By Alexander Paetzold, COO, Trade Technologies & MD, Trade Technologies Germany GmbH, and Co-written by William Evans, Trade Advisory Board Member

A company with guarantees at six banks does not really have a guarantee portfolio. It has six of them, in six systems, in six formats, and the task probably nobody is assigned to is tracking that portfolio and associated exposure. Our overview of managing guarantees and SBLCs end to end argues that operational control is where the real risk sits. This piece is about the version of that problem that appears at scale, when instruments are spread across many banks and platforms, and about why going digital does not make it go away.

Why does a portfolio end up spread across so many banks?

Because using several banks is good practice, not bad. Companies place guarantee business across banks for sound reasons. Not every exposure is one a single bank wants to carry whole, and most of the time treasurers also do not want to depend on just one bank. A guarantee needed in a particular country often has to be issued by a local bank present there, which might be the local subsidiary of one of your core banks or a local correspondent bank. Pricing and relationships vary, and it pays to spread the work. So a large portfolio naturally ends up across several issuers. 

That is a sign of a well-run treasury, not a symptom of disorder. The fragmentation that comes with it is the price of doing the sensible thing, which is worth saying plainly, because the answer is not to use fewer banks. It is to manage the portfolio better.

Why doesn't standardized messaging already solve this?

Guarantee messaging is more standardized than people expect. Banks issue and amend through defined SWIFT messages: an MT 760 to issue a guarantee or standby, an MT 767 to amend one, with further messages for acknowledgements, reductions, and even a dedicated non-extension notification. There is also MT 798, a corporate-to-bank trade envelope built for multi-bank operations, which lets a company apply, receive advices, and handle amendments across several banks through a single channel.  

So why is the portfolio still in pieces? All of that standardizes the message and the transaction, not the portfolio. A structured MT 760 describes one instrument, at one bank, at one moment. It does not assemble a current view of everything outstanding across every bank. MT 798 depends on SWIFT connectivity and onboarding that many corporates, mid-sized ones especially, have not put in place, so in practice a great deal of guarantee business still runs through each bank's own portal, by email, and on paper. Standardization at the message layer is real and useful. It is simply a different thing from consolidation at the portfolio layer.

What actually differs from one bank to the next?

Even when the underlying instrument is similar, the handling is not. Each bank has its own portal, with its own login, its own screens, and its own way of showing status. Reference numbering differs. Amendment histories are laid out differently. Reporting formats differ, so an export from one bank does not line up with an export from another without manual work. Notice and communication practices differ too. 

None of this is a fault. Each portal is built to serve that bank's own customers well, and most of them do. But a treasurer holding instruments at six banks is looking at six well-built islands, each complete in itself and none aware of the others.

Doesn't going digital fix the fragmentation?

This is the part that catches people off guard. Digitizing issuance can make the fragmentation harder to see rather than easier. When every bank offers a capable portal, each instrument feels well controlled, because at the single-instrument level it is. The gap was never at the instrument level. It is at the portfolio level, the one view that shows every live guarantee across every bank, with amounts, expiries, review dates, and country exposure together in one place.  

No single bank platform provides that view, because in a bilateral setup no single bank sees the whole book. Syndicated facilities are the qualified exception. Where guarantees are issued through a fronting bank under a syndicated facility, the fronting bank and the facility agent do have transparency, by default, over total utilization across the banks participating in that deal, because the structure requires it: the fronting bank issues on behalf of the syndicate and is counter-indemnified by the participants pro rata. But that visibility ends at the boundary of the facility. A corporate with one syndicated line and several bilateral lines still has no bank that sees all of them together. So a company can digitize every issuance step, feel thoroughly modern doing it, and still be unable to answer, in minutes, what its exposure is in total or to any one counterparty across all its banks. Digital issuance is not a consolidated portfolio. It is faster issuance into the same fragmented picture.

Why does manual reconciliation survive in a digital process?

Reconciliation is the part the portals do not do. Someone still logs into each bank in turn, pulls what that bank shows, and stitches it into a master spreadsheet that is out of date the moment a new amendment lands somewhere else. This is the same drift described in common operational failures, now multiplied by the number of banks. Every additional bank adds another source of record to chase, another format to normalize, and another place where a change can happen that the master list does not capture.  

The manual work does not survive because a team is behind the times. It survives because digitizing issuance left the hardest task untouched, which is holding one current picture across many separate sources.

What does managing at scale actually require?

One consolidated view that sits above the banks rather than inside any one of them. The aim is not to replace bank portals, which do their job well. It is to add the layer they cannot provide: every live instrument, inbound and outbound, across every bank and platform, in a single current record, and queryable by expiry, by country, by currency, and by counterparty. That is the view a treasurer is actually asked to produce, and it is the one a fragmented setup cannot produce on demand.  

A single question tests where a company stands. Ask for total guarantee exposure to one counterparty, across all banks, as of today. If the answer takes a week of logging into portals and reconciling spreadsheets, the problem is not any one bank. It is that the portfolio has never existed in one place, only in the pieces each bank holds. Closing that gap is what managing guarantees at scale really means, and it is why the thing that helps is consolidation, not another portal.