An agency transaction is a deal in which one party, the agent, is authorized to act on behalf of another, the principal, in dealings with a third party, and the agent’s actions legally bind the principal even though the principal may never meet the third party face to face. This structure runs through real estate closings, stock trades, insurance sales, and countless commercial arrangements. Whether a relationship qualifies as an agency carries real consequences: it determines who bears financial risk, who can be sued if a deal goes bad, what duties the middle party owes, and whether a company reports the full transaction price as revenue or only its commission.
The Three Parties and the Direct Legal Link
Every agency transaction has three roles. The principal grants authority. The agent uses that authority to negotiate, sign contracts, or otherwise deal with outsiders. The third party is whoever the agent transacts with on the principal’s behalf.
The defining feature is that the agent creates a direct legal relationship between the principal and the third party. When a real estate agent negotiates the sale of a home, the resulting contract binds the homeowner and the buyer. The agent facilitated the deal but is not a party to it. That is what separates an agency transaction from a situation where someone buys goods on their own account and resells them.
How an Agency Relationship Gets Created
Agency can arise several ways, and not all of them require a signed document.
The most straightforward path is an express agreement. The principal appoints the agent directly, usually through a written contract or power of attorney spelling out what the agent can do. Oral appointments work in most situations, though certain acts (signing a deed on someone’s behalf, for example) typically require writing.
Agency can also form by implied conduct. If a business owner routinely lets a manager negotiate supply contracts without objection, that pattern can create agency without anyone ever saying so.
Estoppel applies when a principal’s words or actions lead a third party to reasonably believe an agency relationship exists. If the third party relies on that belief and would suffer a loss, the principal is blocked from later denying it.
Ratification works in reverse. Someone acts without any actual authority, and the “principal” later approves the deal. The approval operates retroactively, as if authorization had been in place all along. Ratification can be explicit or implied from conduct, such as accepting the benefits of the unauthorized transaction.
Finally, agency by necessity covers rare emergencies. A shipping company that must sell perishable cargo before it spoils, where getting the owner’s permission is not possible, may act as an agent of necessity for the cargo owner.
What Authority the Agent Actually Holds
The scope of an agent’s authority decides whether a third party can enforce a contract against the principal. There are two main categories.
Actual Authority
Actual authority is what the principal intentionally gives the agent, either expressly or by implication. Express authority comes through direct instructions: a written agency agreement might state that the agent can sign contracts up to a specific dollar amount. Implied authority covers whatever the agent reasonably needs to carry out those express instructions, such as hiring support staff or renting space to perform the work.1Legal Information Institute. Actual Authority
Apparent Authority
Apparent authority does not come from anything the principal tells the agent. It arises from the principal’s outward conduct toward the third party. If the principal’s actions would lead a reasonable third party to believe the agent has certain powers, the principal can be bound even where the agent’s actual authority was privately restricted.2Legal Information Institute. Apparent Authority
Courts weigh several factors together: the title or position the principal gave the agent, the principal’s history of accepting similar acts, representations made to third parties, and industry customs for someone in that role. No single factor is usually decisive. Giving someone a company credit card or business letterhead does not automatically create apparent authority. The question is how the principal held the agent out to the world as a whole.
Agent or Principal: Who Owns the Risk
Being an agent and acting as a principal in your own right are two different legal and economic positions. The distinction comes down to who controls the goods or services and who bears the financial risk.
An agent never takes ownership of what is being exchanged. A stockbroker executing a trade on your behalf matches your buy order with a seller, earns a commission, but never owns the shares. An insurance agent sells policies issued by the insurer without assuming the underwriting risk. A travel agent books flights and hotels on your behalf without buying the inventory first.
A principal acts on its own account. A distributor that buys products from a manufacturer, warehouses them, and resells them to retailers is a principal. The distributor owns the inventory, sets the retail price, and absorbs the loss if the goods do not sell. That assumption of risk is the clearest marker of a principal transaction. The same logic separates a securities dealer who buys bonds into inventory before reselling them from a broker who merely matches buyers and sellers.
What the Agent Owes the Principal
Once an agency relationship exists, the agent owes the principal fiduciary duties. Fiduciary duty is the highest standard of care the law recognizes. A breach can lead to termination, forfeiture of compensation, and significant damages.
Loyalty
The agent must act solely in the principal’s interest. No self-dealing, no secret profits, and no representing parties with conflicting interests in the same transaction. A purchasing agent who steers contracts to a vendor in exchange for kickbacks violates loyalty regardless of whether the vendor’s price was competitive. The prohibition is against the conflict itself.
Conflicts that cannot be avoided must be disclosed. If an agent has a financial relationship with a service provider they are recommending, or a personal interest in property involved in the transaction, the principal needs to know before agreeing to anything. Undisclosed conflicts are treated as a breach of loyalty even where the agent genuinely believed the recommendation was sound.
Care
The agent must act with the skill and diligence a reasonably competent person in that role would exercise. A financial advisor recommending investments must research those products. A real estate agent pricing a home must analyze comparable sales rather than guessing. Where the agent holds specialized credentials, the standard rises to match: a licensed appraiser is held to the standard of a competent appraiser, not a reasonable layperson.
Obedience
The agent must follow the principal’s lawful instructions, even where the agent disagrees. If a seller tells their real estate agent to reject all offers below a specific price, the agent must comply. The agent can advise against the instruction, but the principal calls the shots. The only exception is illegal or clearly unethical directives, which the agent is obligated to refuse.
Accounting
The agent must keep accurate records of all money and property handled on the principal’s behalf and must never mix the principal’s funds with their own. This is why real estate brokers maintain separate escrow accounts and why attorneys keep client funds in trust accounts distinct from their operating accounts. The principal can inspect those records at any time.
Agent Liability When the Deal Goes Wrong
Whether the agent is personally on the hook for a contract depends largely on what the third party knew about the principal.
Disclosed Principal
When the third party knows the agent is acting for a principal and knows the principal’s identity, the principal is “disclosed.” The agent generally has no personal liability on the contract. The deal is between the principal and the third party. If it falls apart, the third party pursues the principal.
The agent can still become liable in narrow situations, such as personally guaranteeing performance or acting outside the authority the principal granted. Those are exceptions to the general rule of agent protection.
Partially Disclosed Principal
When the third party knows the agent represents someone but does not know who, the principal is “partially disclosed” (also called “unidentified”). Here the agent is personally liable on the contract alongside the principal. The reasoning is fairness: the third party extended credit partly based on the agent’s own reputation, since the principal’s identity was unknown. An agent who gets stuck paying can seek reimbursement from the principal afterward, but the third party does not have to wait for that to play out.
Undisclosed Principal
When the third party has no idea an agency relationship exists and believes the agent is acting on their own behalf, the principal is “undisclosed.” The agent is fully liable because the third party relied entirely on the agent’s personal credibility.3Legal Information Institute. Undisclosed Principal
If the third party later discovers the hidden principal, the third party can generally pursue either the agent or the principal. Modern courts in many jurisdictions allow judgments against both, though the third party can collect only once. Once the full amount is recovered from one, the other is released.
When the Principal Pays for the Agent’s Mistakes
Contract liability is not the only exposure. When an agent injures someone or damages property while working, the principal can be liable in tort.
The doctrine of respondeat superior holds a principal (or employer) legally responsible for wrongful acts committed by an agent (or employee) acting within the scope of their duties. Courts apply this even where the principal had no involvement in the harmful act and was exercising reasonable oversight at the time.4Legal Information Institute. Respondeat Superior
There is no single national test for “within the scope.” Most courts use one of two approaches. Under the benefits test, if the agent’s actions were endorsed by the principal and were conceivably beneficial to the business, the principal bears liability. Under the characteristics test, if the conduct is common enough for the position that it could fairly be called characteristic of the job, the principal is liable.4Legal Information Institute. Respondeat Superior
Respondeat superior generally does not reach independent contractors, because by definition the principal does not control the details of how they work. Courts look past labels, though. Calling someone an “independent contractor” in a written agreement does not settle the question. Courts examine the actual relationship: how much control the principal exercises over the work, whether the worker uses their own tools and sets their own schedule, whether they serve multiple clients, and whether the principal controls the specific activity that caused the harm. If the reality looks like employment or agency, the label will not shield the principal.
How Agency Classification Changes the Financial Statements
Agent-versus-principal classification can swing reported revenue by hundreds of millions of dollars for a large company, even where the actual profit is identical either way.
A principal reports revenue on a gross basis. If a retailer sells a product for $100 that it purchased for $70, the retailer reports $100 in revenue and $70 in cost of goods sold, showing $30 of gross profit. The full $100 appears on the income statement because the retailer owned the product and bore the risk of it not selling.
An agent reports revenue on a net basis, recording only the fee or commission. If a ticket marketplace charges a 15% commission on a $200 concert ticket, it reports $30 in revenue, not $200. The $170 flowing to the venue never touches the marketplace’s revenue line. A company processing $10 billion in transactions as an agent might report only $1.5 billion in revenue, while the same company acting as a principal would report the full $10 billion.
Under U.S. accounting rules (ASC Topic 606) and the international equivalent (IFRS 15), the test centers on control. The entity that controls the good or service before it reaches the final customer is the principal. Control means the ability to direct the use of the asset and capture its remaining economic benefits.5Deloitte Accounting Research Tool. Determining Whether an Entity Is Acting as a Principal
Three indicators help determine which party has control. First, responsibility for fulfillment: the entity that guarantees the product works, handles returns, and absorbs problems looks more like a principal. An agent connects parties; a principal stands behind the product. Second, inventory risk: an entity that must acquire goods before a customer orders them, or that is stuck with unsold stock, is bearing principal-level risk. Third, pricing discretion: the entity that sets the price the customer pays, or can adjust it, is exercising principal-level control over the economic benefit. An agent typically earns a fixed fee or percentage set by someone else.
Credit risk also factors in. An entity that collects payment and absorbs the loss if the customer defaults looks more like a principal than an agent passing payment through. No single indicator is conclusive, and companies sometimes land in gray areas requiring judgment. Misclassifying the relationship leads to revenue restatements, which is why auditors scrutinize these determinations closely.
When the Agency Relationship Ends
Agency relationships do not last forever, and acts taken after termination generally do not bind the principal.
Either party can end the relationship voluntarily. The principal can revoke the agent’s authority, and the agent can renounce it. Where the agency was created by contract, early termination may trigger a breach-of-contract claim, but the authority itself still ends. The exception is an “agency coupled with an interest,” where the agent holds a stake in the subject matter. A lender who holds a security interest in property and also acts as the borrower’s agent to sell it has an interest that survives the principal’s attempt to revoke.
Certain events terminate an agency automatically. Death of the principal immediately revokes the agent’s authority in most situations, and the same applies when the agent dies, including the authority of any sub-agents the agent appointed. Mental incapacity of either party suspends or ends the relationship; a durable power of attorney is the specific tool designed to survive the principal’s incapacity, and without that designation, the agent’s authority ends. Bankruptcy of the principal terminates the agent’s authority over assets affected by the proceedings, and insolvency of the agent can similarly end things. If a change in law makes the agency’s purpose illegal, the relationship ends. If the property central to the agency is destroyed, the agency ends because its purpose can no longer be fulfilled.
Even after the agency ends, the principal should notify third parties who dealt with the former agent. Without that notice, a third party who reasonably believes the agent still has authority may be able to hold the principal liable under apparent authority, since the third party had no way of knowing the relationship ended.