ACCT90013 · Financial Accounting Theory And Practice
Measurement and Positive Accounting Theory
The previous chapter trusted an efficient market to price the firm. This chapter asks two harder questions and joins the subject's two halves. First, the measurement question: should accountants stop at supplying information and let the market value the firm, or take a larger role by putting fair values directly on the balance sheet? That is the information vs measurement perspective, and it forces the central relevance-vs-reliability trade-off (plus recognition vs disclosure). Fair value is more relevant but often less reliable; historical cost is the reverse — and every measurement debate in the subject is a version of this tension. Second, the contracting question: once managers and owners have different interests, accounting's job becomes stewardship — writing and enforcing contracts. That is agency theory: the principal delegates to the agent, moral hazard and risk-sharing generate agency costs, and the response is monitoring, bonding and above all executive compensation tied to accounting performance — trading incentive sensitivity against precision and the risk imposed on a risk-averse manager. These ideas are the platform on which Positive Accounting Theory (next chapter) predicts the policies managers actually choose.
What this chapter covers
- 01Information vs measurement perspective on financial reporting
- 02Fair value on the balance sheet vs historical cost
- 03The relevance-vs-reliability trade-off
- 04Recognition vs disclosure
- 05Agency theory: the principal–agent relationship and agency costs
- 06Executive compensation: pay-for-performance, sensitivity vs precision, risk & incentives, governance
Worked example: design the pay contract
- +1(a) Name the setting: principal (board/owners) delegates to a risk-averse agent (CEO) whose effort is unobservable — moral hazard. Pay must motivate effort while controlling the risk imposed on the agent.
- +1(a) The two properties: a good measure has high sensitivity (responds to the agent's effort) and high precision (little noise unrelated to effort). X is sensitive but noisy; Y is precise about value but loaded with uncontrollable market risk.
- +1Trade-off: loading on Y exposes the risk-averse CEO to market shocks (costly risk premium); loading on X rewards effort but lets earnings noise leak into pay. Neither alone is ideal.
- +1(b) Recommend a blend: weight accounting earnings (effort-sensitive) for incentives and add a smaller, relative-to-peers share-price component to filter common market risk — “here, pay on controllable earnings plus relative TSR aligns effort without over-charging the CEO for shocks he can't control.”
Key terms
- Information perspective
- The view that accounting's job is to report reliably and let an efficient market do the valuing — supply decision-useful information rather than try to measure value directly. It pairs naturally with the EMH.
- Measurement perspective
- The view that accountants should take a larger role and build fair values into the financial statements themselves, making the balance sheet a more direct estimate of value — at the cost of reliability.
- Relevance vs reliability
- The central measurement trade-off. Relevance = capacity to change a decision (fair value scores high); reliability / faithful representation = freedom from error and bias (historical cost scores high). Almost every measurement debate is this tension.
- Agency theory
- The analysis of the principal–agent relationship: an owner delegates to a manager with diverging interests and asymmetric information. The resulting agency costs (monitoring, bonding, residual loss) drive compensation, covenants and governance.
- Sensitivity vs precision
- Two properties of a performance measure used in a pay contract. Sensitivity = how strongly it responds to the agent's effort; precision = how little noise it carries. A good incentive measure is sensitive and precise; real measures trade one off against the other.
Measurement and Positive Accounting Theory FAQ
Why are measurement and agency taught together in one chapter?
Because they are the two faces of accounting's dual role made concrete. Measurement (fair value, relevance vs reliability) is the valuation/information role asking how directly accounting should estimate value. Agency (contracts, compensation) is the stewardship/contracting role asking how accounting disciplines managers. Seeing them side by side is the point: the same number that is most relevant for valuation can be least reliable for contracting, which is exactly why the two roles conflict.
Is fair value ‘better’ than historical cost?
Neither is universally better — it's a relevance-vs-reliability trade-off. Fair value is more relevant (closer to today's value, more decision-useful) but often less reliable (estimated, volatile, manipulable, especially for illiquid Level-3 items). Historical cost is more reliable but can be stale and less relevant. The exam wants you to argue the trade-off for the specific asset and use, not to crown a winner.
What exactly are ‘agency costs’?
The total cost of the principal–agent conflict: monitoring costs (the principal watching the agent — audits, boards), bonding costs (the agent credibly committing — covenants, contracts) and the residual loss (the value still lost because alignment is never perfect). Accounting reduces agency costs by providing the contractible performance signals that compensation and debt contracts run on.
How do I answer a ‘design the compensation contract’ question?
Run the agency template: (1) name the principal, agent and the unobservable action (moral hazard); (2) judge candidate measures on sensitivity (responds to effort) and precision (low noise); (3) note the risk imposed on the risk-averse agent — loading on noisy or uncontrollable measures forces a costly risk premium; (4) recommend a blend and justify with the applied sentence. Mentioning relative performance evaluation (peer-filtering) to remove common risk often earns the top mark.
Exam move
Hold two templates ready. Measurement: for any “how should we measure X” prompt, state the information vs measurement perspectives, then argue relevance vs reliability for that specific item (and recognition vs disclosure if it's borderline) — conclude with a position, not a fence-sit. Agency / compensation: name principal, agent and the unobservable action; score measures on sensitivity vs precision; flag the risk premium charged by a risk-averse agent; recommend a blend (often controllable earnings + relative share-price). The single most common lost mark here is reciting “fair value is relevant” without tying it to the asset and use — always anchor the trade-off to the named facts. This chapter is also the springboard to PAT, so keep the contracting vocabulary (covenants, monitoring, conservatism) fresh for the next chapter.
