International payment part 2: pain points and solutions

Ming-Chieh Lee
7 min readNov 25, 2020

--

look into the pain points in international payments and solutions both at application and infrastructure level.

Key takeaways

  1. Pain points in international payments are: 1. cost, 2. speed, 3. no transparency, 4. ease of access. Root cause is lack of international payment rail.
  2. A number of fintech start-ups have sprung up to address these problems, but true disruption would require both at the infrastructure level as well as business level.

Intro

International payment is an area that I am really passionate about. As a practitioner and consultant to international businesses, I experienced the frictions in international payment first hand. As a fintech nerd, I know that international payment is an area for disruption for the following reason:

  1. market size
  2. technology discontinuity
  3. fragmentation: lack of standard
  4. excessive waste or cost / is there an opportunity to create efficiency

In this article, we will start with the fundamentals. Then look at existing current solutions both at the application level and at the infrastructure level. Note, this is part 2 of international payments. Part 1 talks about How International Payment Works: Clearing and Settlement

The fundamentals

Pain points in international payments:

Note some of the pain points I described below have been improved by the banks as we are in a transitional phase.

The pain points in international payments can be summarized into the following 4 points:

  1. expensive
  2. slow
  3. lack of transparency
  4. ease of access

Expensive

For John in US to send money to Bob in China, John would need to use wire transfer from his bank. In US, the bank charges from $15 to $40 for processing / transaction fee. This is on top of the exchange spread that bank makes.

The exchange spread is the money banks make for trading the currency. For example the bank’s spread for $ to ¥ trade may be 6 to 7. Basically banks will use ¥6 to buy $1 from you. And can sell the $1 for ¥7 to someone else. Note the spread should be much smaller, I made it large here so it’s easy to understand. Typical spread is around 2–5% depending on the liquidity of the currency.

In addition, there may be an additional charge at the receiving bank side.

Slow

A typical wire transfer will take 2–3 days. However, since banks are not open on the weekends, and different time zone can have different operating hours, and holidays, there is no guarantee that the money will arrive in 2–3 days. This problem is exacerbated by lack of transparency (next section)

Lack of transparency

Imagine you are sending money to someone that is supposed to arrive in 2–3 days. The money has left your account, and 3 days passed but the recipient still hasn’t received the money. You asked your bank, and they couldn’t tell you exactly where the money is nor exactly when it will arrive. That’s a horrible feeling and not acceptable in today’s digital economy.

Another area that needs better transparency is fees. Most of the bank that we use to send international money will tell you their fees, but will not guarantee that there is no additional charges by other institutions. I think that’s just horrible user experience to have customers go through a service without knowing exactly how much will be charged.

Ease of access

Traditionally, sending international payments from your banks is not the easiest experience. It’s typically not part of the internet banking or mobile app function. So, you had to physically go into a branch and wait in line.

Before we look at the cause of these problems, let’s try to understand the fundamentals of international payments. By understanding the fundamentals, will help us understand better the cause of the problems described above.

How is money moved across border?

Note: I will the discussion on the high level here. For those interested to get a solid understanding of the basics, please refer to my previous article: how foreign exchange works.

The way foreign exchange works is first by netting. Then the actual exchange occurs with Nostro and Vostro account, SWIFT messaging, and either Fedwire or CHIPS as the rail for US. For detail on Fedwire and CHIPS, refer to my article: how payments are cleared and settled.

Netting is the process of consolidating foreign exchange trade. For example, Bank of America (BofA) wants to buy ¥650. And Bank of China (BofC) wants to buy $100. Assuming exchange rate $1 = ¥6.5. In the example above, we have a net of 0, so no actual money is moved across the 2 banks. However, there is money movement within the bank to settle the trade.

BofA will debit $100 from its FX trading account, and credit $100 to BofC’s account at BofA, and BofC will do the same as illustrated in the graph below.

The transaction above is done by Nostro and Vostro accounts, which are accounts that banks have with each other to settle international payments. This is also known as “correspondent banking”.

The cause of the pain (the why)

The reason for the pain points in international payments is lack of international payment rail. There are closed loop rail within each country, such as Fedwire and CHIPS in the US, and there is no standard format / protocol for these national payment rail to communicate with each other.

The team at Ripple explains:

A good way to think about this is email before the advent of the SMTP protocol, which standardized electronic messaging. Imagine being on AOL and not being able to directly email someone using Yahoo. That’s pretty much the situation with international payments. Each region has its own closed-loop system. What’s lacking is a universal translator

Left without a standard solution, banks have come up with their own, called “correspondent banking”. However, these direct relationships between banks internationally is limited due to cost and exposure to risks. And without direct relationship, international payments will take longer, cost more, and lack full transparency.

Source: Glenbrook partners

Current application level solution:

Currently, whether you go through banks, Western Union, or fintech start-ups, the way money is moved across border is The current is:

  1. SWIFT messaging
  2. Nostro and Vostro accounts / correspondent banking,
  3. and wire transfer for domestic money movement.

The problem is that there is no international wire:

To address the problems, a number of fintech start-ups have sprung up (like Transferwise). These companies have done very well and provided the following:

  • Lowered the cost: By netting, pooling, and direct connection to major correspondent banks
  • Provided transparency: Transparency on timing by providing estimates of arrival and show status of where the money is. Also provided transparency on fees
  • Improved delivery time: Speed up timing through direct connection. However, can’t avoid the slowness and limitation of the underlying rail
  • Ease of use: Digital access through computer and mobile. Focused on user experience and product design

The increased competition from fintech has made banks improve their service by lowering fees and enable digital access. While these fintech have done a great job, but the root cause of the problem, lack of international payment rail, has not bee addressed. Hence the following problems still persists:

  1. speed is limited by the limited operating hours of the rail: CHIPS operate from 9 to 5 ET and Fedwire operate from 7:30 to 7:00 ET.
  2. Still too much fragmentation
  3. Still excess costs and wastes without standardized international wire

The fintech start-ups have placed a wrapper around the existing antiquated infrastructure. A wrapper, no matter how beautiful, can only do so much.

Current infrastructure level solution

In today’s digital economy, we should have 24/7 rail that can make international payment in real time. Though we may not actually get real time due to fraud / AML measures, but the capability should be there.

In order for such solution, we need both infrastructure change as well as application level change. On the infrastructure side, there are currently 3 offerings that are promising: Visa’s B2B connect, Ripple, and SWIFT GPI.

SWIFT GPI was rolled out in 2017. The solution promises to process majority of the payments within 24 hours, end-to-end tracking, and fee and FX transparency. SWIFT, the most incumbent out of the 3, built its solution on top of existing infrastructure. Though SWIFT is also looking to further improve its technology with blockchain and has announced partnership with R3, a Ripple rival.

Ripple built an entire new rail with blockchain. Ripple provides real time payment processing.

Visa B2B connect, the last of the 3 to roll out a solution, is also using new rail. Visa announced partnership with IBM, using IBM’s hyperledger fabrics for rail. Visa is also partnering with banks’ core system software provider (FIS) to distribute. I would assume this distribution strategy will keep expand.

Visa B2B connect

Conclusion

It’s hard to say who will be the winner, or whose product is better without looking under the hood. The market is certainly big enough for all 3 companies to do well. With competitions, Ultimately I wish the winner is consumers. Perhaps there ought to be an IBaaS (international banking as a service) company that make it easy to switch between the different rail providers, so that consumers are not locked to a specific platform / solution. This will keep the solution providers competitive and make consumers the true winners.

--

--

Ming-Chieh Lee

passion for #fintech #payments #RTP real time payment #Banking as a Service #digital strategy #blockchain