNB currently trades at $10 with a market cap of $1.2 billion (Ranked 16th on coinmarketcap).


Huobi Chain Project

Centralized exchange Huobi is also planning its own DEX that will coexist alongside its centralized exchange.

This is a heavily funded project, with $166 million set aside to fund its development.

We’re not 100 percent sure if a corporation can be 100 percent autonomous. We’re not sure if it can be 100 percent decentralized either. But we believe there can be some kind of balance in between.

– Gordon Chen, Huobi Chain Project’s executive leader

Huobi Token is ranked 69th on coinmarketcap with a price of $1.86 and a market cap of $93 million.


Decred DEX

In June Decred announced plans for “A New Kind of DEX”.

The project will start as “a simple client-server architecture and then extend it by having the servers relay orders between each other, creating a mesh.”

The Decred team are some of the pioneers of atomic swaps, demonstrating an on-chain atomic swap between bitcoin and litecoin last year.

Decred currently offers tools to manually perform cross-chain atomic swaps. At the moment, support exists for the following coins and wallets:

Bitcoin (Bitcoin Core)

Bitcoin Cash (Bitcoin ABC, Bitcoin Unlimited, Bitcoin XT)

Decred (dcrwallet)

Litecoin (Litecoin Core)

Monacoin (Monacoin Core)

Particl (Particl Core)

Polis (Polis Core)

Vertcoin (Vertcoin Core)

Viacoin (Viacoin Core)

Zcoin (Zcoin Core)


The Decred team explains that “using atomic swaps, it is possible to perform trustless exchanges of supported cryptocurrencies both on-chain and off-chain. In order to generate and maintain an order book, there needs to be a meeting place where users can communicate about prices.”

“Due to the trustless nature of the atomic swap, it is a natural basis upon which to build a DEX…. An atomic swap can handle settlement of matched orders both on-chain and off-chain, so the task of building a DEX becomes a question of handling the order matching process.”

Decred estimates that the DEX can be built for USD 1-5M in 6-18 months.

Decred’s coin is currently priced at $39.04 with a market cap of $334 million (coinmarketcap ranking 30th)


Lightning Network Based Atomic Swap DEXs

The Lightning protocol enables interoperability between chains using the BOLT standard (Basis of Lightning Technology).

Currently Bitcoin and Litecoin have their own Lightning Networks. The compatibility between Lightning Networks enables payments to be sent in Bitcoin and received in Litecoin. Lightning Labs demonstrated a cross-chain Lightning atomic swap between the Bitcoin and Litecoin testnets in November last year.

It’s possible that in the future BOLT based DEXs with extremely low trading costs will be available.

Andreas Antonopoulos explains: “I expect we’re going to see many more currencies added as the BOLT spec is ported to other blockchains…. If a chain does not have a transaction malleability problem and it has the basic capabilities of simple smart contracts, specifically multi-sig hashed timelocks, then you can implement Lightning that’s compatible with the BOLT standard which would then allow you to do multi-currency swaps between that chain, Bitcoin and Litecoin.

“If people port the Lightning software, or implement their own Lightning software using the BOLT specifications, to their blockchain then eventually we will be able to see a multi-currency lightning running across multiple blockchains.

“Other blockchains can build their Lightning Network that is compatible with the BOLT standard, enabling cross-chain Lightning atomic swaps. And I expect that someone’s going to build an easy to use Lapp that allows you to do cryptocurrency conversion in a decentralized manner over Lightning. It’s just going to take time.”

Cross-chain atomic swaps also have important privacy implications (e.g. atomic swapping Bitcoin and Monero). Last month a Monero Research Lab paper was published that describes how confidential cross-chain atomic swaps are possible based on thring signatures.


BTCPay Server’s Open Source Shapeshift

BTCPay Server is a self-hosted Bitcoin Payment processor that plans to create a self-hosted p2p exchange that will enable anyone to be a liquidity provider.

This idea was announced in the aftermath of Shapeshift’s KYC announcement a few weeks ago.

BTCPay Server explains it will be “p2p trustful order matching between servers. (meaning you need to trust your peer up to a certain amount). No fancy stuff, just the minimal amount of work required to replace ShapeShift”.

BTCPay server estimates that around 100 man hours are needed for the initial goal of getting the same UX as Shapeshift (design, coding, testing, marketing), but at that point it would not be a trustless solution. The long term goal is to make it trustless.



Summary of DEXs

Overall, the current state of DEXs is pretty disappointing but the proposed/under development projects sound promising. Atomic swaps seem like an essential feature going forwards.

The Future – Balkanized Chains or Interoperability via Atomic Swap DEXs?

At the moment, all blockchains are entirely segregated from one another. DEXs aren’t doing a great job yet of painlessly bridging them.

Scott Stuart, CEO of Kava Labs gives his perspective on this segregation:

The specialization of blockchains is an outcome we’ve seen in the developing Web3 stack as a result of the scalability trilemma that requires blockchains to choose fundamental trade-offs between decentralization, throughput, and safety.

1) dApp Platforms (ETH, EOS, ADA, etc)

2) Store-of-Value Blockchains (BTC)

3) Privacy Blockchains (XMR, ZEC)

4) Stability Blockchains (Basis, DAI)

5) Specialized Application Blockchains — designed for specific applications (SIA, STEEM)

The key question is how will these segregated chains fit together in a future open financial system that’s built to reduce costs and friction?

If everything goes well, seamless DEXs, atomic swaps and advances in interoperability protocols means you’ll be able to pay anyone instantly in any currency with an automated trustless conversion.

Ari Paul lays out the vision for this interoperable cross-chain world:

“We can readily imagine base protocols optimized for specific use cases that beat Bitcoin at those use cases currently, like IPFS for decentralized file storage, or Ethereum for hosting the assets of a mobile game.

“At some point in the future, programs will likely be able to seamlessly convert one currency to another via decentralized exchanges without the user even being aware.

“Can friction be reduced enough such that users are simply happy to use IPFS?  From their perspective, they’ll be using BTC or USD, but their currency may be converted in realtime to Filecoin.”


Market Share Among Cryptos in an Interoperable Atomic Swap DEX World

It’s interesting to think how this highly interoperable future would play out in terms of which coins would accrue most of the market cap.

One theory is that low cost instantaneous interoperability means the network & any network effects is the entire cryptocurrency network, not only a single coin, such as bitcoin.

People cheering for Atomic Swaps and DEXs: note that the more widespread, cheap and convenient they become, the weaker you “network effect! winner takes-all!” arguments become. – imaginary_username‏

The opposing theory is that, like today, most of crypto’s market cap would be stored in the most secure and tested censorship resistant coin, probably even more so.

If we really do end up with instant and frictionless exchange, this theory makes more sense to me because features that can only belong to 1 coin (the most secure, censorship resistant, inflation resistant, with the longest track record) will be even more important when deciding what to hold. Whatever features are lacking on bitcoin could just be accessed on demand by painlessly atomic swapping into the smaller, less secure coins with the required features as needed.

If atomic swap DEXs and future interoperability developments can create a crypto ecosystem this frictionless it will be mind-blowing. We seem a long way from that though. Interoperability is hard. In the healthcare industry it’s “always 5 years away”.

It’s possible that converting into other coins might not turn out be as smooth and seamless an experience as hoped, even with atomic swap DEXs. Moving into different blockchains for different features might remain a slow and costly experience.

interoperability is like this dense fog which completely impedes my ability to even guess at what the industry will look like in 3 years.Nic Carter

Overall, much progress is needed to improve on today’s limited DEXs, but the ideas being put in place for the next few iterations of these projects have the potential to completely change the game.


Follow me on Twitter @leebanfield1

Bitcoin: 3ETVv9gvyDo6DQwt1EnXKfiirjToeRbPJf


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: