Building Blockchains - Ripe Model for Principal-Agent Problem
Blockchain technology produced a new business model; one that enables companies to profit through creating and releasing open-source code. The decentralisation of business orgs creates new corporate governance challenges.
Management theory, in broad terms, deals with the relationship between managers and business entities. Inherent in this relationship is the principal-agent problem. This problem arises because the interests of a manager (agent) can — and often do — diverge from the interests of the owners of the business (principal) that he or she is managing.
Classic management incentivization: the carrot and the stick
Business organizations mitigate the principal-agent problem by use of incentive games that better align manager and business owner interests.
Example 1 (Reward-Based Game): A manager is incentivized to generate revenues for a business because this is a performance metric that will influence his or her compensation. Revenues also benefit the business and its owners by increasing a company’s equity value (benefiting shareholders), enabling the company to pay down debt (benefiting creditors), allowing employees to be paid on time, etc.
Example 2 (Deterrent-Based Game): A manager is deterred from acting in a manner that incurs excessive risk and liability for the business owners. One way this is achieved is through legal mechanisms such as vicarious liability or ‘piercing the corporate veil.’ The former may allow a manager to be held directly liable for the injury, or illegal conduct, of his or her employee; the latter may allow a manager to be held personally and solely liable in the context of fraud, etc.
I would hazard that the modern ‘business organization stack’ is built upon hundreds of different reward and deterrent incentive games, each playing a part in collectively establishing a Nash equilibrium between the ‘players’ within a business (i.e., managers and owners). These games are prevalent at all layers of the stack — e.g., compensation structures, human resources policies, governance policies, laws and regulations, etc. — and each game provides ‘checks and balances’ to the principal-agent problem that are fundamental to the viability of the organization.
Enter the ‘Cryptoeconomic Business Model’
With the advent of Blockchain-based assets — and the exponential influx of capital into the Blockchain industry over the past few years — we have witnessed the birth of a novel business model. This model enables companies to make money in new ways through the creation of open-source protocols and code (an invaluable service for which we once relied upon the altruism, rather than profit motive, of developers to provide).
I refer to this as the cryptoeconomic business model. This can be defined as any business model predicated on making profit by building a cryptoeconomic system, i.e., a peer-to-peer cryptographic network which functions on providing incentive payments to (assumed) adversarial nodes. Virtually all public/permissionless Blockchains today are ‘cryptoeconomic systems’ by this definition.
The cryptoeconomic business model upsets the classic principal-agent equilibrium that is often achieved by using reward and deterrent incentive games. This is done by introducing an entirely new class of stakeholder into the ecosystem — the Keepers of a Blockchain network (e.g. tokenholders and other participants who provide a form of ‘paid labor’ into the network, such as validators, miners, etc.)
If the traditional business has two classes of players (managers and owners), the cryptoeconomic business has three (managers, owners and Keepers).
These new entrants complicate the game theory model because, now, instead of the acting only on behalf of owners, there are two sets of stakeholders (owners and Keepers) whose interests depend on the efforts of a manager. What happens when the interests of these different sets of stakeholders diverge? In whose interests would (or should) an agent be motivated to act?
Token offering events & the risk of divergence/dilution
Value creation in a traditional business model is different than value creation in a cryptoeconomic business model.
In a traditional business, the final milestone of success is achieving profitability. Managers are incentivized to achieve profitability, and then to perpetually increase profitability, because the fruits of this labor accrue 100 percent to the business entity benefiting both owners and managers. Simple enough.
This is not exactly the case for a cryptoeconomic business model. Early in the cryptoeconomic business life cycle, each milestone benefits managers and owners collectively — but upon a company’s token offering event milestone (note: because the term ‘ICO’ is a faux pas) there is a fundamental shift.
Value creation no longer accrues to the business entity, but directly to the product/output of that business (i.e., the cryptoeconomic system)
In a cryptoeconomic business model, the final milestone is not profitability per se, but in the value of the Blockchain network/token, which recent scholarship suggests may be measured as a token’s current utility value (“CUV”) and discounted expected utility value (“DEUV”). CUV/DEUV come into play immediately following the token offering event milestone, concurrently with the introduction of Keepers into the stakeholder set.
This point is illustrated below:
So how does this impact our thinking on managerial incentives?
The immediate observation is that managers and owners will only benefit from working to increase a network’s value to the extent that they retain some amount of that network’s native tokens. In practice this amount might be in the ~20–50 percent range for the business entity, which is sizable, but significantly less than the 100 percent value retention model of a traditional business.
In theory, managers have ‘skin in the game’ by virtue of these token holdings and should be motivated to drive growth in the token’s CUV/DEUV with the expectation of selling those retained tokens for a profit at some later date. This outcome would be ideal as it implies an alignment between manager-owner-Keeper interests.
But the problem is that the dilution from 100% value retention (in a traditional business model) to ~20–50 percent value retention (in a cryptoeconomic business model) may also dilute a manager’s motivation to create long-term value for the network. Without sufficient reward/deterrent games in place, managers are prone to instances of moral hazard and myopic thinking.
It is plausible, for instance, that this may result in some degree of friction between the profit motive of managers, which incentivizes a manager to retain a significant portion of the tokens for the core business and the interests of the other Keepers/tokenholders who would benefit from those tokens being distributed more broadly thus creating network effects that could increase the CUV/DEUV of the token. This would be an example of misalignment between manager-owner-Keeper interests.
Other challenges in managerial motivation post-genesis block
Another challenge is due to the fact that revenue models (i.e. ‘rent-seeking’) may not be viable in cryptoeconomic systems. If a manager were to extract profit/revenue from a network by coding a centralized fee* into a protocol or dApp (i.e. any type of transaction fee that remits value back to the business), a likely outcome is that the protocol or dApp would either: (i) fail to gain adoption, or (ii) be hard forked by users (or duplicated by a competitor) to remove the fee from its code base thus making the network more cost-efficient.
*Note: To clarify my point on centralized fees, certain platforms use sustainable fee models as a feature of the platform’s cryptoeconomic design (e.g,. Factom and Counter-Party, wherein a portion of fees are burned to increase the scarcity of the token). Also, as the use cases for dApps/protocols continue to proliferate, centralized fees may prove to be an accepted business model for certain applications of Blockchain technology.
Here are a few of the other ramifications of this challenge:
Profiting upfront; creating value later: The creators of cryptoeconomic networks (currently) realize value for the business entity primarily via two streams: (i) the proceeds of token offering events, and (ii) the retention of some amount of the offered tokens. Both of these milestones occur relatively early in the life cycle of a business. Given that the majority of a manager’s compensation/profit is front-loaded, experience has shown that some managers will opt to simply complete a token offering event before ‘jumping ship’ to the next project, rather than working to generate value for their current project.
CUV/DEUV is a bad indicator of managerial competence: We may not yet have the best tools to evaluate managerial performance in cryptoeconomic business models. CUV/DEUV are inherently different metrics than earnings per share, EBITDA, return on equity, etc. (the latter are some of the tools used to evaluate CEO performance in a traditional business). CUV/DEUV is driven by supply and demand; more fitting for valuing a commodity than equity. To evaluate a manager’s performance on the CUV/DEUV of a token is akin to evaluating a gold company CEO’s performance on the price of gold.
The lack of legal mechanisms to protect Keepers/token-holders: There exists an elaborate body of corporate, securities and employment law designed to address the principal-agent problem between participants in traditional business structures (e.g., vicarious liability, ‘piercing the corporate veil,’ fiduciary duties owed by directors to shareholders, etc.) These protections do not (yet) exist for the Keepers/token-holders of cryptoeconomic systems. Granted, there is free market mechanism in play by virtue of the Keepers’ ability to hard fork a protocol in retaliation to mismanagement, but this overhaul should only be used as a last resort.
Singular token offering events: For traditional start-ups, the process of raising capital occurs in tranches (i.e. Seed, Series A, Series B, etc.) and each tranche is largely tied to a manager’s ability to demonstrate progress towards profitability since the previous tranche. Token offerings — on the other hand — are mostly structured as singular events. This structure alleviates the much needed external pressure on managers to deliver on building their products on time and on budget. It also fails to backstop losses for investors in the event that a manager fails to deliver.
These are just a few examples of how the principal-agent problem can manifest itself in the context of new, cryptoeconomic business models— each of which will eventually be solved by new incentive games designed for the tripartite (i.e., manager-owner-Keeper) environment.
I suspect that the study of management theory in the context of cryptoeconomic business models will continue to be an evolving field — and a very relevant one at that.
Many thanks to Ryan Zurrer, Lawrence Krimker, Lauren Furman, Marc Pontone and Josh Teichman for their contributions to this article. I welcome any comments or feedback.