If you’re struggling to get your head around the endless fintech buzzwords and jargon in trade finance, you are probably not alone. Sanne Wass explains the most important concepts you need to know about.


Blockchain and distributed ledger technology

Blockchain is a digital, decentralised ledger in which transactions between two users can be recorded in a secure, verifiable and permanent way. It’s a form of distributed ledger technology, where every transaction goes into a “block”, which can be likened to a page of a public record book. Each block contains a reference to the block that came before, hence the name “block-chain”, which makes it nearly impossible for a user to tamper with previously recorded transaction data.

The technology allows for a peer-to-peer network, which is updated in real-time without the need for intermediaries.

While blockchain is what underpins cryptocurrencies, many blockchain applications today involve recording assets other than currencies: in healthcare, for example, it could be used for patient records, in real estate it could record and transfer land titles, while in trade finance, it’s being explored for the exchange of letters of credit and bills of lading.

One can generally distinguish between public, “permissionless” blockchains on the one hand, and permissioned blockchains on the other, explains Alisa DiCaprio, R3’s global head of research and trade.

“In a public blockchain, anyone can run a node and download the full history of the ledger. While the identities of the parties on the ledger are abstracted by combinations of numbers and letters called addresses, the transaction details are publicly accessible,” she says.

Bitcoin, the world’s first cryptocurrency, is one of the largest public blockchain networks in production today. Another is Ethereum, which was launched in 2015, and is being used to create applications beyond supporting a digital currency.

But there are disadvantages of a public blockchain: for example, a substantial amount of computational power is needed to maintain a distributed ledger at a large scale, and other activity on the network can impact the capacity. A famous example is that of CryptoKitties, a blockchain-based virtual game that allows players to purchase, breed and sell virtual cats. The game’s popularity in December 2017 congested the entire Ethereum network, causing it to slow down significantly.

Another drawback of a public blockchain is the lack of security and privacy for transactions. These challenges have led many firms to deploy permissioned, often private, blockchains for enterprise use.

As opposed to a public blockchain, a permissioned blockchain network requires an invitation and must be validated by the network participants or owner. “Permissioned blockchain is a term often used interchangeably with private blockchain,” DiCaprio explains. “They require that potential participants undergo a vetting process so that legal entities can be associated with the transactions. While an entire blockchain can be permissioned, sometimes just specific business networks within the ledger will be permissioned.”

The idea of a permissioned blockchain is one that is crucial for financial services like trade finance, Keith Bear, vice-president of global financial markets at IBM, explains. “The critical thing for trade finance, and this is true for anything which involves financial services institutions, is that they will always need to know who the client or counterparty is, for anti-money laundering and know your customer reasons. So when we’re seeing blockchain being deployed in financial services, be it in post-trade, trade finance or anything else, the technology to support that has to be permissioned.”


Fabric vs Corda

A range of consortia are currently working to bring blockchain technology to trade finance, with IBM (which uses the Hyperledger Fabric framework) and R3 (which launched the Corda platform last year) being two of the most active blockchain firms in this space. HSBC and ING recently conducted their first live, commercial letter of credit transaction on blockchain using the Corda platform. Meanwhile, we.trade, a platform for open account SME trade developed by a consortium of nine European banks and IBM on Fabric, is now in production.

Fabric and Corda are both examples of enterprise blockchains which fall under the category of private permissioned ledgers. However, the two are designed slightly differently. Fabric uses so-called “channels” to let participants set up mini private blockchains, where data is shared with all participants. With Corda, meanwhile, data is shared at the level of individual deals or agreements. Corda also departs from the architecture of conventional blockchains in that it actually doesn’t use “blocks”. This means that it technically speaking isn’t blockchain, but distributed ledger technology.


The internet of things (IoT)

The internet of things refers to devices and sensors connected to one another that automatically collect and exchange data. IoT is what powers fitness trackers and smartwatches that monitor physical performance. It’s what car insurers use to offer cheaper insurance by installing a device in vehicles that tracks mileage and driving habits.

IoT is now also being incorporated in supply chains all over the world: Switzerland-based Arviem, for one, installs smart IoT sensors on containers and cargo to track anything from the location of the goods, to vibration and container openings, as well as conditions such as humidity and temperature. Combined with advanced analytics and a range of other data sources (such as weather data), this helps business professionals analyse and optimise their physical supply chains.

IoT is a proposition that will appeal to trade financiers: it could enable them to obtain real-time information on the physical flows that they are financing, and use this information to better assess working capital funding risks, rather than solely relying on balance sheet strength. This will ultimately benefit those companies who need their support but are struggling to obtain financing today.

Arviem is currently exploring this use case, developing a supply chain finance solution where financiers and investors can bid on financing for shipments based on their risk profile, calculated using real-time sensor data, historical data, as well as the value and type of cargo. The use of IoT data will in essence “make assets in transit financeable, which is not doable today”, says Stefan Reidy, the co-founder and CEO of Arviem.


Smart contracts

A smart contract is a piece of software that is programmed to automatically complete transactions once certain conditions are met. For example, a smart contract could automatically ensure that a supplier is paid once the goods have arrived at their final destination.

In a first-of-its kind trade transaction in 2016, Commonwealth Bank of Australia, Wells Fargo and Brighann Cotton combined smart contracts with blockchain and the internet of things. It involved the tracking of a shipment from Texas in the US to Qingdao in China, and when the goods arrived, the smart contract automatically triggered the release of funds.

According to a recent report by law firm Allen & Overy, there are a range of use cases for smart contracts in the finance context, such as proxy voting, the settlement of securities, payments under a derivatives contract and the recording of financial data. However, it notes, “smart contract technology is still in its infancy and is unlikely to be deployed for complex and elaborate agreements yet”. In the immediate future, the technology will most likely be used for automating simple and standardised tasks, such as straightforward payment flows, identity verification and index pricing. As such, smart contracts can help to automate transactions that used to require manual paperwork.

A smart contract doesn’t necessarily have to be a legal contract, writes Allen & Overy. “It depends on the relevant applicable law to determine whether a contract has been made, just as with any other type of alleged contract.”

R3’s DiCaprio explains that “not all smart contracts are created equal”. So while contracts on Corda typically link legal prose documents with smart contract code, smart contracts on Ethereum are “simply defined as self-executing code”.

“Only smart contracts accompanied by legal prose would be legally binding,” she says.

While smart contracts have gained renewed interest since blockchain technology has come to life, it was actually a concept invented back in the 1990s, and doesn’t need blockchain to work.

DiCaprio explains: “Smart contracts existed prior to blockchain. All that is needed is a system that enables automatable and enforceable agreements. Historically, this required a single operator overseeing closed databases. The reason smart contracts are so commonly associated with blockchain is because blockchain allows smart contracts to exist between distrusting parties without the need for a single operator.”


Optical character recognition (OCR)

OCR is a technology that converts different types of documents, such as scanned paper documents, PDF files and images, into a digital format.

Trade finance is a particularly good use case for OCR, given its traditionally paper-heavy nature. Banks are increasingly using OCR technology to digitise and interpret a wide range of trade documents, such as letters of credit and bills of lading. Many banks see this technology as the necessary first step to wider digitisation within the industry.

OCR is a field of computer vision, the science of giving machines the ability to visually interpret images, which also includes biometrics and face recognition technology. It is commonly used together with machine learning and other artificial intelligence to analyse the document’s data to spot trends and assess risk. Traydstream, for one, launched a solution to automate regulatory compliance screening last year. To do so, it developed its own OCR engine, which can read, scan and instantly structure paper-based information digitally.

It then processes the documents and rigorously checks them against a library of tens of thousands of global and regional trade finance regulations and rules. Using machine learning technology, the system only improves as it is fed more data.



A cryptocurrency is a digital or virtual asset that uses cryptography (also referred to as encryption, namely the process of encoding information so that only authorised parties see it) to secure financial transactions and regulate the generation of units. Bitcoin, which was launched in 2009, was the first cryptocurrency to capture the public imagination. Powered by blockchain technology – which serves as a public financial transaction database – it is completely decentralised, meaning there is
no central controlling authority.

Over the years, bitcoin has become a contentious issue, with critics pointing to the fact that its popularity hinges on its anonymity and thus the ability to use it for illicit activity.

But cryptocurrencies have been used for much more than that. In fact, it’s making a real impact on trading companies around the world, helping them to lower transaction costs and speed up settlement times.

BitPesa, for example, uses bitcoin to help African businesses make or receive international payments. In Africa, using conventional methods to pay a supplier abroad or receiving a payment from a regional customer is expensive – the average cost is 12%, says Elizabeth Rossiello, BitPesa’s founder. It’s also a highly manual and inconvenient process, and can take weeks.

So how does BitPesa work? Using the example of a Nigerian customer wanting to transfer money to a Chinese suppler: through the BitPesa app, the Nigerian firm makes a transfer to BitPesa – in the local currency and through its local bank. BitPesa then buys bitcoin and sells it on to a Chinese broker, a licensed partner, who makes sure the money goes into the recipient’s local bank account in China. In this case, the cryptocurrency is used as a bridge currency between fiat currencies – the customer doesn’t touch the cryptocurrency themselves, and the instant settlement removes the risk associated with cryptocurrency’s price volatility.

It’s a method that’s been adopted by other players. IBM’s new Word Wire solution, for one, enables real-time cross-border payments using lumens (XLM) as a bridge currency. Lumens is a cryptocurrency built on the Stellar network, an open platform for building financial products. Ripple’s xRapid product, on the other hand, uses XRP (a digital asset created by Ripple) to offer an on-demand option to source liquidity for cross-border payments.

Meanwhile, other firms have sought to create their own cryptocurrency on top of existing blockchains, which they call tokens. Tokens are a representation of a particular asset or utility and are typically created and distributed to the public through an initial coin offering (ICO), a means of crowdfunding, to fund project development.

Populous, is one example of a fintech startup using tokens to power its invoice finance platform. Built on Ethereum, the platform takes advantage of smart contracts and a decentralised network, but none of its operations use or rely on cryptocurrencies like bitcoin or Ethereum’s cryptocurrency, ether (ETH). Instead, the flow of funds within the Populous platform is made possible through the use of custom tokens, called Pokens.

When an investor makes a bid on an invoice, he places fiat money into the platform, which is then exchanged to Pokens. These tokens are then held in a smart contract until the auction has concluded, after which the contract executes itself, and the invoice seller automatically receives his tokens. Again, these can be withdrawn in any fiat currency. According to Populous’ founder and CEO Stephen Williams, this is a more straightforward, quicker and cheaper process than conventional methods, especially for cross-border activities.

Populous’ token is a stable token – often referred to as a stablecoin – which in simple terms means it is price stable. It’s seen as a way to prevent the volatility that often fuels the use of a cryptocurrency as a speculative asset.

“A stablecoin is a digital coin which is pegged to a fiat currency, for example the US dollar,” explains Bear at IBM. “In order to make that work, you have to have it backed by a real dollar, so it becomes a shadow representation of a real dollar that is sitting physically in a bank somewhere.”


KYC utility

A KYC utility works as a shared service or repository in which multiple institutions can manage and share due diligence information. The aim is to ease the know your customer (KYC) burden on banks by eliminating the high level of duplication currently happening between and within institutions. The model saves each bank using the service from having to reach out to its clients to gather the information individually, while customers who have already provided their information to a participating member won’t have to supply it again.

Many market players have labelled blockchain as the ideal technology to power a KYC utility. The peer-to-peer nature of blockchain technology could give customers more control over their data, while the elimination of third-party data aggregators and centralised repositories could drive greater operational efficiency (read more in this magazine’s fintech feature).



An acronym for ‘application programming interfaces’, APIs are the technology that enables different platforms, apps and systems to easily connect and share data with each other. It is, for example, the technology behind the familiar “Sign in with Facebook” button you see popping up on websites other than Facebook.

APIs present a world of possibilities to innovate financial services, particularly now that legislative changes (the EU’s PSD2 and the UK’s open banking regulation) are forcing banks to make open APIs available. This will give third parties the ability to easily access customer data (with the customer’s permission, of course), draw insights and use the data to create innovative products tailored to customer needs.

For example, APIs now allow consumers and businesses to initiate and track payments on third-party applications by connecting directly to their bank. In trade finance, too, the technology could bring new opportunities. With the sharing of data via an API, applying for a loan with a non-bank lender, for instance, could become easier, cheaper and more secure. Through an API, the customer could easily enable a third-party lender to access past and current cash flow data, allowing the lender to make an accurate and ongoing assessment of a company’s risk and creditworthiness. “If you become less risky, then you’d expect the rate you pay to reflect that, because it should be real-time and dynamic,” explains Louise Beaumont, co-chair of the open bank working group at techUK.


Machine learning

Machine learning is the science of making computer systems learn from data and improve this learning over time in an autonomous way. A branch of artificial intelligence, machine learning platforms allow for large data sets to be analysed to reveal patterns, trends and associations with minimal human intervention.

In trade finance, machine learning is typically used in the field of compliance, to detect fraud and financial crime. Fortytwo Data’s machine learning-enabled platform, for example, helps banks monitor and screen transactions and other data for unusual activity. The fintech firm claims that this new technology could eradicate 55% of the millions of false leads – red flags that turn out to be innocent – that legacy systems generate every year.

This type of platform typically utilises two forms of machine learning: supervised and unsupervised. Supervised machine learning means the machine is told what the correct output historically looks like in different scenarios – like a student learning from the results of previous exams. The machine’s task is then to spot complex trends and patterns in the data that humans are typically unable to detect, and use that learning to solve new issues. In the case of Fortytwo Data’s platform, for example, the machine learns from a feedback loop with a human user. When an employee receives an alert for a transaction, they can tell the machine why the activity is not suspicious – information that the platform will later use in its identification process.

Unsupervised machine learning, on the other hand, can detect abnormal behaviour that may not be caught by knowledge-based rules or human review, as well as new types of unusual behaviour. “This is important in anti-money laundering where new money laundering techniques are invented, and organisations need to detect and respond extremely quickly,” explains Luca Primerano, head of strategy at Fortytwo Data

A lot of fintech startups are also exploring how to use machine learning to carry out more accurate credit scoring and thus provide cheaper financing. This approach could be revolutionary for SMEs, which often find it hard to access trade finance due to banks’ inability to accurately assess their credit risk.

One such example is MatchPlace, which launched a peer-to-peer invoice financing service in the UK earlier this year. The firm’s CEO Benjamin Gedeon explains: “We analyse different sets of characteristics of an invoice, such as the company who is submitting it, the type of good/service, the duration, the amount, and of course, the third party on the invoice, the company that will pay for it.” Over time, he adds, the algorithm will “learn” in such way that it will reward good past performance with cheaper financing.

This method is radically different from the traditional bank credit scoring approach, writes Michael Boguslavsky, head of AI at Tradeteq, a trade finance distribution platform, in a recently published whitepaper: “Traditional credit scoring requires a small number of accounting entries and ignores most information available in typical company accounts, as well as any non-accounting information that is available for a company.”

Instead, he argues, by combining machine learning techniques with a broad set of available and emerging data sources (such as geographical and trade network data), it is possible to exploit categories of company information previously uncaptured, and ultimately improve access to trade finance.


Cognitive computing

Cognitive computing is a technology that simulates human thought processes. It’s a term that is used by IBM in particular and brings together various AI elements such as machine learning, natural language processing and computer vision.

The aim is to make computers mimic the way the human brain works and make high-level decisions in complex situations. IBM’s Watson is one such example. The tech giant first trained Watson to operate in the healthcare space, but is now introducing it to other domains, including financial services. Since 2016, after IBM acquired compliance consulting firm Promontory Financial Group, it has been teaching Watson to crunch complex legislation. It means Watson can now read through regulations and other regulatory content (commentary, guidance, speeches and legal cases), pre-process those regulations and identify the potential obligations for financial services firms.



While regtech isn’t a technology as such, the term refers to a range of innovations that can help banks and firms resolve regulatory issues and manage compliance risk. An abbreviation for ‘regulatory technology’, it has drawn particular interest from banks as they continue to try and cope with the tsunami of regulation that has hit them since the financial crash of 2008.

AI and machine learning, for example, can be used to identify and flag suspicious activity, while natural language processing tools and KYC utilities can help undertake due diligence.

Regtech is particularly relevant for trade financiers, and it could well help improve their ability to offer finance. According to over 90% of banks surveyed for the ICC Banking Commission’s latest annual review, compliance and regulation are the biggest barriers for access to trade finance.



Like regtech, insurtech is an umbrella term that covers a range of emerging technologies specifically designed to make insurance more efficient. The broader insurance industry is already seeing incredible moves towards more technology-driven policies, with AI and machine learning
helping insurance companies to build more finely delineated groupings of risk based on data, thus allowing products to be priced
more competitively.

Insurtech is also growing as a concept in the trade finance space. Euler Hermes, for one, is currently testing a range of new technologies through its Euler Hermes Digital Agency, a project established in 2015. It led to the launch of Credable, its first stand-alone brand, earlier this year. The platform allows SMEs to easily check their customers’ creditworthiness, and then offers on-demand single invoice cover based on that score – something that was not previously available to them.

A range of blockchain initiatives have also emerged in the insurance space. For example, since 2016, the B3i, a joint effort between 13 insurance and reinsurance companies, has been exploring how blockchain technology and smart contracts can help make data exchange between insurance and re-insurance companies more efficient. The aim is to improve the way data, claims, capital and payments are disclosed, automated and managed.