The principal-agent problem arises when a principal has an agent who is able to make decisions on their behalf. That’s the principal-agent relationship.
The problem arises when their goals are misaligned or differently incentivised, and the agent pursues their own goals rather than their principal’s.
The principal-agent problem: possession, access, information, goals
For example, suppose you own a property in a country halfway around the world. You can’t visit often but you don’t want to sell. So, you hire someone to look after the place.
You’re now the principal; they are now the agent, empowered to make decisions about how to keep your property safe and secure. But you and they have differing needs and goals. You want your property kept liveable and saleable. They have the keys to a big empty house that no-one but them checks on. Will they have a big party? Sublet it room by room? Move family members in rent-free? Collect the fee and never bother to even visit the property? Or will they meet the terms of their engagement faithfully?
The principal and the agent have different goals and intentions. This, when it’s applied to financial management, economics, and supply chains — and essentially any other circumstance where someone acts as an agent on behalf of another person or entity — is the principal-agent problem.
The origin of the term is in dispute, but the most common attribution is to Michael C. Jensen and William H. Meckling’s 1976 paper, “Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure”.
For instance, corporate managers like CEOs are agents; their shareholders are the principal, to whom they have a fiduciary duty. Elected officials are agents; citizens are the principal. Lawyers are agents; clients are principals.
The origins and pitfalls of the principal-agent relationship
We reach for agents when we need expertise, labour or availability we don’t have ourselves; there’s a reason lawyers charge so much. Yet we also know that our agents aren’t “in the same boat” as we are. They may have different goals. They almost always have asymmetric information — that’s what principals engage agents for, in many cases.
So it’s difficult for principals to ensure that agents are really acting in the best interests of the principal — rather than, potentially, in the best interests of the agent themselves. Is your lawyer’s legal strategy really right? You could lose a major civil or even criminal case if not, but they get paid either way. Is your financial advisor’s plan for your portfolio going to earn you the best returns — or them the best fees? You probably don’t have the subject-area knowledge to answer those questions.
This is known as moral hazard — the temptation for agents to increase their exposure to risk, because they will share in the profits if the risk pays off, but pay no part of the price if it does not. Economist Paul Krugman defines the term as “any situation in which one person makes the decision about how much risk to take, while someone else bears the cost if things go badly”. That’s one reason why personal and institutional trust is so important.
Where agents have one set of interests and principals another, you have a conflict of interest. This is a circumstance where efforts in pursuit of one interest can work against another — as when a financial advisor could make trades that earn them fees but reduces your returns. This doesn’t always become a problem, but it is the reason the same lawyer can’t represent both parties in a divorce, for instance. If they were to become inimicable, for whom should the lawyer advocate?
Where an agent deviates from the pursuit of the principal’s interests, we have agency cost. This is the cost to the principal arising from reliance on the agent. In many cases the agency cost is nonzero, yet the benefit of using the agent outweighs the cost. Where this isn’t, or is feared not to be, the case, avoidance of it imposes an opportunity cost on the principal. You pay more in time and energy to avoid potential principal-agent problems. Again, the importance of trust.
Solutions to the principal-agent problem
There are solutions to the principal-agent problem. The two most relevant are professional ethics and codes of conduct, and regulation.
Professional ethics are there to identify the interests of the professional with those of the client. Doctors are professionally obligated to advocate for their patients, attorneys for their clients. Corporate managers have fiduciary duty. Codes of conduct and ethics within professions help to generate trust by defining the parameters of the professional role in terms of identification with the principal’s interests. For example, The CFA Code of Ethics, the ethical code for investment professionals in the USA, says members “must:
- Place the integrity of the profession and the interests of clients above your own interests
- Act with integrity, competence, and respect
- Maintain and develop your professional competence”
This is clearly aimed at creating a professional role that isn’t vulnerable to developing principal-agent problems.
In finance, as in many other areas, there’s regulatory backing for these professional structures, as well as overarching regulation that seeks to punish or avoid moral hazard.
What does the principal-agent problem look like on the ground?
One good example is realtors. When you engage a real estate broker to sell your home, it’s not unreasonable to expect that they’ll sell your home for more than you would. But it’s not true.
According to Stanford economist B. Douglas Bernheim and Stanford grad student Jonathan Meer’s 2007 paper, “How Much do Real Estate Brokers Add? A Case Study”, the answer is:
“We find no evidence that the use of a broker leads to higher average selling prices, or that it significantly alters average initial asking prices. However, those who use brokers sell their houses more quickly”.
In other words, brokers are primarily interested in getting homes sold, so they can get their commission — not in getting higher prices for homes. The principal and the agent have different goals.
Another example comes from within the financial industry. Enron was once one of the biggest companies in the US, but began losing money. Concerned for share prices, management hid the losses using complex financial instruments.
Many shareholders were unable to understand; the board of directors had a legal and ethical responsibility to intervene, but was otherwise incentivized to permit the company’s drift into illegality.
Enron went under in 2001, amid a slew of lawsuits, criminal proceedings and questions about how to avoid something similar happening again. The principals — Enron’s shareholders and other stakeholders — and the agents, the company’s management and directors, had radically different goals and incentives.
What about the world of digital assets?
In many ways the world of digital assets is about the rediscovery of traditional finance — the realization that even with blockchain-powered security, you have to come off-chain sometime; at that point, digital assets are vulnerable just as any other asset, and ancient forms of larceny have been rediscovered apace with the traditional financial solutions to them.
Our case is Gerald Cotten, whose QuadrigaCX was once one of Canada’s biggest digital asset exchanges. Then, in 2018, Cotten died in India from complications from Crohn’s Disease and was promptly cremated, at the age of just 30. At first the big story was that he had personally held the private keys for the QuadrigaCX cold wallets, which held between $190 million and $250 million in digital assets, mostly BTC.
Shortly thereafter, the story began to pivot, as investigators discovered that these wallets were not just locked: they were empty. Where was the BTC? No-one knows — but it was probably never held or invested as it was supposed to be. “What Quadriga really did with the money that customers gave it to buy Bitcoin”, says digital asset tracking company Chainalysis, “remains a mystery”.
As an agent, Cotten was entrusted by many principals with their investments. He presented himself to them as a person who would manage their finances in a responsible and trustworthy manner. Whatever he did — and theories abound — it wasn’t that. A review by the Ontario Securities Commission found that “[t]he bulk of the asset shortfall—approximately $115 million—arose from Cotten’s fraudulent trading on the Quadriga platform”.
While Cotten’s conduct was particularly dramatic, it wasn’t particularly unusual “Exit fraud” — meaning, take customers’ money out of a project and run away with it — was a common form of theft in the digital asset space, though reputable projects were unaffected for the most part and losses were low. (This was small comfort to those who suffered such losses.)
Again, the problem is that the agent had no incentive to line up their activities with the interests of their principals.
How can this problem be combatted specifically in the digital asset space?
Solutions to the principal-agent problem in digital assets
One answer is the smart contract.
Smart contracts are self-enforcing. They always and only perform operations when the conditions have been met, meaning they can hold funds in escrow during transactions and release them when work has been done or when markets reach certain conditions.
There is undoubtedly a place for smart contracts in protecting digital asset investors, but the main area of progress will be the adoption of the professional and regulatory wisdom of traditional finance.
What let Gerald Cotten abstract millions from those who trusted his business was the absence of these professional structures: no divisions of power, regulatory oversight, professional training or ethics concerns, just a recent business school graduate personally holding the keys to millions of dollars of other people’s money. The opportunities for abuse are as glaring as they were when banks held other people’s money in gold; the solutions are the same too.
We’ve seen that the digital asset space likes to consider itself a world apart, untrammeled by the restrictions and bureaucracy of traditional finance. But we’ve also seen that, time and again, it’s having to rediscover concepts like oversight, fiduciary duty, conflict of interest, and the traditional trust and custody arrangements that millions rely on to hold their assets safely, securely and reliably.