ACCT90013 · Financial Accounting Theory And Practice
Information Economics
Chapter 1 said accounting exists to reduce information asymmetry. This chapter builds the micro-foundation: a single rational investor deciding under uncertainty. The payoff is a precise definition of why a financial report is useful — not because it is true, but because it can change the decision a rational person would make. That is the decision-usefulness approach, the bedrock of the information perspective. You model a risk-averse investor (concave utility) who chooses the act with the highest expected utility; treat information as evidence that can change a decision; and revise prior beliefs into posteriors using Bayes' theorem. Accounting is useful precisely when a signal moves the posterior enough to flip the chosen act. Scaling up from one investor to the market re-motivates the asymmetry problems — adverse selection and moral hazard — and shows how disclosure reduces them. The exam tests whether you can run a small expected-utility or Bayesian update and say what it means.
What this chapter covers
- 012.1 Decision theory under uncertainty (states, acts, payoffs)
- 02Risk aversion and concave utility; the expected-utility rule
- 03Information as evidence that can change a decision
- 04Bayesian revision: prior → information → posterior
- 05The decision-usefulness approach to financial reporting
- 06Asymmetry in the market: adverse selection (lemons, signalling) & moral hazard
Worked example: does the signal flip the decision?
- +1(a) Set up Bayes: P(High|good) = P(good|High)×P(High) ÷ P(good). Numerator = 0.80×0.60 = 0.48.
- +1(a) Denominator (total probability of a good signal): P(good) = 0.80×0.60 + 0.30×0.40 = 0.48 + 0.12 = 0.60.
- +1(a) Posterior: P(High|good) = 0.48 ÷ 0.60 = 0.80 — the good signal lifts the belief from 0.60 to 0.80.
- +1(b) Decision-usefulness: the signal is ‘useful’ only if that revision is enough to change the chosen act — e.g. if Pass was optimal at 0.60 but Buy becomes optimal at 0.80. Information that never flips the decision has no decision value, however accurate.
Key terms
- Expected utility
- The probability-weighted average of the utility of each outcome of an act. A rational decision-maker chooses the act with the highest expected utility — not the highest expected money, because utility is concave for a risk-averse investor.
- Risk aversion
- A preference captured by a concave utility function: a risk-averse investor values a certain amount more than a gamble with the same expected money, and so demands compensation for bearing risk. It is why diversification and risk-sharing in contracts have value.
- Bayes' theorem
- The rule for revising a prior probability into a posterior after observing a signal: P(state|signal) ∝ P(signal|state)×P(state). It is how a rational investor incorporates an accounting signal into their beliefs.
- Decision-usefulness
- The criterion that financial information is useful when it can change a decision — the signal moves the posterior enough to flip the chosen act. It defines usefulness by decision impact, not by truth, and underpins the information perspective.
- Signalling
- A response to adverse selection in which an informed party credibly reveals private information (e.g. a costly audit, a warranty, a dividend) so that good types separate themselves from bad. It lets the market price firms apart instead of pooling them.
Information Economics FAQ
Do I have to do the Bayesian arithmetic, or just explain it?
Both, but the marks split. ACCT90013 sets bounded quantitative items — a small expected-utility comparison or a one-step Bayesian update with clean numbers. Do the arithmetic to get the posterior, but the larger marks are in the interpretation: say what the revised belief means for the decision and tie it back to decision-usefulness. A correct number with no interpretation leaves marks on the table.
What does ‘information has value’ actually mean here?
Information has value only if it can change an action. Formally, a signal is valuable when, after updating, the optimal act differs from what you would have chosen on the prior alone. This is why “true but irrelevant” disclosure has no decision value, and why relevance (capacity to change a decision) is the headline qualitative characteristic of useful information.
How does a single-investor model connect to the rest of the subject?
It is the engine room. The single rational investor defines why accounting is useful (decision-usefulness), which the EMH then scales up to a whole market of such investors (capital-markets research). The same model's risk-aversion and asymmetry features motivate agency contracts and PAT. Master the one investor and the market-level chapters become applications of it.
Adverse selection vs moral hazard again — why does this chapter revisit them?
Because here you see the mechanism, not just the labels. Adverse selection (the ‘lemons’ problem) shows why hidden pre-contract information makes markets pool and good sellers withdraw — and why signalling/disclosure fixes it. Moral hazard shows why hidden post-contract action makes flat contracts fail — and why incentive pay/monitoring fixes it. The exam wants the mechanism and the matched mitigation, anchored to a scenario.
Exam move
Build two clean templates you can run cold. Template A — expected utility: list states, attach probabilities, compute the utility (not money) of each act, pick the highest, and say one sentence about risk aversion. Template B — Bayesian update: write P(state|signal) ∝ P(signal|state)×prior, normalise by the total probability of the signal, and then — the part that earns the most — state whether the revised belief flips the decision (decision-usefulness). Keep the arithmetic tidy so the interpretation sentence is where you spend your words. Finally, rehearse the asymmetry pair (adverse selection → signalling/disclosure; moral hazard → incentive pay/monitoring) as a matched set, always anchored to the named facts.
