Robo-Rules and First Principles

When my daughter joined her senior school, I was totally bemused. Brought up in the school system of the ‘70's which had infant 1 and 2, Junior 1, 2, 3 and 4, followed by senior 1 to 5 with lower and upper 6th, I had just adapted to the Year 1 – 11 format, but was now faced with words like syntax, poetry and rudiments.

I really struggled to relate these words to a student’s progression through school until someone kindly explained they were all about how one learns a language. The sequence follows thus

• Elements
• Rudiments
• Grammar
• Syntax
• Poetry
• Rhetoric (1&2)

Or

• Basic components
• First principles
• The rules
• The structure
• The addition of intensity and expression
• Effectiveness and persuasion

Once I had this it started to make sense, though as one who was scientifically trained the construction of text is not one of my top strengths. Despite that I can appreciate the complexity and beauty of the written word and the varying strengths of different languages.

Why do I explain this? Well, there is much talk in financial services about the creation of “Robo-Rules” or a set of code(?) that can be read unambiguously by machines and applied to business processes and data to monitor compliance and identify issues. The driving force is the huge effort and expense that is currently wasted by each firm interpreting, or paying lawyers to interpret, each new regulation and change to regulation. Even then there is inconsistency with almost as many interpretations as there are interpreters; what is meant to simplify and protect more often confuses.

The eventual rules are mostly driven by some for of regulation which is bound in the relevant legal system. Drafted by lawyers, debated by experts(?) and enshrined in archaic forms (eg olden terms and limited punctuation), these rules differ structurally from jurisdiction and indeed often differs within the same regulation as compromises are reached. Additionally, the legal code upon which a regulation is built differs from jurisdiction to jurisdiction, eg the Law of Tort on the UK, the Napoleonic Code in France, etc.

I applaud the effort to create a world with the simpler and more consistent application of well-constructed rules, but I don’t underestimate the problem. This is where the scientist in me kicks in. Faced with a problem I go back to first principles, make sure I understand the root causes and true objectives and then construct a robust solution. This is why current efforts create an unease in me.

It feels as if the industry is looking to “robotise” the top-level output; the poetry and syntax of the rules if you like, while glossing over the flawed foundations ie it is working “top down”. It undoubtedly creates a great platform for some and makes the problem seem easier to solve. Unfortunately, experience tells me that it will not be a sustainable solution. It will be limited and will need fundamental reworking at some stage.

Personally, I would prefer to remember by daughter’s schooling and try and build from the base ie “bottom up”. Use the enormous intellectual resource the industry has to develop the right set of language elements to support effective regulation, establish the rudiments of constructing rules and push through the grammar and syntax. 

I realise that this suggestion challenges a large number of vested interests, but if we are to improve matters and rebuild trust in the financial system with better regulation then we need to invest, not just protect. Indeed, maybe we need both approaches with a plan to unify them when they meet?