The sequel to the LIBOR scandal threatens to be a chiller for financial institutions sitting on an undead graveyard of unstructured data. But Ann Heron, Chief Strategy Officer at Pendo Systems, is teaming up with the good guys to drive a stake through its heart
So, what was it that so moved the Federal Bank of New York’s General Counsel Michael Held – not an office holder given to using inflammatory language – to describe it as ‘a DEFCON 1 litigation event if ever I’ve seen one’?
It was five letters that spell out a concept so far removed from most people’s sphere of reference that the BBC had to post Janet and John-style explanations of what they meant when LIBOR first penetrated the public consciousness around 2012. (And, even then, it got the acronym wrong).
The LIBOR scandal, as it came to be known, saw a sequence of major international banks, all members of the cosy London Inter-Bank Offer Rate-setting club, hang their heads in shame and pick up hefty fines from regulators for manipulating a benchmark that impacts around $400trillion of financial products – from floating derivatives to your monthly mortgage.
That was the first hint that LIBOR, once described as the world’s most important number, could be on a slow slide to oblivion. And by 2021 it will be well and truly dead, the nail happily banged into its coffin by regulators across the world who have instead been working up alternative, reference rates. In the US, the preference is for the Secured Overnight Financing Rate (SOFR); the UK is leaning towards the Sterling Overnight Index Average (SONIA); while other markets have chosen EONIA for trading in Euros, SARON for the Swiss franc and TONAR for the Japanese yen. (And you thought LIBOR was tricky?).
The problem comes not in setting rates for future products, but in determining what the rate should be for those pre-existing agreements that were pegged to LIBOR in the past. Not only is the adjustment not going to be a like-for-like swap, but financial organisations must first identify from the millions of records they hold, what needs to be updated and how. Legacy is probably an over-used word in financial services now, but the mathematical and administrative headache that it is creating for financial services companies is not one you’d want to inherit.
The whole exercise has raised the terrifying prospect of a ‘Zombie LIBOR’ that persists in undead form in legacy contracts long after the real world has moved on. US advisors Arnold & Porter point out that ‘an estimated $350trillion of currently outstanding LIBOR-linked financial transactions expire after 2021(in comparison, the US national debt is $22trillion)’. So, what can be done about it?
Ann Heron, Chief Strategy Officer at Pendo Systems and herself a former senior officer at the New York Fed, has been close to the issue for some time.
“The industry has never seen a challenge of this scope, scale and complexity,” she says. “A manual approach to navigating the LIBOR documentation challenge simply will not work. Data discovery is the first critical step. The sooner a firm knows its LIBOR exposure, the sooner it can begin the onerous task of remediation and related operational system changes. So, time is of the essence.”
With the clock ticking, Pendo Systems will be revealing its LIBOR Fallback Engine at London’s Sibos – it’s one that uses its advanced AI and machine learning ‘from data discovery to contract remediation through to data curation’.
“This solution is made possible by combining two distinct capabilities,” says Heron. “The first is around unstructured data, the data that doesn’t fit into a table so it’s not machine-readable. LIBOR contracts and financial statements are great examples of unstructured data. If you can’t access and manage your unstructured data, it’s going to be nearly impossible to be successful with the LIBOR documentation challenge.
“The second capability is workflow. Workflow is an intelligent system that is needed to handle the waterfall of logic and business rules associated with LIBOR decommissioning.”
Pendo uses its proprietary software to tackle the discovery phase. Once it’s ingested all of the documents and determined if they’re LIBOR or not LIBOR, it scripts the data extraction rules to meet the client needs, creates a structured data set, which is the output summary of all the results, and provides a real-time audit trail for testing that’s designed to feed workflow tools seamlessly.
“After Pendo has digitised these unstructured documents, we think it’s critical to keep them safe in an automated system,” explains Heron. “We don’t want them to slip back into the ether, back into a manual process. Up to this point we’ve made all of the LIBOR data attractive to a machine readable format, that can then be made ready for a subject matter expert’s review in workflow.”
The whole process has been put together with close regard to what the regulators are looking for – even if, as yet, that’s not 100 per cent defined.
The Alternative Reference Rates Committee (ARRC), a US industry group convened by the Federal Reserve Board and the New York Federal Reserve Bank to plan the market’s transition away from US dollar LIBOR, for instance, wants firms to measure progress in terms of forecasting.
“And it has to be fact-based, so it has to link back to evidence and documents,” says Heron. “In terms of differentiators, the solution needs to work at both the enterprise scale and, ideally, be extendable to the broader industry.
“Regulators will look to see that the solution is also sustainable. It can’t be something that’s just spun off on Excel spreadsheets and emailed across an organisation. This needs to work for the next couple of years and, given the legal litigation tail on this, it has to be an enduring solution, so sustainability matters. It requires very advanced automation, but it needs to allow for just enough human insight to see around corners and, ideally, again given the litigation tail, the data should persist.
“The regulators keep talking about ‘trust but verify’ or ‘how do you know what you know?’. The persistence of the data will allow enterprises to ask these hard questions a few months from now or a few years from now.
“Most importantly, the logic needs to be transparent. Black boxes and intellectual property won’t fly. Auditors and regulators and an organisation’s counterparties will want to understand the decision logic and, ideally, the solution should be ‘green’. By green we mean it’s the opposite of one-and-done. It’s our intent that this solution can check all the enterprise risk management boxes – data discovery, remediation, and curation – at the core. This is the journey.”
Heron doesn’t see the Fallback Engine as just being useful in a crisis, however.
“There’s a lot of waste created if this is viewed as an administrative task just for LIBOR. What if the solution could also lift the efficiency and effectiveness of the underlying credit agreement processes and derivative contracts. That’s why the Pendo solution is a differentiator,” says Heron. “What if you can get more out of this than just looking for LIBOR?”
Leaving traders to set a rate from which they could profit, based on numbers they plucked out of thin air, turned out, as Katie Martin, capital markets editor for the FT, later put it, to be a bad idea. But if Pendo and its partners, in helping the industry row back from DEFCON 1, create a more positive legacy, then maybe some of the pain will have been worth it… and we can keep the zombies at bay.