In February 2014, Mt. Gox went bankrupt. US$460 million gone. The company that had not long before managed up to 80% of all bitcoin trading collapsed. Investors had little information or rights.

They can now claim some of their money but for prices that are small fractions of what bitcoin is worth today. Hackers took advantage of poor technology infrastructure, and the central exchange for cryptocurrencies went down the drains, taking most of the stored coins with it.

The unnerving part of this story is that the crypto world promises security through decentralization, and here centralization was the pain point.

What happened to Mt. Gox is a perfect example of why we need blockchain technology: Centralised systems are central failure points. They are hypocritical antitheses to the crypto world’s fundamental promise — to decentralize and democratize, and thereby also make the world more secure.

The Solution: Decentralized Exchanges

They offer security by taking away the single failure point of centralized token and currency storage and returning control over funds to users. Through decentralized trade mechanisms, they may create a global, incorruptible, price-driven marketplace.

The Levels of Decentralization in Exchanges

Six levels of decentralization currently exist in cryptocurrency exchanges:

  • Traditional full centralization, including market leaders Coinbase, Bitfinex, Kraken, and others. They dominate the market with 99% of trades.
  • Bancor’s unique project which has created smart tokens that hold several coins and self-convert.
  • Decentralized exchanges with centralized order books, most prominently EtherDelta.
  • Open protocols which allow decentralized applications to run the order books, most notably 0x and, if it goes through with current plans, OmiseGo.
  • Decentralized exchanges that allow child chain applications like IDEX in the Aurora ecosystem
  • Decentralized exchanges that use atomic swaps, like Blocknet and Airswap

The first level, as Mt Gox and many other failures have shown, is dangerous and unsustainable; it also betrays the fundamental promise of blockchain technology. For users to abandon them, however, other models must show that they are better.

Overcoming the big barrier to decentralization

A big obstacle for decentralization is an order book, where traders post buy and sell requests. Individual users and especially market makers — traders who bring liquidity to markets and put buy and sell orders for many different assets or tokens — add or remove orders constantly. The sheer amount of trading orders is too much for an on-chain solution to handle.

One decentralized solution is Bancor’s smart token, which operates the exchange itself. Bancor not only eliminates the third-party broker, but also the exchange trading partner. To exchange a token, a user only needs to interact with a smart token that holds tokens in reserve, and so trades do not need to rely on the presence of other offers. However, because the Bancor protocol encodes and influences the price discovery through its tokens, it limits market efficiency.

To preserve an open market, most solutions still maintain order books, for which there are different approaches.

Centralized order book, decentralized settlement

EtherDelta secures tokens in smart contracts and thereby decentralizes token storage. However, their order books and matching mechanisms for traders are still centralized to provide the necessary transaction speed.

Open protocols like 0x are taking the first step beyond this. They allow for market makers, called relayers, to offer their own off-chain order books on top of the network. The relayers do the matching of buy and sell offers. This eliminates the full centralization of order books, and through incentive systems brings liquidity providers onto the network, because the network rewards relayers with tokens for storing coins and performing the trades. All this runs on top of smart contracts that cryptographically secure both the order books and the trades, similar to EtherDelta.

Decentralized order books, shared liquidity

A more ambitious project is the Aurora DAO ecosystem. Their solution is to connect a network of different exchange platforms into one large shared liquidity pool. Although for now it looks more like EtherDelta, it plans to allow many different market makers to operate autonomous sidechains and to then connect them to share liquidity. This combines the needs for scalability and speed through the interoperability with sidechains with the need for liquidity. The ecosystem incentivizes placing orders with coins, and then users can just execute on any order.

Peer-to-peer swaps replace the order book

The most promising option for future scalability and transaction speed are atomic swaps, which will make order books redundant. Atomic swaps rely on cross-chain channels directly between two users or, in Blocknet’s case, communities of users, who can exchange tokens rapidly off-chain. The channels are hash time-locked contracts, meaning that they give exchange traders limited time to confirm payment or else the failed payments are enforced on-chain, which secures the transaction. Watch for the launch of the Lightning network for the future of this technology.

What is holding them back?

Although these projects have a lot of potential, they still have to overcome development hiccups before widespread adoption.

Firstly, current technology is not scalable enough for thousands of on-chain token and currency exchanges every second. Along with the lack of transaction capacity comes, for now, increasing cost and slow speed of transactions.

Secondly, many current approaches offer no convincing user interface and may scare people away because of the added responsibilities for managing one’s own private keys.

Lastly, most decentralized platforms suffer from a lack of liquidity which makes the exchange of coins very difficult and inefficient. Even in the crypto world, many are unaware of decentralized exchanges. This is a ‘chicken-and-egg’ problem because once more users will trade on the platforms, they will also bring liquidity. The most challenging part will be the development of fiat currency liquidity pools.

Decentralized exchanges are the future

The different projects and protocols will need to work on their marketing and their user interfaces so that many more people in the crypto world see the need to switch. They should advance in their development of trustless peer-to-peer marketplaces that make middlemen redundant. Most importantly, they need to emphasize and implement intelligent incentive structures to provide for liquidity and establish efficient pricing.

Many already have useful approaches to this but will need to keep upgrading their services to appeal to wider audiences.

We are as curious as you are about the future of blockchain and how it can change this world for the better! Join us in talking about this and many other exciting developments on our Telegram.

Want to work on the future of the blockchain? Join the Zerion team, email us at [email protected]