Shorter settlement cycles — Haven’t we done this before?

In the move to T+1, Gerard Smith, head of product, digital assets, Nasdaq, asks what lessons can be learned from China’s existing T+0 settlement environment.
By Chris Lemmon

As the industry prepares for the move to T+1 in the US and Canada, with similar conversations taking place in Europe, it’s easy to lose sight of the fact that the world’s second largest equity market, China, is already at T+0.

China aside, global equity markets typically have a delayed settlement model with a two-day period between trade execution and settlement. This delay allows for several post trade processes to take place, including the matching of allocations – typically between asset manager and brokers – and the instruction of settlement through both buy- and sell-side value chains.

Shortening the settlement delay reduces counterparty risk and therefore should result in lower margin requirements. In that context, efforts to reduce the settlement cycle seem logical. Nevertheless, there remains general disagreement about whether moving to T+1 offers sufficient net benefits, whether it is ambitious enough, or if the industry is sufficiently prepared.

However, largely missing from the debate is how China’s T+0 model is already widely used by asset managers, brokers and custodians the world over, operating a settlement deadline just four hours after the close of trading. It also has the added complication of a less liquid settlement currency, CNH, Renminbi that’s traded offshore from mainland China such as in Hong Kong.

International investors and intermediaries can therefore draw significant lessons from this market and the impending ramifications for operations, risk and market structure.

Navigating shorter settlement times

At present, much of the industry still manages allocations after trade confirmation. The process of matching therefore needs to be compressed, allowing settlement instructions to be sent on T. To deal with this, on Hong Kong’s Stock Connect market the practice of ‘single sided settlement’ has become commonplace. This work-around is an agreement from the asset manager passed down to the local custodian giving authority to match the broker allegement at the CSD, regardless of whether the settlement instructions match. Any issues are resolved after the fact which, while not good practice, is widely adopted.

Another way of mitigating the risk of settlement failure is for asset managers to subscribe to an ‘integrated model’, whereby trade processing stays within the business of a single firm. The broker, clearer and local custodian operate under one roof and therefore have full control over end-to-end trade processing. Participants using this model can leverage banking credit lines to fund their trades at a later point in time and they have no settlement risk between the broker and custodian. However, they lock asset managers into a smaller broker panel – and hence run into frictions with regulatory requirements around best execution.

We also see asset managers opting for segregated accounts at CSD level. These accounts offer better asset protection but add to the administrative burden and complexity of operating in this market, contributing to settlement risk when the communication of SSIs goes awry. Meanwhile, for some firms, the challenge of dealing with T+0 settlement has simply been outsourced to their broker. Instead of trying to complete the end of day allocation process, FX and other post-trade activities in a compressed timeframe, they simply ask their broker to sit on their net purchases for another day, albeit with the accompanying financing costs.

Given the move to shorter cycles is a global trend, now is a good time for market operators to think more holistically about marketplace structures and reinvent post-trade processes, rather than adopt a sticking plaster approach.

Longer term solutions

The introduction by SWIFT of the Unique Transaction Identifier to securities transactions is an industry initiative that should result in a significant reduction in settlement fails. The UTI is attached to each trade and propagated through the value chain between institutions. Use of the UTI aids transparency within the settlement lifecycle and allows participants to track settlements within the process. In turn, this allows participants to identify issues and bottlenecks. Increasing adoption of the UTI by CSDs and other actors throughout the value chain will unlock the end-to-end visibility that is required to manage fails effectively and identify weak links in the system.

Whilst the UTI proposal aims to bring transparency to the existing structures, other initiatives seek to reinvent the way in which the market operates. The use of Distributed Ledger Technology (DLT), for example, has long been spoken about in terms of its profound implications for the structure of the industry.

Some would argue that the pace of adoption of DLT within financial infrastructures is speeding up, but we have yet to see broad adoption of the technology within market infrastructure firms except for a growing number of use cases focussed on securities issuance, but not settlement.

Rather than replacing legacy systems, I believe we are more likely to see more DLT-inspired solutions that will enhance how existing structures operate. This approach will still achieve end-to-end visibility, offering accurate data to all participants in a way that is simultaneous rather than sequential, but maintain the roles of current actors in the system. It will also support the use of smart contracts, which have the power to automate real-time workflows without the need for a fully distributed topology.

One specific area we expect to see develop over the coming months is increasing tokenization of assets, either representing new asset classes such as carbon or well-established securities like government bonds, as market infrastructures seek to lay the initial foundations of an emerging digital asset ecosystem. They will be built on private permissioned networks, on single nodes that can be built on a central database with the option to migrate to distributed topography later, therefore reducing the go-to-market cost.

This approach can be integrated with existing CSD architecture, is more practical, and will support a transition to T+0 over time. But crucially, it is more likely to see emerging technologies adopted at a greater rate than we are currently seeing in the market.