The Numbers Behind: Kyber Network
7 min read

The Numbers Behind: Kyber Network

The Numbers Behind: Kyber Network

What is it?

Kyber is a decentralised exchange protocol that enables the exchange of over 70 tokens without the need for trusting a centralised counterparty. Token projects, automated market-makers and developers looking to create alpha on idle assets can use Kyber to trade with individuals for a small return. One way to think of Kyber is as an API layer that can be plugged into a DApp or platform to enable instant conversion across multiple pairs. It allows the creation of margin platforms (Nuo), exchanges with no custodial risks (1inch) or even payment relays that can take a wide variety of erc-20 assets and return dai to the vendor. More complex applications such as portfolio management and lending have been created on top of Kyber.

Image Source

Why does this matter? Because we are in a sea of tokens with little liquidity and over time, cross-way liquidity without centralised services will be a staple for the ecosystem to grow. If making a payment to a Coinbase vendor that accepts USDC requires you to sign up on Binance, transfer your erc-20 token, exchange them - withdraw to your wallet and then do a payment odds are low you will go through with the payment. An integration of Kyber, on the other hand, allows you to exchange the asset for USDC, ETH (or any asset supported by them) - and make the payment. The likes of 0x, Kyber and Bancor are a trifecta that have codified liquidity into DApps today. Their growth will be the backbone of what DeFi as a financial ecosystem will become.

The other reason why Kyber has been interesting is the fact that they have managed to handle over $1 billion in trade volume while being (mostly) decentralised. In 2017 when Dex's were coming to market, imagining an open, permissionless project handling so much in volume may have been hard to perceive. They have crossed the metric. They have done it while creating over $1.3 million in fees burned as KNC tokens. These may be early indicators that protocols are finding protocol<>market fit.

User-Behavior on Kyber

About 40% of Kyber's supply is in different exchanges including prominent ones like Binance and Coinbase. An indication of the fact that it is quite frequently used for speculation rather than "hodling".  It could also be that a large chunk of Kyber hold the token in centralised exchanges exchanges. The irony for that would be real.  Either ways - it is becoming clear that as the DeFi ecosystem grew and the possibilities of KNC burn increased -demand for Kyber has increased. Another likely chance is that an increasing number of individuals have been accumulating Kyber for their DAOs launch set to be released later in the year.

0x and Kyber share a similarity in the sense that both tokens have undergone a period of speculatory mania (early 2018) and are now catching up in terms of active users to prior levels while price remains far from there. What remains worrisome for me however is the fact there are only about ~1000 addresses interacting with both protocols on a good day. It may stay this way as increasing amounts of volume gets passed on from wallets and DApps. Individuals may never interact with Kyber directly. To check if the network is indeed growing - we will have to explore volume on trade, the number of trading wallets and the number of trades.

$1.02 Billion Cumulatively Traded So Far

Kyber Network benefits heavily from volatility in the market.  One proof for this is that its highest volume was on March 12th  when around $26 million worth of assets were traded on it. As of writing this, Kyber has had $1.02 billion traded through it since January 2019. At $594 million till Q2 2020, the network has already crossed volume from 2019 ($425 million). There are a number of factors that may contribute to volume traded on Kyber increasing in the coming months

  1. Launch of the DAO could incentivise market-makers to be more pro-active on the network
  2. The token specification explicitly explores inter-operability. With Cosmos and Polkadot coming of age, it is likely that we see live implementations by Q3
  3. The number of tokens that choose to do direct listings (eg: UMA protocol on Uniswap) will increase and Kyber could be used for it
  4. I don't see DeFi coming to a slump anytime soon. The trading apps are trojan horses. As consumer apps come alive on these networks, we will very likely see network congession as traction surges

Number of Traders Up 6x

The definition of a "trader" here is slightly convoluted because they are not restricted to smart contract addresses. They could be wallets, DApps or third party services that redirect funds on behalf of users. Over time, we may see a trend where volume and number of trades increase drastically but the number of "traders"  shown below don't show an uptick as that volume wil come from third parties. As of writing this - there is a wide suite of services using Kyber directly across payments, portfolio management and decentralised exchanges. It is likely that more developers use it with on-ramps (eg: Pay in NFT tokens, receive USD) and savings related products (eg: convert yield from a staked asset to DAI) in the coming months. The number of traders on Kyber is one of the few metric I have seen to have a hockey-stick growth in the space over the past few years and it is impressive for that reason.

Another way to measure traction on the network would be to explore the dollar value of each trade and trader on the network. The highest for this figure was also on March 12th - at roughly ~$16,000 dollars coming in from each contract doing a trade and ~$3500 moving on the average trade. The higher these figures are - the more likely indication that whales do use decentralised finance products. There's two things to note here

  1. The average trade on a DeFi protocol will likely be higher than what you see on a centralised exchange given resources (fee, time) spent. More proficient users that are comfortable with handling their private keys and interacting with dex products use these.
  2. The chance of this figure dropping drastically is quite high with scaling coming to the market. An early example of this is what Synthetix has with synthetix's exchange

$1.3 Million In Fees Generated

Kyber Network's fee model relies on exchange transactions accruing a small amount in KNC as fees. These tokens are then burned over a period of time. The model is similar to that of Binance's. What I found interesting is that this amount has been steadily rising. Higher volatility means more traders use dex's and this converts to a fee for the network. I anticipate this figure to only continue rising as the number of applications built on the protocol increase in count

If Kyber were a software business at a monthly revenue of $200k - it would have been one of the more impressive startups in the space. However, it isn't. It is a protocol. And as a protocol - where its value accrues is a reasonable question.This is not to suggest it's feats as a protocol is not major.  On the flipside, one could argue that token based incentives is exactly what has made it possible for Kyber to scale to $1 billion in volume and an early stage start-up may have struggled to show similar figures. My key take-away from working through these numbers were that

  1. The idea that protocols have no product <> market fit is almost nonsensical. They are here and scaling healthily enough to be considered a series A startup. How we get them to the levels of a unicorn is where the challenges of the next 2-3 years lie
  2. Open-permissionless finance has strong network effects that relay back to certain base protocols that power core utility (eg: exchange protocols). These are primed for growth
  3. Token economics really matter. In comparison to $1.3 million on Kyber (since Jan 2019), 0x only had about ~$60k in fees generated (since December 2019). Granted 0x has a more robust ecosystem in terms of market-makers - how a design is modelled matters heavily.

While these criticisms may seem harsh - to give credit where it is due, both protocols have been solving for exactly this. 0x is about to release a new token economic model (v3) and Kyber is in the process of launching a DAO to make stakeholder driven governance a reality. I should be covering them in the days to come.

See you sometime later this week with a market-map on Asian blockchain startups


Notes :
1. Data sources : Santiment, Nyctale & Nansen
   Hit me up - I may be able to get you a discount on those
2. I have no affiliation to the Kyber team nor do I hold any Kyber tokens
3. This is obviously not investment advice
4. If you want to do a DCF model on Kyber and discuss how scaling / inter-operability will affect DeFi - check the about page and drop me a DM / e-mail.