Finance 3.0

Law Lab co-director John Clippinger makes a case for Finance 3.0

The current financial crisis has forced a rethinking of the structure and governance of global financial networks: specifically, how to restore trust, transparency, and accountability, while at the same time, not stifling innovation nor genuine value creation. Looking forward, there looms the prospect of the heavy hand of governments intervening in financial markets with the usual “unintended consequences”. Yet absent government regulation, there is the even more daunting prospect of an endless cycle of bubbles and busts, an eventual global depression.

There is a third alternative. Rather than turning to worn out regulatory methods, there may be an opportunity to design into financial networks mechanisms to achieve commonsensical governance principles, such as, auto-auditing, independent rating mechanisms, reputation assessment, authentication of claims, and principled self-governance. This is what we mean by Finance 3.0. It entails experimenting with the design of new kinds of financial networks that combine innovations in regulation and technology to create more transparent, accountable and innovative financial systems.

Components For Self-Governing, Trusted and Transparent Networks

Part of the research at The Law Lab is to explore the feasibility of encoding within financial networks principles and mechanisms that perform some of the same functions that are normally undertaken by regulatory agencies and legal processes. One of our goals is to build an open cloud platform to research and pilot test some of the possible principles and mechanisms needed to create trusted, resilient networks.

The following components are under consideration or development.

Global Identity Systems:

How important is it to authenticate participants and have role based access privileges? Is contextual, authenticated anonymity an important requirement for building trusted networks?

Protected and Selective Sharing of Information:

Can “zero-knowledge proofs”, pseudo-anonymity, role based disclosure and other mechanisms enable sharing of information without compromising confidentiality and privacy?

Robust Reputation System

Can there be sufficiently robust and diverse reputation metrics that are not hackable or gameable? Can such measures be derived reliably from behavioral meta-data and other sources?

Transparency and Accountability

How much information needs to be collected, audited and disclosed to achieve robust transparency and accountability?

Independent Rating

Are there mechanisms to enable independent, competitive and open rating methods that are not easily compromised?

Collusion Detention/Prevention

Can methods be devised to identify or flag incidents of collusion? Are there robust mechanisms to significantly reduce its likelihood or efficacy?

Self-Healing

Can mechanisms developed for Grid and Cloud computing be adapted to identify “trust/risk” failures and automatically generate repairs?

Resiliency

Can networks be designed with sufficient redundancy, distributed risk, and rapid recovery capabilities that they are resistant to catastrophic failures?

Dispute Resolution and Enforcement

Are there private law practices and digital mechanisms to quickly and efficiently resolve a large class of disputes and to effectively enforce them?

A Vision of Digital Governance

One of our research goals is to have an open cloud platform whereby anyone, anywhere, anytime could subscribe to different governance services and combine them to form, legal and digitally enacted operating agreements. These operating agreements would define, manage, and actively govern “member” networks, where the members have agreed to be bound to the terms of the operating agreements. The Vermont Digital LLC project is just the beginning of a series of real world “field experiments” to develop digital operating agreements that can be potentially used for networks with millions of members.

Evolvable Rules and Contracts

One of the reasons for drafting contracts is to protect different parties from anticipated future contingencies. Yet this is virtually an impossible task given the limitations of the written word. For example, even if it were possible to enumerate all possible contingencies and fairly assign the appropriate remedies, the list would be of such a length and complexity that for many signatories it would defy comprehensibility.

Yet the fairness and soundness of a contract is in no small measure dependent upon it shouldering this impossible burden while preserving transparency, comprehensibility. This is impossible in written prose using rules or principles that do nothing more than iterate fixed lists. What is needed are types of rules that “evolve”, in effect, dynamic algorithms or grammars, that can generate an infinite number of variants that preserve certain prescribed relationships or conditions over time under varying conditions.

In one sense, that is the difference between a literal rule and a generative rule. A literal rule is just that; it creates a literal match between something new and old – something prior and something consequent. It does not allow for variants that do not match some initial definition. A literal rule for keeping a child warm and healthy, for instance, might prescribe specific articles of clothing and behavior, whereas an evolvable rule might allow for any means of behavior and clothing that keeps body temperature at 98.6.

In the case of contracts, the intent is to have evolvable rules that preserve the interests of the parties under future, variable conditions. In the language of evolutionary biology, it is a matter of having evolvable rule sets that match the “fitness functions” of different parties to a contract under variable environmental conditions. One areas of initial experimentation for us is to see if it is feasible to have evolvable provisions for a Term Sheet for the investment of a Series A round into a start up. The term sheet was taken from Wilson, Sonsini Goodrich & Rosati website that contains a “term sheet generator”.

The goal is to have a natural language processing program parse the term sheet into a Lisp based “contracting language” and then have a genetic programming language evolve multiple variants based upon the fitness conditions of the different parties, in this case, their relative holdings in the company capitalization tables under variable business conditions and investment scenarios. Ideally, one should be able modify the language of a term sheet and determine the outcomes or implications for the different parties or classes of shareholders. This research, however, is in a very preliminary phases and we do not expect to have any preliminary results until the next three months.

Outline for Developing Finance 3.0 Research and Approaches: Work in Progress

We are now investigating different approaches so develop a shared under standing across different levels of analysis and action.

These are:

Strategic: High level policy formulation; Setting the Mission/Vision, more emphasis on direction and outcomes and less concerned with means. Presidential and Thought Leader level

Tactical: Concern more with actual implementation of mission or vision by agencies, institutions or organizational means.

Operational: Concrete program, budgets and allocation of resources to achieve specific results as implementations of the prior two levels, where the rubber meets the road, boot on the ground.

Technological – These are technologies and methods that can significantly affect what is possible in the scope of a vision or how it is implemented. It can create new possibilities and cost structures that affect all three levels.

The goal is to create dialogue and eventual consensus across these different levels of activity.

—John Clippinger