Many people will be familiar with the famous motto of the boy scouts: “be prepared.” Since the financial crisis this could also be the motto for banks. Post crisis, people are much more sensitive to systemic risks. This includes increased concern about counterparty risk in areas of the system, such as securities servicing, that were more or less taken for granted pre-Lehman.
Having learned a bitter lesson in 2008, both banks and regulators have for example been scrutinizing custody networks. Regulators have been writing new rules to protect client assets held by third parties, and banks have been carrying out far more detailed due diligence on their providers.
These are welcome developments, but do not solve the problem completely. Even the most careful due diligence cannot guarantee that a counterparty will not fail. Nor is total failure the only concern. Today’s financial system is subject to more rapid change than in the past, making flexibility in custodian setups increasingly desirable for other reasons as well. Against this backdrop, the long-dormant discussion around custody-provider contingency planning has been reanimated.
What, me worry?
At the heart of the issue is the fact that custody setups are notoriously complex. Under normal circumstances, it can take up to six months to find and onboard onto a new provider. It takes time to choose potential candidates and then negotiate terms. Complex contracts have to be drawn up and signed: a typical service level agreement (SLA) can run up to 100 pages. Then the setup needs to be implemented, with all the necessary accounts and interfaces.
Feriz Hasani, Head Relationship Management Global Custody FI at UBS, says the bank has seen increased interest among clients in contingency planning, but also a certain level of uncertainty. “The complexity and potential cost involved is forcing clients to consider some fundamental questions,” he says.
One such question is what exactly to worry about. The headline concern is what to do in the case of sudden, unexpected custodian failure – perhaps due to some physical, technical, market or business-related problem that renders the custodian incapable of providing its services. Behind that are concerns about more slow-moving disruptions: a regulatory change, a provider exiting a market. These can generally be handled in a more orderly fashion, but still require the client to find an alternative.
There is also a middle ground between failure and exit during times when providers are experiencing turbulence. A perfectly sound provider might have problems providing full service if volumes spike suddenly for any reason. “At the height of the financial crisis many providers turned out to be under-resourced,” says Hasani. “Today people are looking at ways to judge how their provider can handle extreme conditions. That has become part of the equation.”
As a result, clients need to carefully consider a broad range of risk scenarios and decide how flexible they want their contingency plans to be.
“The complexity and potential cost involved is forcing clients to consider some fundamental questions.”
Act fast, but how?
The next basic question is how to approach contingency. Hasani says there are two main methods. One is to spread the risk by using more than one custody provider. “Some clients have split their networks, and do their custody over two, three or more providers. Not putting all your eggs in one basket is sensible from a risk management perspective, of course, but it is very expensive, requiring clients to set up and maintain parallel networks. Not only will they likely not be able to get the best volume discounts, they also have to devote extra resources internally to oversee the relationships.”
For that reason, banks are increasingly looking at contingency plans involving the ability to switch providers quickly in case of an emergency. “This generally entails the establishment of a shadow provider and can make sense in many ways,” says Hasani, “both in terms of cost as well as flexibility.”
Running hot and cold
Hasani says there are different choices clients can make depending on the contingencies they are planning for. “Within UBS we generally differentiate between ’cold’ (passive) and ’hot’ (active) setups. In a cold setup, the transfer can generally take place within a month or so. In a hot setup, the transfer can take place within 24 hours. This is the best option to provide coverage for any eventuality, but it does cost more.”
This cost difference is related to the effort involved on the provider’s side to keep operationally ready. In a cold setup, the two parties generally come to a preliminary agreement, but no specific actions are taken. No detailed contracts are negotiated or signed, no SLAs are drawn up, no accounts are opened and no technical work is carried out.
In a hot setup, the backup provider is always ready to go. Contracts and SLAs are signed, accounts are opened and the IT is implemented. The idea is to perfectly mirror the setup of the incumbent. The only difference is that no actual securities trades are made. This requires more work on the contingency provider’s side: even dormant infrastructure needs to be managed to be kept operational. If the client makes any changes in the setup at the incumbent, these need to be reflected in the contingent setup as well.
The contingent setup also needs to be constantly tested, a subject of some debate. “There is discussion on the best testing methods,” says Hasani. “In one scenario, the contingent provider carries out regular readiness testing. In another, the client opens a small account at the contingent provider, which runs live. This keeps the machine warm, so to speak.”
One important element of the hot setup not to be overlooked is the establishment of clear communications processes for the switchover. In an emergency, everything must work. There is no time for fine-tuning, for a lack of clarity on processes, or indeed for miscommunication of any kind. In a well-thought-out contingency setup, both parties will know in advance who needs to call whom. To facilitate this, templates for the call and the official notice (generally in writing) may be worked out in advance as well.
Of course, in a real crisis, clients will not just be thinking about their custody needs. “Our clients do cash, FX, equities and derivatives – a whole palette,” Hasani says. “However, cash and custody setups are prerequisites for all the various other products.”
In this context, Hasani points out that contingency planning in the cash world, for example in CLS, is also a very hot topic at the moment. “Service providers often provide large intraday liquidity provisions,” he notes. “What happens if such a provider goes bankrupt? Who is going to absorb the necessary liquidity? On the provider side, discussions have been ongoing about whether national banks should intervene in such situations, or if the remaining providers should agree to proportionally take on the business the bankrupt provider cannot handle. This just shows how interconnected the securities world is.”
Know your customer, well
As a major player in the custody industry, UBS has been thinking of ways to assist its clients considering custody contingency plans. “Providing these kinds of solutions also requires a great deal of thought on the provider side,” says Hasani. “Our biggest challenge has been that we need to understand clients’ needs at a whole new level of detail. To provide robust contingency services, you have to have an in-depth picture of how each client operates.”
As a result, Hasani says UBS favors the hot setup option. “We believe that to be truly prepared, you will need an active shadow provider,” he explains. “For this reason, we are working on developing a full contingency product range covering a wide range of asset classes, and based on a switchover in 24 hours with guaranteed availability.”
With such a contingency plan in place, clients can be confident in a robust custody setup. They will have turned “be prepared” from a motto, to a reality.
Some like it hot
While numerous variants are possible, custody contingency setups can broadly be divided into two categories:
cold (passive) and hot (active).
Ca. 1 month
Within 24 hours
Very flexible; allows for quick reaction in a crisis
Less flexible; cannot react quickly in a crisis