Like many competitive endeavours, pairing top human intuition and judgment with custom technology in finance often exceeds the capabilities of either alone.
Organizations like Citadel, Renaissance and Jane Street have been beating the banks and indexes for years now by merging engineering and finance talent.
They benefit from the essentially unlimited opportunity offered by global financial markets to convert skill into cash flow. Imagine for a moment that there is an appetite for transaction surety and regulatory compliance that could support such a market for legal work.
What would that look like? And what would that market's Jane Street look like?
The Liquid Legal Market
Posit a new legal market. This legal market might look like a futuristic logistics market.
Where Ubers and Waymos service individuals with driverless cars, and Aurora Innovation serves transport, this market has a number of semi-autonomous law firms servicing individuals and small businesses through human-lawyer-mediated online portals.
Consider the tenant in London who participates in this market. He snaps a photo of a lease on his phone and loads it into a portal for some advice.
Behind the scenes, sophisticated bidding models from global law firms go to work pricing advisory risk and negotiation capabilities. Some of these firms use configurable off-the-shelf solutions, and some firms have rolled their own bidding algorithms. These models ingest client needs (e.g. a phone call or a mark up?), biometrics and location data. They check conflicts against both a centralized database or their own private databases. The firms each come up with a price. Some are too slow or drop out. This bidding and ranking takes place in milliseconds, and a firm is selected.
That firm applies a tested, rigorous negotiation playbook through a custom workflow implemented on one of a dozen legal tech platforms. The results are reviewed in minutes by a Welsh solicitor, and the tenant gets a light markup back a few minutes later. He can ask the solicitor follow-up questions for a small additional fee.1
The lease goes back to the landlord who puts those comments onto the same market. The process repeats until it's agreed.
Some problematic early-termination clauses are eliminated, and the lease is signed in a few hours.
These small-stakes matters are the genesis of this new legal market. Most of these future clients are unrepresented today in the same way their predecessors never rode in a black car until Uber put one on their phone.
Liquid Institutional Tier
Institutions take notice.
Individiual "trusted adviser™" relationships largely dictated the allocation of institutional legal spend before the advent of this market, but those institutions need sophisticated legal work on immediate turnaround times.
And no business relationship is impervious to the relentless march of efficiency.
Consider the client that loads a set of ISDA documents into something like a web browser.
After some of that same background bidding, they click a button to select from a few winners, and a man appears in a small box on screen.
He's a lawyer in his early 30s. He speaks with the client's accent. He graduated magna cum laude from Stanford Law School. He's perhaps not familiar with the client but is up-to-date on their firm's history, key personnel, precedents and legal playbook.
The conversation is privileged and confidential.
Software on the lawyer's end has parsed the legal task into standardized data structures evaluated by his firm's proprietary algorithms which feed back analysis to him via a heads-up display on the perimeter of his desktop.
Within 15 minutes, the client and lawyer have hashed out some big-picture issues and the client has a marked up redline with talking points on her desktop. The lawyer's advice reflects the opinion of one of the world's most reputable law firms.
What would the client pay for that?
It must be worth what they're paying now for the same work that takes a day or two, maybe four, to come back. That could easily be $5,000. Does an instant response make the work worth $6,000? $50,000?
What if the client's a hedge fund and one of its traders asks in-house legal for time-sensitive advice on material non-public information?
"Can we trade?"
What an institution would pay for real-time responses to that is an important question for another day.
JSL
Consider the lawyer on the other side of the market.
He works at what could be called the Jane Street of Legal, or JSL. His firm is staffed by computer programmers who have studied hornbooks or legal practice guides and lawyers who grew long beards and started writing Haskell or Rust. Where the ratio may be 50:50 legal to business personnel at big firms today, JSL has 10 programmers to every practicing lawyer.
The firm operates out of Dallas, Hamilton, Miami or George Town, strictly for tax reasons, but it's in-person five or six days a week. JSL often builds and almost never buys, mostly because there's no supplier for the technology it needs.
Its employees are bound to durable non-disclosure and non-solicit provisions with significant deferred compensation, and leaks are unheard of.
JSL dominates the top end of the liquid legal market for the select set of high-stakes legal services it provides. Services like CFIUS defense and antitrust M&A advice. JSL's technology allows it to deliver industry-leading work product fast, and it leans on that capability to win every bid its algorithm submits.
The Quants
JSL programmers have been recruited away from algorithmic trading shops, automated driving startups, hedge funds, banks and high energy physics programs. Many have PhDs. Physics, statistics, mathematics or linguistics are common disciplines.
Why JSL?
Legal work is almost unrivaled in its profitability, and ethics rules around fee sharing and non-lawyer compensation have loosened in response to international competitive pressures.
The United Kingdom and traditional red US states were early movers, but now most important jurisdictions' ethics rules allow it.
Language models owned by JSL and built by the quants have converted the art of natural language processing and legal judgement into a science of vector operation pipelines.
Also JSL invests heavily in R&D, and it pays six and seven-figure performance-based bonuses as a result.
The Normies
But law is a people business.
It's not just about correctness, it's also necessary to sell that correctness to the client, counterparty or judge.
The lawyers at JSL are, for that reason, great rhetoricians. They turned down offers at firms with names like Quinn Emmanuel, Wachtell and Slaughters. They studied hard sciences in undergrad and maybe sat for the patent bar during law school.
But they can talk to people.
Why did they join JSL? Because JSL engineers build the best, most secretive weapons in the industry. JSL lawyers—and only JSL lawyers—get to wield these weapons.
The lawyers, like the engineers, earn hard cash but also equity in certain JSL entities which have active but closed secondary markets. Their wealth becomes an amplified side effect to the feedback loop of wielding these weapons to defend the moat. And their earning potential as a JSL junior associate as a result is more than what some partners earn at their traditional competitors.
If this hypothetical liquid legal market coalesces, history indicates there will only be one real JSL.
So who's it going to be?
[1] Here it is important to note that this structure will not relieve firms of their ethical requirements to be informed and vigorous in their representation of their clients. That ethical obligation, it is expected, will be met through the application of these technical facilities.