Singapore-based Eximchain has gone live with a blockchain-based solution for supply-chain finance that relies on an innovative method to enhance security, in order to expand the service’s reach beyond an anchor corporation’s direct relationships.
The tool is a consensus mechanism called quadratic voting, or QV, and Eximchain’s co-founder and CEO, Hope Liu, says her company is the first to put it into practice.
“We’re talking about trusting data from strangers,” she said. “You have to be able to trust the entire network, instead of the single party you know.”
The potential that blockchain brings to supply-chain finance is letting companies up- or downstream from an anchor corporate buyer (e.g., a Walmart or a Samsung) get access to loans or financing by proving their relationship to that big corporation.
Traditionally banks will service a direct supplier, but a fourth-, fifth- or sixth-tier company cannot prove its relationship in the value chain. The anchor company may not even know these suppliers exist. And investors don’t have a way of judging a borrower’s risk without a means of confirming such relationships.
Those gaps add up: there is an annual $2 trillion financing gap for small- and medium-sized enterprises, says the International Finance Corporation.
The company was set up in 2015 at Massachusetts Institute of Technology. It went live October 5 on Quroum, a blockchain developed by J.P. Morgan, but modifying its usual consensus mechanisms. Its first clients including a Fortune 500 tech company that it would not name, and a China-based company, YOOSourcing, that has a digital platform matching companies with suppliers.
The 35,000 suppliers on YOOSourcing’s platform can use Eximchain to verify their relationship with large corporate buyers, with a long-term goal to secure financing and new contracts at favorable terms.
We’re talking about trusting data from strangers
Liu says Eximchain intends to create a marketplace to securitize and trade SMEs loans. “These are assets that are either backed by the anchor buyer’s credibility, or by inventory validated by the DLT network where everyone can check the history of the record”, she said. “Everyone knows this model makes sense, but there have been issues at the infrastructure level.”
Those issues are security and scale.
Scalability is a common problem for blockchains. Eximchain’s solution is to only use the blockchain for validating transactions, but not as a database. “To store data,” Liu said, “you should use other traditional databases, which are a thousand times more functionable.”
It’s in the realm of security that Eximchain is bringing a new model.
Voting as intensity
Putting all of these suppliers, investors, corporations and banks on such a platform makes its ledger more like a public blockchain, so security is important. Participants want to avoid a situation such as with bitcoin, where a handful of miners now have massive influence.
Eximchain therefore introduced QV to create consensus on the blockchain. QV lets people vote based on intensity of interest: you can vote more than once, but each additional vote costs more. Participants are given a budget in the form of tokens. Instead of just voting yes or no, they can amass their budget for issues that they feel strongly about.
The cost per additional vote increases quadratically, i.e., it squares with each iteration, like placing bets in backgammon: first vote=1 token; second vote=4 tokens; third vote=9 tokens...
What this does is make it incredibly expensive for someone to control the vote.
Eximchain will soon launch an app so participants can acquire tokens and do quadratic voting; the company issued tokens this year that users can purchase to cast votes. Tokens are redistributed evenly across the network after each vote, so that someone who only cast one vote will receive extra, a passive form of income, while those who want to influence the vote end up with fewer tokens handed back.
This way there is a chance to provide rewards to users and to encourage more participation.
Double sale on blockchain
In the case of validating blocks, QV in theory can help avoiding double sales on blockchain.
Paul Sin, Deloitte’s Asia-Pacific blockchain lab leader, says participants lack a central authority (such as a stock exchange) to verify a transaction on public chains.
“After you bought an asset, you need to update the data to all the nodes, so the data nodes can prove that you own the assets. As long as 51% of the nodes vote for your ownership of this asset, you own it,” he said.
Imagine a power company sells its spare quota of carbon dioxide emissions to several different parties, and then leverages its low electricity cost to mine tokens and uses these tokens to vote, and eventually controls 51% of the chain to vote for its ownership of the quota even after it’s sold.
We need to think how those attacks will work
“If this blockchain only has 10 nodes, the power company can validate another transaction with 11 nodes as long as the additional revenue can justify the marginal cost,” he said.
QV has emerged as a tool to mitigate against this happening; it has been cited by Ethereum founder Vitalik Buterin as a leading candidate to governance problems.
Eximchain’s Liu says QV means no single player can continuously control the voting. “He would lose a big amount of tokens even if he managed to win a vote,” she said.
Looks good...on paper
However, what looks good in theory has yet to be deployed in the real world.
“It’s good to figure out a way to make the blockchain more decentralized, but we first need to think how those attacks will work,” said Duncan Wong, founder of CryptoBLK, a Hong Kong blockchain developer working with HSBC.
Lei Chen, professor at Hong Kong University of Science and Technology, says researchers continue to seek ways to make blockchain more safe and efficient. No consensus mechanism is satisfactory: proof of work consumes too much electricity and results in new centres of a few mining pools; proof of stake gives too much power to rich people; another consensus mechanism called Byzantine Fault Tolerance requires too much resources to engage communications of all nodes. QV might prove better – but it needs time to prove itself, he says.
Another problem of QV is that participants could multiply their influence by misrepresenting themselves as multiple individuals. One reason why Eximchain chose to use Quorum was to take advantage of its KYC protocols.
Here’s Liu’s more technical explanation:
“The changes we have made to the consensus mechanism turn the consensus model to a proof of authority mechanism managed via smart contract. It is important to understand the difference between how participants vote on consensus for new blocks (this does follow a Byzantine Fault Tolerance/Proof of Authority model) and the governance mechanism (quadratic voting) which democratically controls who is part of the consensus pool. The end goal here is to stabilize transactions’ gas costs, and hold consensus participants accountable to the users of the network.
“As an example: while on vanilla Ethereum anyone may anonymously mine blocks and charge arbitrary transaction costs (effectively meaning large pools control the price of gas), on our network block makers (“miners”) must put their identity at stake and propose the minimum gas price they will accept (and be voted out if they deviate) in order to be elected to participate in consensus over new blocks.”