ICC Global Survey on Trade Finance 2016

The ICC just released an excellent report on trade finance. 

While the title suggests the report is about 2016, it's actually on 2015.  The report is based on a survey, input from a range of third party public and private sector entities, and SWIFT-provided data.

The report is a comprehensive analysis of trends in the global economy, international and regional trade, various trade finance instruments (including forfaiting) supported by some 95 tables.  SWIFT-provided data on letter of credit and collections messages sent, including data on average L/C size (including by regions with separate data for import and export L/Cs), proposed and rejected transactions by region and country (e.g., Russia). 

There is literally something for everyone with an interest in global trade.

Well worth a close read.

SWIFT RMB Tracker August 2016 Issue - Further Thoughts on the Increase in RMB Usage in South Africa

SWIFT issues an excellent free monthly publication on use of the RMB in payments outside the PRC:  RMB Tracker.   As part of tracking RMB usage, RMB Tracker provides a monthly value-based comparison of the top twenty currencies so even if you're not interested in RMB usage you can track the fate of other currencies. 

You can either sign up for a monthly email or browse issues at SWIFT’s website.

The report is based on customer and bank payment orders (SWIFT MT Series 100 and 200 instructions).  SWIFT, of course, isn’t the only messaging system that financial institutions use.  However, it is widely used—probably the most widely used method for sending payment orders.  Thus, the report should give a very good indication—though more directional than precisely locational—about changes in the use of the RMB.  

The August issue focused on the Republic of South Africa (RSA).  Immediately below are extracts from the August RMB Tracker.  I've used boldface to highlight key points as well as the comments I will focus on. 

"Brussels, 24 August 2016 – SWIFT data shows that Renminbi (RMB) usage in payments in South Africa increased by 65% over the last 12 months and by 112% in the last two years, moving the country from position #30 in July 2014, to #24 in July 2016. Excluding domestic traffic, RMB payment messages increased in volume by 70% in the last 12 months. In addition, nearly 40% of RMB payments by South African institutions have been offshore payments exchanged with countries other than China and Hong Kong, compared to 16% in July 2015."

 “South Africa has experienced a major shift in RMB growth over the last two years, strengthening the country’s trade relations with China and Hong Kong,” says Harry Newman, Head of Banking, SWIFT. “The establishment of an RMB clearing centre in South Africa in July 2015, as well as Singapore’s increased use of the RMB for payments with South Africa, have been a catalyst for RMB growth in the region.”

A couple of things caught my eye.

But before I begin a technical note:  SWIFT’s analysis on the RSA is based on the change in the number (not value) of RMB payments to/from the RSA.  (Page 2).   The number of transactions is one way to measure adoption of the RMB for transactions.  However, I’d prefer analysis based on value, assuming both weren’t available.   If the volume increase is being driven by a large number of small value payments that would put a different spin on the increase figures than if the increase in number were closely tracking value.  

Why doesn’t SWIFT provide this?  They have the data.  In addition to the (free) RMB Tracker, SWIFT have a paid subscription sister publication and  presumably aren’t going to cannibalize its appeal by providing more information in Tracker.

As you read what follows, bear in mind that the lack of information on aggregate payment value constrains the analysis that follows.

Now to my comments.

As written, the report could be read to imply that South Africa has dramatically increased its RMB denominated transactions.   This is not strictly speaking the case.  Since the PBOC’s appointment of Bank of China’s Jo’burg branch (hereafter BOCJ) as the RMB clearing bank in RSA, BOCJ has marketed RMB correspondent banking services elsewhere in Africa.  The increase in volume reflects transactions from other African countries, not just the RSA.

We can find details of this activity in a February 2016 article by Yu Meng in The Peoples' Daily Online and which the South African China Economic and Trade Association relayed on its website.  I’ve highlighted the relevant portions in red and added some comments of my own in boldface blue.

"According to Bank of China, in the "first ten months of 2015, the Renminbi trade settlement and investment amount in and out of Africa (AA: that is, not just RSA) has grown 35% to 126.6 bn yuan (R254 bn).  Half the amount is facilitated by Bank of China’s Johannesburg branch (AA: Here we have a USD equivalent number.  Note BOCJ handles one-half of the aggregate value amount for Africa not just RSA) , which was authorized in July 2015 by China’s central bank, the People’s Bank of China, to serve as the clearing bank for Renminbi business, the first in Africa."

"At the moment, the Johannesburg Branch’s Renminbi service has reached more than 30 African countries, and opened 69 Renminbi clearing accounts for financial institutions in Africa."

(AA:  While we don’t have the volume of these non-RSA transactions nor the amounts, they could well account for a significant portion of the RMB payments attributed to RSA by SWIFT).

"Li, executive vice president of the Johannesburg branch, oversaw a 2.7 bn yuan (R6.08 bn) transaction on Nov 30th for a Mauritian client — the bank’s biggest single clearing deal in Africa. “We wouldn’t have even imagined a transaction in Renminbi this big last year and because of the PBoC’s authorization we expect to do more,” says Li."

(AA:  The wording is ambiguous.  It’s not clear if this was an outgoing or incoming payment.  In either case, it seems a rather large amount for single commercial transaction especially given the amount and nature of bilateral trade between the two countries. Mauritius annually imports roughly US$1 billion equivalent from the PRC – making this single payment over 40% of annual imports. 

The PRC is not among Mauritius five largest export markets.  RSA is in fifth place and annual exports are roughly US$200 million equivalent so it’s unlikely this was an incoming payment for Mauritian goods.   

Given that and the fact that Mauritius is an offshore financial center,  AA suspects that this transaction was financial not commercial.  This was a relatively large RMB transaction—apparently the largest or one of the largest in Africa.  If it is an outgoing payment, its origin may not be Mauritius.  If it is incoming ,its final destination may not be Mauritius.  

Given reports of Chinese parties engaging in “gray” cross-border transactions (AA’s euphemism of the post), AA imagines that a lot of parties might have an interest in finding out more details, particularly if this were a repeating transaction.

Second, Singapore’s increase in RMB payments to the RSA is also interesting. 

If they are trade related, that would imply that the RSA has either redenominated invoicing on existing trade or has "new" trade conducted in RMB.  If so, I’d expect that that pattern would be evident in RSA trade with other countries.  That’s based on the assumption that if the RSA is moving to a greater use of the RMB in its trade, it would be doing so with more than one country. 

That doesn’t have to be the case of course.  Perhaps, there was a unique opportunity to increase sales with Singaporean entities.  As well, the information SWIFT has provided on the increase is based on number of payments not value. 

Assuming that value is tracking to some extent the number of payments (again this need not be the case), as above, I suspect that transactions are financial not commercial.  Perhaps, BOCJ is obtaining RMB for its operations from Singapore via for example FX transactions. While ICBC Singapore is the PBOC-designated RMB clearing bank (in Singapore) and not BOC Singapore, Singapore has some 110 or so banks.  Besides local banks that are active in the RMB, the country has a variety of foreign banks including banks from the PRC (BOC has a full branch) and from Hong Kong SAR who likely have RMB-denominated business and could serve as counterparties.

CIPS (China International Payment System) Hysteria Largely Dispelled?

AA Reporting from His Secure Location: Sometimes You Need More Than Just the Hat

When the PRC announced its intent to create CIPS--the China International Payment System, now renamed the Cross- Border Interbank Payment System--some trumpeted it as a replacement for SWIFT and another, perhaps, the fatal step in global de-dollarization, heralding the end of the primacy of the US dollar.  Here’s one particular brilliant “insight” cast in terms of “white and black hats” and “blessings”. 

Most of this commentary was ideologically-driven:  a preconceived notion in desperate search for  validation. When your ideology gives all the answers, analysis is not only very easy but very facile. You just fit the facts or what might pass as “facts” around the preordained answer.

My favorite exemplar of this way of “thinking” is the assertion that the ruble is fully supported by gold and thus in some way superior to the US’s “fiat” money.  That’s “supported” like Golden Belt Sukuk was “asset backed” –“way back” it turned out.  No doubt AA is missing the opportunity of a lifetime by not converting his deposits to rubles and depositing them in a Russian bank. What could possibly go wrong?  But then I resisted buying Golden Belt Sukuk so I have a track record for missing out.  

Some of it was based on the latent hysteria that often informs discussions of politics and economics, particularly when certain actors are involved.  Here there was a veritable trifecta:  possibly including China, Russia, and Iran – an “axis of evil” for some.  But as the video above shows for others an “axis of virtue”.  Count your blessings!   And note the white cowboy hat on the wall in the video!

At the time there were more balanced analyses like this one from the FT which cast CIPS  as an attempt to simplify the process of making cross-border RMB payments to promote internationalization of the RMB and perhaps lessen exposure to alleged spying and the threat of denial of SWIFT services.  As it turns out these latter goals appear not to have been met as will be outlined in a following post.  CIPS may well facilitate greater “regulatory supervision” (but not “spying” for sure) of offshore RMB-denominated transactions by PRC authorities.

SWIFT’s 25 March 2016 press release by and large put an end to nonsense about CIPS as a global replacement for SWIFT or should have. 

“The MOU sets out plans for a strategic collaboration to develop China’s Cross-border Interbank Payment System (CIPS) using SWIFT as the secure, efficient and reliable channel to connect CIPS with SWIFT’s global user community.”

But there are still holdouts.

Ideology is a powerful thing in the face of mere facts. Here’s one from Jim Willie over at the aptly named website “Before It’s News” carrying a dateline four days after SWIFT’s press release.  I guess the site name means the articles are written before looking at the news.

What’s clear or should be now is that CIPS is not a replacement for SWIFT.  It’s analogous to CHIPS – a New York-based bank-owned utility that makes the bulk of cross-border payments in the US dollar. That is, CIPS is a payment utility like CHIPS or CHAPS, not a secure messaging system like SWIFT. 

Why create CIPS? 

The PRC is looking to promote the internationalization of the RMB.  CIPS is designed to streamline the currently cumbersome process of making cross border RMB payments.  It is also a way to bring this payment “traffic” within the PRC.  As a result of bringing that traffic onshore, PRC authorities will have more visibility into offshore RMB payments than they previously had.  

But could we have figured out that CIPS was highly unlikely to replace SWIFT when CIPS was first mooted?  Yes!

Because there’s a lot of apparent confusion about precisely what SWIFT does, let’s start with some basic facts as they’re part of the foundation of the argument to follow.

SWIFT not hold any accounts, nor does it execute any payments.  It transmits payment instructions to financial institutions that then “make” payments either (a) through utilities (like CHIPS or CHAPS) or national payment systems (like FEDWIRE, CNAPS2—China’s domestic RMB payment system) or  (b) on their own books (book transfer) when they hold the account of the intermediary or final beneficiary. 

That is, assume Bank A receives a SWIFT message from Bank B ordering Bank A to debit Bank B’s account and pay Bank C for account of Person or Company K. 

1. If Bank A holds the account of Bank C, it will make payment on its own books (an internal payment or book transfer) unless Bank B instructs it otherwise. As the name implies, the payment is made within Bank A.  Bank A’s assets and liabilities don’t decline when the transfer is made.  All that happens is that the amount of the payment is moved from Bank B’s to Bank C’s account –an internal shift within liabilities.   
2. If Bank A doesn’t hold Bank C’s account, it will pay the correspondent Bank that holds Bank C’s account either through the utility or national payment system (in banker speak a “wire transfer”).  In this case Bank A’s assets and liabilities will decline by the amount of the payment and Bank C’s correspondent bank’s assets and liabilities will go up by the amount of the payment.  Here there’s an actual movement of funds out of Bank A to Bank C’s correspondent.

While SWIFT is probably the most convenient and most used method to send and receive payment instructions, it is not the only way.  If SWIFT denies a financial institution access to its system, that financial institution can use a variety of other methods (proprietary PC based systems that mimic SWIFT, or other methods like telex, facsimile, email, letters) to send payment instructions to its correspondent.  These will often be more cumbersome and costly and less secure, but if the correspondent is willing to accept them, the payment can be made.

What’s involved in creating a global replacement for SWIFT?  

First a side digression to place this question in context. Creating a bi-lateral alternative or regional alternative would be much easier.  Most internationally active banks offer their customers PC-based systems for transmitting payment orders and receiving account information though such services are limited to two-way communication between the specific customer and a single bank.  Banks in Russia and China offer these services and could easily set up a communications system. If there was a preference to avoid the internet, a cable could be laid.  But that would not be a global replacement.

Putting aside the not inconsiderable cost of creating a new SWIFT, one would have to build a system offering similar services at the same or lower costs, persuade existing SWIFT clients that there was a good reason to shift from SWIFT, and at least for now convince them to shift a large portion of their transactions to the RMB.  That of course would require that the banks’ customers shifted their transactions to the RMB away from the dollar or other currencies.  All this seems to AA to be well beyond a hard slog.

SWIFT works.  That’s why it is the global communication utility. It is as embedded in the payments world as Microsoft software is in the PC world, though I’d argue that SWIFT has the better product. 

Here’s the hill to climb.

1. More than 11,000 financial institutions in more than 200 countries use SWIFT. 
2. SWIFT supports more than 150 payment and settlement systems around the globe.  
3. In 2015 SWIFT handled over 6.1 billion messages related to payments (in a wide variety of currencies not just one), securities transactions, treasury operations (FX dealing, interbank deposits, swaps, loan syndications, gold trading), trade finance (letters of credit, collections).   With this sort of volume fixed cost per transaction is low.  Variable cost is very low.  How does an upstart match this advantage? 
4. SWIFT offers a robust range of products beyond “mere” payments, including financial messaging, bulk file transfers, secure internet browsing/web access for SWIFT members’ clients, a comprehensive suite of compliance and analytical tools --anti-money laundering, sanctions enforcement, etc. 
5. And does so with 99.999% reliability year after year after year.  Its credibility as a reliable partner is proven.

Beyond these obstacles there is another very serious impediment to getting banks to embrace CIPS as an alternative to SWIFT: currency.

CIPS is likely to be limited to a single currency—the RMB.   Today only about 40 percent or so of SWIFT payments are in dollars (Page 5).  What do foreign banks do with the other 60% of their payments?  Run two communication systems?  SWIFT for everything but RMB? And CIPS for RMB?  That’s operationally cumbersome and thus expensive in an environment when cost minimization is key.

On top of that the RMB is less attractive than the dollar and likely to remain so for a long time because of concerns about transparency, business ethics and fair dealing, legal redress, market size (availability of sufficient investable RMB denominated assets), and their liquidity and credit quality to name just a few of the challenges the PRC faces in making the RMB a true alternative to the US dollar. 

Don’t mistake this comment as a Pollyannaish view that the dollar’s place in the world economy is unassailable.  It isn’t.   Sadly, our “own goals” are likely to be the main factor in the dollar’s fate.  Richard Dunne step aside for the real pros!  Any alternative currency will also have to offer the same or greater benefits than the dollar.

Next post will discuss some of the issues arising from how CIPS is likely to operate.