Artificial intelligence in general, and contract analytics in particular, continue to dominate headlines. There’s even an app store for legal tech products now.
But what makes a tech platform a legal tech platform? It’s not easy to draw a hard line, and for good reason. Law firms and corporate legal departments have achieved impressive results by using the ContraxSuite platform and the LexNLP natural language toolkit. Organizations in other sectors, though, have also put these tools and others like them to effective use in document analysis.
There are plenty of contracts outside the legal domain. Organizations in finance and real estate deal with huge quantities of contracts every day. An organization in one of these industries, or any other business of sufficient size, will have a lot of contract paper. Looking through documents for important data – like LIBOR fallback clauses – can be made easier by utilizing software tools over other methods like time-consuming keyword search.
Anyone Can Analyze Documents With ContraxSuite
An element of ContraxSuite that sets it apart from similar software platforms is the LexNLP toolkit. LexNLP operates within ContraxSuite, but can also be used as a separate software library. In the past, it’s even been integrated into an organization’s pre-existing contract, matter, and e-billing management solutions in order to enhance outcomes.
An advantage of ContraxSuite’s document analysis interface is that it is trained on documents found in the SEC’s EDGAR database. EDGAR contains records of the majority of borrowing by SEC-registered companies, and many of these are commercial loan agreements tied to LIBOR. And as we’ve seen before, the LIBOR issue is both complex and critical at this moment in time.
There is another reason that commercial loan agreements make a great ContraxSuite use case. LexNLP is designed to retrieve and extract real, unstructured legal text. But legal professionals are not the only people who handle a lot of legal text everyday. Legal text is everywhere; the specific grammar and vocabulary of legal contracts is the life blood of every sector of the economy. Commercial loan agreements are a ubiquitous type of contract perfect for deeper analysis. There are multiple questions to ask. Chief among them: How do you extract data from contracts without spending a lot of time, money, and human effort on line-by-line reading?
Case Study: The We Company
WeWork is a company that leases office space to other businesses, and has grown substantially since its founding in 2010. Last month, WeWork made headlines (and changed its official name to “The We Company”) after filing their S-1 with the SEC. The filing of an S-1 is a necessary step in a company launching an IPO. As a result of this filing, We’s inner financial workings became public, though only to those willing to read hundreds of pages and spot the details. It took several days for the salient elements of the S-1 to be fully documented by the news media and other industry professionals.
Along with their S-1, We Co. also filed a Credit Facilities Commitment Letter. This agreement contains the particulars of the borrowing arrangement between We Co. and its lenders.
The We Company is not a law firm. They are not a large financial institution. Yet these documents, and many other agreements We Co. has with individual tenants, can all be analyzed by ContraxSuite for all sorts of red flags.
Finding LIBOR Dates
An important – and easy – first step when analyzing any large contract is to look for any dates, and the context in which those dates appear. We Co.’s Credit Facilities agreement has several dates, all of which can provide important insight into how this agreement works. LexNLP can scan any document in seconds, and provide a list of unique dates – and the frequency of each date – within a document:
Three of these dates raise red flags, as they occur after the slated end of LIBOR in 2021. Any money loaned to We Co. under this agreement may not have a stable reference rate. What rate will govern any attempted loans?
Upon further inspection with LexNLP, December 31st, 2022 is a maturity date for a Delayed Draw Term Facility, with an amortization percentage. This maturity date is not directly tied to LIBOR, so that’s a bit of good news:
However, the other mention of December 31st, 2022 opens the door to loans that may be based on LIBOR, long after LIBOR is no longer available:
There are potentially billions of dollars on the line for this contract’s lenders. Will they want to loan that money in 2022 without a stable reference rate? We need more information about LIBOR’s impact on this agreement.
Finding LIBOR Fallbacks
Checking the dates in this document was just a quick preliminary step. To seriously assess LIBOR exposure, we have to look beyond the presence of certain dates, and look for LIBOR-related content within the rest of the agreement. Maybe this contract uses multiple different reference rates. Maybe the parties here devised some other schema. Turns out, this Credit Facilities agreement does in fact hinge on LIBOR.
We don’t have to comb through the whole document to discover the LIBOR connection. As with the quick date search above, we can use LexNLP to find all percentages in an agreement in just a few seconds.
From here, we can find and analyze the small handful of clauses that discuss interest rates. For example:
The “rate otherwise applicable” is ambiguous here. We have a clue in the reference to “ABR”, an alternate base rate. Elsewhere in the document, “LIBOR” is not mentioned in relation to ABR. However, “Eurodollar Rate” is, which means there is inherent LIBOR-related risk:
Where is the ceiling? How high can the interest rate on a loan rise under this agreement? The text of this passage tells us to look to the “Interpolated Rate”. The definition of “Interpolated Rate”, however, does not provide easy answers to our questions:
Further clarification is not forthcoming. These are the only two instances of the phrase “Interpolated rate” in the entire agreement. What we see after we follow the breadcrumbs, is that there is no defined ceiling to how high the interest rate term structure can climb, and that the Administrative Agent in this agreement is charged with setting the rate. This wouldn’t be so bad, except every interest rate in this agreement is defined in some way by LIBOR, with no alternative or “fallback” provision able to pick up the slack when LIBOR is discontinued.
The discontinuation of LIBOR poses a risk to the parties to this Credit Facilities agreement, not only because of the potentially volatile interest rate term structure, but also because of the ambiguous fallback language. We Co. may not be a law firm, but it is a global real estate company with billions on its balance sheet. Without clearer fallback language, We Co. and its lenders could be exposing themselves to tens of millions of dollars of risk.
This case study demonstrates one of the biggest draws of the LexNLP library. Without having to read through the entirety of We Co.’s Credit Facilities agreement, in just a few minutes a user can identify problem areas in any contract. The clauses mentioned in this article were found in less than 10 minutes with LexNLP and ContraxSuite. And this example only scratches the surface of what LexNLP can do.
Elevate is a global law company, providing consulting, technology and services to law departments and law firms. The company’s team of lawyers, engineers, consultants and business experts extend and enable the resources and capabilities of customers worldwide. In November 2018, Elevate acquired LexPredict, the legal AI technology consultancy that created ContraxSuite. Elevate has been ranked as a top global law company by Chambers & Partners for the past five years in a row. It has also been ranked among the Inc. 5000 Fastest Growing Private Companies. More at elevateservices.com