MBA Compensation Design Concept Card Template

Compensation Design Concept Card: [Your name, course or project, and date]

A core-course concept card for designing pay: the pay mix (fixed against variable, cash against equity), incentive design (measures, thresholds, caps and accelerators), equity compensation (restricted stock units and stock options, dilution and burn rate) and the maths of pay for performance (the optimal incentive intensity of agency theory, pay-performance sensitivity and the value of an option grant). It is for MBA students in organisational behaviour, human resources, corporate governance and entrepreneurship courses, for founders who must design their first sales plan and option pool, and for managers who sit on a remuneration or compensation committee. Worked example A designs a sales commission plan and an executive equity grant for Corvane Robotics and computes a clawback, Worked example B is a half-solved store manager scorecard for Northfield Outdoor Co., and Exercise C asks for incentive intensity, an option repricing and a plan critique. Source basis: Kerr (1975), Jensen and Meckling (1976), Holmstrom (1979), Lazear and Rosen (1981), Jensen and Murphy (1990), Holmstrom and Milgrom (1991), Gibbons (1998), Lazear (2000), Bebchuk and Fried (2003), Hall and Murphy (2003) and Black and Scholes (1973), written up in original words with fictional numbers; dated rules (US SEC Rule 10D-1, UK Corporate Governance Code 2024, EU Pay Transparency Directive) are linked to official pages. Replace every [bracketed] value and every row marked Example. This is a learning aid, not legal, tax, accounting or financial advice.

Document control

Contents

• When to use it
• The idea in plain language
• Formulas and notation
• Assumptions and limits
• Rules that shape pay design
• The design process step by step
• Worked example A (fully solved): Corvane Robotics sales and executive pay
• Worked example B (half solved): Northfield store manager scorecard
• Exercise C (unsolved): intensity, repricing and a plan critique
• Connections to other cards
• Common mistakes and misconceptions
• Self-test questions
• My notes and next steps
• Sources and further reading
• Answer key

When to use it

Use this card when:

• you design or review a variable pay plan: a sales commission, an annual bonus scorecard, a team incentive or a long-term incentive for executives;
• a start-up must set its option pool, decide between options and restricted stock, and explain an offer's equity to a candidate (mba/mba-offer-negotiation covers the candidate's side);
• a remuneration or compensation committee must judge whether executive pay is linked to performance and whether the plan meets governance rules on clawback and disclosure;
• a case asks why an incentive plan produced the wrong behaviour (gaming, short-termism, neglect of unmeasured work, unhealthy competition);
• you need to compare the value of pay packages that mix salary, bonus and equity.

Do not use incentive pay to fix a problem of selection, training or tools. Where output is hard to measure, depends mostly on luck or on teamwork, or where intrinsic motivation is strong, a higher salary and good management often work better than a sharper incentive (mba/mba-motivation-card).

The idea in plain language

An employer wants effort and good decisions it cannot fully observe; the employee chooses how hard to work and what to work on. This is the agency problem (Jensen and Meckling 1976). Paying for measured performance aligns interests, but it has two costs. The first is risk: measured performance depends on luck as well as effort, so a strong incentive loads risk onto the employee, who must be paid a premium to bear it. Holmstrom (1979) showed that a good contract uses every measure that tells the employer something about effort, and filters out noise the employee cannot control (for example by comparing a salesperson's results with peers in the same market). The second is distortion: when a job has several tasks and only some are measured, a strong incentive on the measured task pulls effort away from the unmeasured ones (Holmstrom and Milgrom 1991). Kerr (1975) gave the managerial version: firms often reward A while hoping for B.

Three further ideas complete the theory. Selection: incentive pay attracts people who expect to do well under it; Lazear (2000) found that a move from hourly wages to piece rates at a windshield installer raised productivity by about 44%, roughly half from more effort and half from attracting and keeping more productive workers. Tournaments: promotions and rank-based bonuses reward relative performance, which filters out common shocks but can discourage cooperation (Lazear and Rosen 1981). Executive pay: Jensen and Murphy (1990) measured how little chief executives' wealth changed with shareholder wealth in their sample and argued for more equity-based pay; Bebchuk and Fried (2003) argued that executives can influence their own pay, so pay design is also a governance problem; Hall and Murphy (2003) explained why stock options cost the company more than they are worth to risk-averse, undiversified executives.

Real plans translate these ideas into a few design choices: the pay mix (how much of target pay is variable), the measures (what is rewarded and how noisy each measure is), the payout curve (threshold, target, maximum, accelerators and caps), the time horizon (annual cash, multi-year equity, deferral and holding periods) and the safeguards (malus, clawback, discretion, disclosure). The maths in the next section tells you which way each choice should move; the worked examples show what the numbers look like.

Formulas and notation

Linear principal-agent model (Holmstrom and Milgrom 1987 and 1991, as commonly taught): output $x = e + \varepsilon$ with noise $\varepsilon \sim N(0, \sigma^2)$, pay $w = s + b\,x$, private cost of effort $c\,e^2/2$, agent's constant absolute risk aversion $r$. The agent chooses $e^* = b/c$, and the incentive intensity that maximises total surplus is

Total certainty-equivalent surplus at intensity $b$:

Commission payout with threshold $\theta$, quota $Q$, base rate $k = I_{target}/Q$, accelerated rate $k_a$ and cap at attainment $\bar{a}$, where $a = \text{bookings}/Q$:

Pay-performance sensitivity (Jensen and Murphy 1990): change in executive wealth per USD 1,000 change in shareholder value, for $n$ share-equivalents held and $N$ shares outstanding (an option counts as its delta $\Delta$):

Black-Scholes value of a call option (Black and Scholes 1973), used for grant-date fair value:

Equity budget measures:

Excel: the commission is =IF(A<Thr,0,IF(A<=1,Rate*A*Quota,Target_inc+Acc_rate*(MIN(A,Cap)-1)*Quota)); the optimal intensity is =1/(1+r*c*sigma^2); the option value is =S*NORM.S.DIST(d1,TRUE)-K*EXP(-rf*T)*NORM.S.DIST(d1-vol*SQRT(T),TRUE) with d1=(LN(S/K)+(rf+vol^2/2)*T)/(vol*SQRT(T)).

Assumptions and limits

• The linear model is a teaching model. Real effort costs, risk attitudes and noise are not known numbers. Use the formula for direction: noisier measures, more risk-averse employees and harder-to-move output call for weaker incentives.
• One measure, one task in the formula. With several tasks, the multitask result applies: if an important task cannot be measured, weaken the incentive on the measured one or measure both (Holmstrom and Milgrom 1991).
• Payout curves assume honest reporting. Every threshold, cap and period end invites timing games (pulling deals into the quarter, holding them back once the cap is reached). Smooth curves and multi-period measures reduce this.
• Option values use Black-Scholes with an expected term instead of the contractual term, no dividends and constant volatility. Accounting standards (ASC 718 in US GAAP, IFRS 2 internationally) require a fair-value measure at grant; private companies must also estimate the share price itself. The employee's subjective value is lower than the fair value when the employee is undiversified (Hall and Murphy 2003).
• Tax and accounting are ignored in the worked numbers. In practice, the timing and rate of tax on options and restricted stock, social charges and deduction limits change the cost and value of each instrument by country.
• Fictional numbers. Corvane Robotics and Northfield Outdoor Co. are fictional, and every figure is an example to test the formulas, not a market benchmark. Use a current pay survey for real levels.

Rules that shape pay design

Dated facts, checked 2026-10-30. This is a summary for learning, not legal advice.

The design process step by step

Start from the behaviour and the business result you want, then choose measures, then the amount at risk, then the curve, then the safeguards, and finally test the plan at low, target and high performance before anyone sees it.

Worked example A (fully solved): Corvane Robotics sales and executive pay

Example inputs: Corvane Robotics (fictional) sells autonomous warehouse robots and fleet software. It is preparing for a listing and wants a commission plan for account executives and a long-term incentive grant for a new chief operating officer.

Step 1: the account executive commission plan (USD)

On-target earnings (OTE) are 200,000 with a 60:40 pay mix: base salary 120,000 and target incentive 80,000. Quota is 1,600,000 of annual contract value. Below 50% attainment no commission is paid; from 50% to 100% the rate is 5.0% of bookings (=80000/1600000); above quota the accelerated rate is 7.5% (1.5 times the base rate); payout is capped at 200% attainment.

Read the curve as the salesperson will. The jump from 0 to 40,000 at 50% attainment is a cliff: a salesperson at 45% near the end of the year has a strong reason to pull a deal forward or discount it to cross the threshold, and one at 30% has no reason to try. The accelerator makes the marginal deal above quota worth 7.5% instead of 5.0%, which is the point of an accelerator, and the cap stops the marginal reward at 200%, where a salesperson may delay deals into the next year. Cost of sales falls as attainment rises (from 20.00% of bookings at threshold to 10.00% at the cap), so the company gains from strong years even with the accelerator.

Step 2: how strong should the incentive be?

Normalise output so that $c = 1$ and take $r = 0.5$. The optimal intensity $b^* = 1/(1 + 0.5\,\sigma^2)$ falls as the measure gets noisier:

The lesson for Corvane: enterprise robot deals are large, few and lumpy, so a single account executive's annual bookings are a noisy measure of effort. That argues for a lower share at risk than a high-volume software sales team would use (60:40 rather than 50:50), for a multi-year or rolling quota, and for team or pipeline measures alongside bookings. Comparing each executive with peers in the same region filters out common shocks (Holmstrom 1979).

Step 3: the executive equity grant

The new chief operating officer receives a long-term incentive with a grant-date value of USD 600,000. The share price is USD 10.00 and there are 50,000,000 shares outstanding. Option inputs: strike 10.00 (at the money), expected term 6 years, volatility 50%, risk-free rate 4.0%, no dividends.

Value to the executive at vesting for four share prices (intrinsic value, before tax):

The two grants cost the same at grant, but they behave differently. RSUs keep value when the share price falls, so they retain the executive in bad years and have a lower PPS. Options give nothing below the strike and more than RSUs only above about USD 21 (the share price at which 114,453 x (P - 10) equals 60,000 x P, which is =114453*10/(114453-60000) = 21.02); they pay for upside and encourage risk taking, which suits a growth company only if the board wants that risk. Many plans mix the two, and listed companies add performance conditions.

Step 4: dilution and the equity budget

Corvane grants 1,400,000 shares' worth of awards in the year, has 4,200,000 awards outstanding and 2,800,000 shares still available in the plan, against 50,000,000 weighted average shares outstanding.

Investors and proxy advisers watch both numbers. Granting options instead of RSUs for the same value uses almost twice as many shares (114,453 against 60,000), so the choice of instrument changes the burn rate as well as the incentive.

Step 5: a clawback under a restatement

The chief operating officer's annual bonus pays 50% of a USD 300,000 target at revenue of USD 86.0m, 100% at 92.0m and 200% at 98.0m, straight-line between points. Reported revenue was 92.0m, so the bonus paid was 300,000. A year later an error in revenue recognition leads to a restatement: revenue becomes 88.5m.

Under a Rule 10D-1 policy (PAY-R01) the company must seek to recover 87,500, computed before tax and regardless of whether the executive was at fault. The design lesson: write the recomputation method into the plan before the first payout, including how to treat measures such as total shareholder return where the effect of a restatement must be estimated.

Worked example B (half solved): Northfield store manager scorecard

Northfield Outdoor Co. (fictional) pays its store managers a target bonus of 20% of a USD 80,000 base salary (USD 16,000). The scorecard has three measures:

The sales score is =100%+(5.0%-4.0%)/(8.0%-4.0%)*50% = 112.5%. For the shrink measure, remember that a lower value is better, so measure progress from the threshold downwards.

Design questions: [ ]

1. Which measure is most exposed to luck (weather, a new competitor nearby), and how would you filter it?
2. Which important part of a store manager's job does this scorecard not measure, and what does the multitask result predict about it?
3. Should the bonus pay anything if one measure is below threshold but the others are strong? Argue both sides.

Exercise C (unsolved): intensity, repricing and a plan critique

C1. In the linear model with $r = 0.5$ and $c = 1$, the noise of a measure is $\sigma = 1.2$. Find the optimal intensity $b^*$ and the total surplus at $b^*$. Compare it with the surplus if the firm uses $b = 1$ (full residual claim). What does the comparison say about "making every employee an owner"?

C2. A year after the grant in Worked example A, step 3, Corvane's share price has fallen to USD 6.00. The options now have five years of expected term left; volatility and the risk-free rate are unchanged. (a) What is each existing option worth, and the whole grant? (b) What is a new at-the-money option (strike 6.00, five years) worth? (c) How many new options would a value-neutral exchange give the executive for the 114,453 old ones? (d) What would a one-for-one repricing cost the shareholders in incremental fair value? (e) Give one argument for and one against repricing.

C3. A sales director proposes: "Pay 10% of every deal's value, no base salary, no cap, paid when the contract is signed." Write five sentences on what behaviour this plan will produce at Corvane and how you would change it.

Connections to other cards

Common mistakes and misconceptions

1. Rewarding what is easy to count instead of what matters (Kerr 1975). Bookings without margin, calls without quality, short-term earnings without investment.
2. Ignoring noise. A strong incentive on a noisy measure pays for luck and costs a risk premium; filter common shocks with relative measures or longer periods.
3. Cliffs and caps without a reason. Each kink in a payout curve creates a timing game; use smooth curves unless a kink serves a purpose you can state.
4. Too many measures. A scorecard with ten measures gives each a weight too small to change behaviour; three to five is a common range.
5. Comparing options and RSUs by share count. Compare by fair value and by how value changes with the share price; one option is worth less than one share.
6. Treating the option's fair value as its value to the employee. An undiversified, risk-averse employee values an option below its cost to the company (Hall and Murphy 2003).
7. Forgetting dilution. Every equity award is paid by existing shareholders; track burn rate and overhang.
8. No recomputation rule for clawback. Without a method fixed in advance, recovery after a restatement becomes a negotiation.
9. Assuming more incentive is always better. Where tasks are hard to measure or teamwork matters, weaker individual incentives with good management can outperform (Holmstrom and Milgrom 1991; Gibbons 1998).
10. Designing pay without the pay-equity check. Objective, documented criteria are now a legal expectation in the EU (PAY-R06) and good practice everywhere.

Self-test questions

1. State the two costs of paying for measured performance.
2. In the linear model, how does $b^*$ change when risk aversion doubles? Why?
3. What does the multitask model predict when one important task cannot be measured?
4. In Worked example A, what is the commission at 120% attainment? Show the Excel formula.
5. Why does the 50% threshold in Worked example A create a gaming problem, and how would you remove it?
6. Explain the two sources of the productivity gain Lazear (2000) found after a switch to piece rates.
7. What does a pay-performance sensitivity of USD 1.20 per USD 1,000 mean?
8. Why does an option grant of the same fair value as an RSU grant use more shares?
9. At what share price do the option and RSU grants in step 3 give the executive the same value?
10. Under Rule 10D-1, does recovery depend on the executive's fault? How is the amount computed?
11. What is the difference between malus and clawback?
12. Give one argument from Bebchuk and Fried (2003) for why boards may not set executive pay at arm's length.

My notes and next steps

Sources and further reading

Primary works:

• Kerr, S. (1975). On the folly of rewarding A, while hoping for B. Academy of Management Journal, 18(4), 769-783.
• Jensen, M. C., and Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3(4), 305-360.
• Holmstrom, B. (1979). Moral hazard and observability. The Bell Journal of Economics, 10(1), 74-91.
• Lazear, E. P., and Rosen, S. (1981). Rank-order tournaments as optimum labor contracts. Journal of Political Economy, 89(5), 841-864.
• Jensen, M. C., and Murphy, K. J. (1990). Performance pay and top-management incentives. Journal of Political Economy, 98(2), 225-264.
• Holmstrom, B., and Milgrom, P. (1987). Aggregation and linearity in the provision of intertemporal incentives. Econometrica, 55(2), 303-328.
• Holmstrom, B., and Milgrom, P. (1991). Multitask principal-agent analyses: Incentive contracts, asset ownership, and job design. Journal of Law, Economics, and Organization, 7 (special issue), 24-52.
• Gibbons, R. (1998). Incentives in organizations. Journal of Economic Perspectives, 12(4), 115-132.
• Lazear, E. P. (2000). Performance pay and productivity. American Economic Review, 90(5), 1346-1361.
• Bebchuk, L. A., and Fried, J. M. (2003). Executive compensation as an agency problem. Journal of Economic Perspectives, 17(3), 71-92.
• Hall, B. J., and Murphy, K. J. (2003). The trouble with stock options. Journal of Economic Perspectives, 17(3), 49-70.
• Black, F., and Scholes, M. (1973). The pricing of options and corporate liabilities. Journal of Political Economy, 81(3), 637-654.

Official pages for dated rules (checked 2026-10-30):

• 17 CFR 240.10D-1, https://www.law.cornell.edu/cfr/text/17/240.10D-1, and SEC adopting release page, https://www.sec.gov/rules-regulations/2022/10/listing-standards-recovery-erroneously-awarded-compensation.
• SEC, Pay versus performance, https://www.sec.gov/rules-regulations/2022/08/pay-versus-performance.
• 26 U.S. Code section 162, https://www.law.cornell.edu/uscode/text/26/162.
• Financial Reporting Council, UK Corporate Governance Code, https://www.frc.org.uk/library/standards-codes-policy/corporate-governance/uk-corporate-governance-code/.
• Directive (EU) 2017/828 (Shareholder Rights Directive II), http://data.europa.eu/eli/dir/2017/828/oj.
• Directive (EU) 2023/970 (Pay Transparency Directive), http://data.europa.eu/eli/dir/2023/970/oj.

Textbooks (further reading, by name only): Milgrom and Roberts, Economics, Organization and Management; Lazear and Gibbs, Personnel Economics in Practice; Brickley, Smith and Zimmerman, Managerial Economics and Organizational Architecture.

Answer key

Worked example B: shrink score 70.0% (=50%+(2.0%-1.8%)/(2.0%-1.5%)*50%); customer score 75.0% (=50%+(74-70)/(78-70)*50%); weighted score 92.75% (=50%*112.5%+20%*70%+30%*75%); bonus USD 14,840 (=16000*92.75%). Design questions: comparable sales growth is most exposed to luck; compare with stores in the same region or adjust for local market growth. The scorecard does not measure staff development, safety or compliance; the multitask result predicts that those tasks will be neglected, so add a qualitative gate or lower the weight on sales. On a below-threshold measure, a gate protects the measure's importance; a weighted sum without gates keeps the bonus motivating when one measure is hit by bad luck.

Exercise C:

• C1: $b^* = 1/(1 + 0.5 \times 1.2^2) = 0.5814$ (=1/(1+0.5*1.2^2)); total surplus 0.2907. At $b = 1$ the surplus is 0.1400 (=1-0.5-0.5*1.2^2/2), less than half, because the agent bears the full risk premium. Full ownership is efficient only when noise is low or the agent is close to risk neutral.
• C2: (a) each existing option is worth 2.02, the grant 230,723; (b) a new at-the-money option is worth 2.89; (c) a value-neutral exchange gives 79,849 new options (about 1.43 old for 1 new); (d) a one-for-one repricing adds about 99,989 of fair value (=114453*(2.8895-2.0159)); (e) for: underwater options no longer retain or motivate, and the executive may leave; against: shareholders bore the fall in price and the executive is protected from it, which weakens the link between pay and performance and invites expectations of future repricing.
• C3: the plan pays for signed value, not for margin, collection or customer fit, so expect discounting, oversized or poorly qualified deals and neglect of renewals; no base salary loads all risk on the salesperson, so expect high turnover and a preference for safe, small deals or, at the other extreme, risky promises; no cap invites windfalls on deals won by others. Changes: a base salary with a 60:40 or 70:30 mix, commission on margin or collected revenue, payment split between signature and go-live, clawback on cancelled contracts, and team measures for large deals.

Self-test:

1. Risk (the employee bears noise and must be paid a premium) and distortion (effort moves to measured tasks and away from unmeasured ones, and measures can be gamed).
2. It falls, because $r$ is in the denominator: a more risk-averse agent needs a larger premium for each unit of risk, so the efficient contract shifts pay towards salary.
3. Lower the incentive on the measured task, or find a measure for the unmeasured one, so that the agent does not abandon the unmeasured task.
4. =80000+7.5%*(1.2-1)*1600000 = 104,000.
5. A salesperson just below 50% gains 40,000 by crossing it, so deals are pulled forward or discounted; one well below it stops trying. Replace the cliff with a ramp (for example a lower rate from 30% rising to 5% at 50%).
6. Incentive (the same workers worked harder) and selection (the plan attracted and kept more productive workers).
7. The executive's wealth rises by USD 1.20 when shareholder value rises by USD 1,000.
8. Each option is worth less than a share (5.24 against 10.00), so the same value needs more options.
9. At about USD 21.02 (=114453*10/(114453-60000)).
10. No. The amount is the incentive compensation received minus what would have been received on the restated numbers, computed without regard to taxes paid.
11. Malus reduces or cancels an award before it vests or is paid; clawback recovers pay already delivered.
12. Executives influence the board that sets their pay; directors have social and financial ties to the chief executive; the main constraint is outside "outrage" rather than bargaining, so pay is camouflaged in forms that are less visible to shareholders.

Revision history

Standard and licence note

Based on the agency theory and personnel economics literature listed above (Kerr, Jensen and Meckling, Holmstrom, Lazear and Rosen, Jensen and Murphy, Holmstrom and Milgrom, Gibbons, Lazear, Bebchuk and Fried, Hall and Murphy) and the Black-Scholes model, with dated rules taken from the official pages named in the sources and checked on 2026-10-30. The text, the plans, the numbers and the exercises are original; no textbook, survey, consultancy material or published case is copied or paraphrased, and Corvane Robotics, Northfield Outdoor Co. and every number in this card are fictional. The template is independent and not affiliated with or endorsed by any regulator, school, employer or publisher. It is not legal, tax, accounting or financial advice: check the current official texts and a qualified adviser before you design or approve a real plan. Template licence: MIT.