FSA 2 – Employee Compensation_ Post-Employment and Share-Based
For a great many companies, paying people is the single largest operating cost. Compensation is also a strategic instrument: the package a company offers determines whom it can attract, whom it keeps, and what behaviour it rewards. Because recruiting and training replacements is expensive, retention has a direct effect on reported margins. An analyst who does not understand how a company pays its workforce does not understand its cost structure.
Compensation reaches employees in many forms, from cash wages and commissions through to medical cover, life assurance and shares. The mix is set in the market for human capital and varies with the role, with local labour law and with industry custom. Throughout this lesson the word employee is used broadly, covering contract workers and members of the board as well as salaried staff, because the accounting treats all sources of human capital in the same way.
The five categories
Accounting standards sort compensation into five categories. The two dimensions that separate them are the length of time between the employee performing the service and the employer paying for it, and the form in which payment is eventually made.
| Category | What it covers | Typical items |
|---|---|---|
| Short-term benefits | Amounts due to be settled inside 12 months. | Wages and salaries; yearly bonuses; benefits in kind such as medical cover; social security scheme contributions; paid holiday |
| Long-term benefits | Amounts due to be settled more than 12 months out. | Extended paid leave, a sabbatical for instance; disability cover of long duration |
| Termination benefits | Amounts payable when employment ends. | Severance; ongoing medical and other benefits in kind; career counselling and outplacement help |
| Share-based compensation | Pay delivered in employer shares, or measured against them. | Restricted stock; stock options |
| Post-employment benefits | Amounts payable once the employee has retired. | Pensions and lump sums for retirees; life cover and medical care in retirement |
Short-term benefits dominate the total for almost every company. The other four categories matter out of proportion to their size because they are where the measurement problems live.
One principle, three dates
IAS 19 Employee Benefits imposes a single organising principle across all of these categories: recognise the fair value of compensation as an expense in the period in which the employee provides the service. Under US GAAP the same idea is spread across several sections of the Accounting Standards Codification rather than gathered into one standard, but the principle is the same.
Three dates give the accounting its shape.
- Grant. The employer sets out the terms of the award and the employee accepts them. Nothing is recognised on the financial statements at this point, but for some forms of compensation this is the date on which the amount is measured.
- Vesting. The employee becomes unconditionally entitled to the compensation, which creates an obligation for the employer. Vesting normally runs alongside the delivery of service, so this is when expense is recognised.
- Settlement. The employer actually pays, in cash or in another form such as shares.
Short-term benefits, and the capitalisation exception
Short-term benefits are the simple case. Expense and a matching current liability are recognised as the compensation vests, which is normally the same moment the employee works. When the employer pays, cash leaves and the liability is removed. The cash outflow sits in operating activities.
There is one important variation. Some compensation costs are not expensed as incurred but capitalised into an asset, so that the charge to profit is deferred until the service is consumed. Manufacturing labour is the standard case: wages attach to inventories and only reach the income statement as cost of sales when the goods are sold.
A company hires an employee into the legal department on 1 January on an annual salary of SGD 82,200, paid every two weeks. The first payment date is 14 January. Separately, on the same date it hires an employee into the manufacturing division on an annual salary of SGD 102,200, also paid every two weeks with a first payment date of 14 January. The goods that the second employee helped to make are sold to customers on 3 April.
At grant on 1 January nothing is recorded. Over the vesting window of 1 to 14 January the company records general and administrative expense of +3,162 and accrued compensation of +3,162 on the balance sheet. At settlement on 14 January the income statement is untouched, accrued compensation is reduced by 3,162, and cash flow from operations shows an outflow of 3,162.
Grant on 1 January: nothing recorded. Vesting from 1 to 14 January: no income statement effect, because the cost is capitalised. The balance sheet shows inventories +3,931 and accrued compensation +3,931. Settlement on 14 January: accrued compensation is reduced by 3,931 and cash flow from operations shows an outflow of 3,931, with no income statement effect. Sale on 3 April: cost of sales +3,931 and inventories reduced by 3,931, with no cash flow effect.
The cash left the business on 14 January but the expense appears on 3 April. That gap is the whole point of the capitalisation model.
Why the other categories are harder
The same underlying model applies to share-based compensation and to post-employment benefits, but three features change the arithmetic.
| Feature | Short-term benefits | Share-based compensation | Post-employment benefits |
|---|---|---|---|
| Typical vesting period | Days or weeks | Years | Years, decades |
| Form of payment | Cash | Shares | Cash |
| Amount recognised over the vesting period | Undiscounted salary, wage and similar | Fair value, measured on the grant date | Present value of estimated future benefits |
Some companies pay share-based compensation that is settled in cash. Cash-settled awards are accounted for in the manner of short-term benefits, not in the manner described later in this lesson.
Where the expense is reported is a separate question from how it is measured. Most companies push compensation expense out to the functional line that matches the employee role, so that everything relating to research staff lands in research and development expense and everything relating to the sales force lands in selling, general and administrative expense. Termination benefits are the usual exception: because they normally arise inside a restructuring programme, they are often shown on a discrete line such as restructuring charges.
Why companies grant shares
Share-based compensation is typically a bonus layer for highly paid employees, managers and technical specialists. For senior executives at many listed companies it makes up the majority of total pay. Awards are made under formal plans that specify eligibility, the instrument used, the maximum number of shares issuable and the vesting conditions. Plans are approved by the board and, at exchanges such as the NYSE and NASDAQ, by a shareholder vote. Sea Limited, the NYSE-listed internet company based in Singapore, grants awards to employees and directors under its 2009 Share Incentive Plan, which shareholders have amended and re-approved several times. Companies frequently run more than one plan at once.
The advantages are real. Paying in shares aligns employee wealth with shareholder wealth and softens principal-agent conflicts. It lets employees share in value they help create, and it is often paired with minimum shareholding requirements for managers to reinforce that alignment. Multi-year vesting improves retention because leaving early forfeits unvested awards. Finally, no cash goes out at grant, which preserves liquidity. That matters most for young companies that could not otherwise outbid established employers for talent.
The disadvantages are equally real. An individual employee usually has little influence over the share price, so the reward may bear no relation to personal performance. Concentrated ownership can distort risk appetite in either direction. A manager who already has salary and reputation tied to one employer, and now holds shares as well, may avoid profitable but risky projects for fear of a price collapse. Option grants push the other way: because an option pays only on the upside, a manager holding options may take more risk than is appropriate, chasing short-term gains at the expense of long-term viability. And employees lose wealth when the share price falls or simply lags alternatives, which damages the retention benefit the plan was meant to deliver.
One caution about the no-cash-outlay argument. Issuing shares to employees has an implicit cash cost, because those same shares could have been sold to investors for cash. Many companies make the cost explicit by repurchasing shares in the open market to offset the dilution caused by employee issuance.
An employee stock option is a non-tradeable call option on the shares of the employer, normally struck at the money, meaning the exercise price equals the share price on the grant date. If the share price rises above the strike after the award vests and before it expires, the employee can exercise and capture the spread.
Measuring fair value
For restricted stock and RSUs, grant-date fair value is observable: it is the share price. For options it must be estimated. An option is worth its intrinsic value plus its time value. An option struck at or above the current price has zero intrinsic value, yet its time value can be large, so the notion that an at-the-money grant is free is simply wrong.
Companies use option pricing models, typically the Black–Scholes model or a binomial model, both covered elsewhere in the curriculum. Neither IFRS nor US GAAP mandates a particular model. The method chosen must be consistent with fair value measurement requirements, be grounded in established principles of financial economic theory, and reflect all substantive characteristics of the award.
Because the estimate depends on assumptions, companies must disclose the material inputs in the notes. The directional effects are worth memorising.
| Assumption | Effect of an increase on estimated fair value |
|---|---|
| Expected volatility | Increases |
| Estimated life of the option | Increases |
| Risk-free interest rate | Increases |
| Dividend yield | Decreases |
Volatility is the most subjective of the four and therefore the one an analyst should watch. Companies usually anchor it to something market-derived: the implied volatility on exchange-traded options over their own shares, or the historical volatility of their own share price. A volatility assumption that drifts downwards year after year while the shares themselves become more turbulent is a signal to look harder.
Settlement is where options differ
Beyond measurement, the second difference between RSUs and options is settlement. When an RSU vests, settlement happens automatically as the unit converts into common stock, and the only entry is a transfer within equity from the share-based compensation reserve to common stock and paid-in capital. When an option vests, nothing settles until the employee chooses to exercise, and that choice depends on the share price. On exercise the company receives the strike price for every option exercised, which is recorded as a cash inflow in financing activities.
Working from the same Equity Compensation Plan, Workflow Corporation awards its executives 25 million stock options on 1 January 20X1, which vest on 31 December 20X3. The strike is set at the money. On the day of the award the shares change hands at JPY 4,360 and each option is valued at JPY 1,288. Expiry falls seven years after the award.
One-third vests each year, so the annual charge is 32,200 ÷ 3 = JPY 10,733 million. In each of the three years the income statement carries general and administrative expense of 10,733 million, and equity rises by 10,733 million through the share-based compensation reserve. Cash flows are untouched, although under the indirect method the 10,733 million is added back in reconciling net income to cash flow from operating activities.
Note what is absent. The share price of 5,400 at exercise appears nowhere in the entries. It determines whether employees exercise, and how much they personally gain, but the cash inflow is fixed by the strike price and the equity entries are fixed by the grant-date fair value.
The shift from options to restricted stock
The other response to fair value expensing was a change of instrument. Companies moved away from options and towards restricted stock, particularly RSUs. In 2021 options featured in fewer than 50 percent of compensation packages for chief executives of S&P 500 companies, with boards preferring RSUs.
Accounting was only part of the reason. Employees often prefer RSUs, because provided the vesting conditions are met and the share price does not go to zero, an RSU retains some value even in a downturn, whereas an option can expire worthless. Underwater options were a widespread problem after the technology bubble burst in the early 2000s and again after the Global Financial Crisis. RSUs also align interests better, since the holder is exposed to downside as well as upside, while the asymmetric payoff of an option can encourage risk-taking that damages long-term performance. And RSUs are simpler: easier for employees to understand, more straightforward for personal tax, and they do not require the employee to find cash for an exercise price.
Post-employment benefits cover cash pensions and non-monetary benefits provided to retired employees. Practice varies enormously with local law and custom. In the United Kingdom companies are required by law to offer pension plans and to enrol employees automatically. The United States does not require employers to provide post-employment benefits but grants tax advantages to those that do. In countries where health care is government-sponsored, employer-provided retiree health cover is much less common.
Defined contribution plans
In a defined contribution plan the sponsoring employer pays agreed contributions into the plan, and employees may contribute too. Employees typically choose among designated investment options, usually mutual funds and exchange traded funds holding equities and bonds. Once the employer has made the agreed contribution, its obligation ends. It is not required to make further contributions, gains and losses on plan investments accrue to the employee, and the employee carries both the investment risk that the assets fall short and the actuarial risk of outliving them. Because the obligation is fixed and short-dated, forecasting it is easy.
The global drift has been towards defined contribution plans, particularly in the private sector, as employers shed risk. The pattern is not universal: in countries such as the Netherlands and Japan defined contribution plans remain rare.
Defined benefit plans
A defined benefit plan is a promise to pay a specified amount after retirement, either as a lump sum or as periodic payments until death. The amount usually comes from a formula built on years of service and pre-retirement compensation. A typical formula pays an annual amount equal to 1 percent of final-year salary multiplied by years of service. Plans normally set a service threshold for qualification, for example five years. Additional service earns additional benefit, but an employee who qualifies and then leaves in year six still keeps the entitlement earned.
Regulation usually requires employers to pre-fund these promises by setting assets aside in a separate legal entity such as a trust. Contributions are made to meet regulatory minimum funding levels or at the discretion of the sponsor. Plan assets are typically invested in bonds, equities, derivatives, cash and other assets, and the combination of contributions and investment returns funds the payments to retirees. In many jurisdictions contributions are tax deductible, so the timing of contributions is often a tax planning decision. A company in a jurisdiction that limits tax loss carryforwards, for instance, may choose to contribute only in years when it has positive taxable income.
Unlike the defined contribution case, the employer here bears the investment risk that plan performance disappoints and the actuarial risks attached to retirement ages, life expectancies and future salaries. Alongside the shift to defined contribution plans, many defined benefit plans have been closed, meaning new employees can no longer join, or frozen, meaning existing beneficiaries no longer accrue additional benefits so their future payments are fixed. Affected employees are usually moved onto defined contribution arrangements.
Other post-employment benefits, abbreviated OPEB, are defined benefit plans that deliver non-monetary benefits such as retiree life insurance and medical care. Regulation often does not require pre-funding of OPEB, partly because governments do not usually insure these benefits, partly because the liability is normally much smaller, and partly because such plans are easier to discontinue if costs become burdensome. Many OPEB plans are therefore unfunded, with the company simply paying benefits as they arise on a pay-as-you-go basis. That does not remove the obligation, and the sponsor still bears investment and actuarial risk.
| Type | Amount of benefit to the employee | Obligation of the sponsoring company | Pre-funding of the future obligation |
|---|---|---|---|
| DC plan | Not defined. Depends on contributions and the investment performance of plan assets. Investment and actuarial risks are borne by the employee. | The obligation, that is the contribution, is defined each period. It is typically paid periodically with no further future obligation. | Not applicable. |
| DB plan | Defined by the plan formula, often a function of length of service and final-year compensation. Investment and actuarial risks are borne by the company. | The future obligation, based on the formula, must be estimated in the current period. | Companies typically fund by contributing to a pension trust. Regulatory funding requirements vary by country. |
| OPEB, for example retiree health care | Depends on plan specifications and the type of benefit. Investment and actuarial risks are usually borne by the company. | Eventual benefits are specified. The future obligation must be estimated in the current period. | Companies typically do not fund OPEB obligations. |
Reporting for defined contribution plans
The financial reporting mirrors short-term benefits almost exactly. Contributions are recognised as expense, grouped into the relevant functional operating expense category rather than shown on a discrete line. The only balance sheet effect is a current liability for contributions that have vested but not yet been paid. Contributions are a cash outflow in operating activities. The plan itself is a separate legal entity with its own financial statements, so plan assets, plan liabilities and withdrawals by employees never appear on the accounts of the employer.
Every two weeks a company pays into a defined contribution plan an amount worth 5 percent of salary. One member of the legal team is on SGD 82,200 a year. Payments start on 14 January.
At grant on 1 January there is no financial statement impact, although the plan contribution is estimated at SGD 158. Over the vesting window of 1 to 14 January the income statement carries general and administrative expense of +158 and the balance sheet carries accrued compensation of +158. At settlement on 14 January accrued compensation is reduced by 158 and cash flow from operations shows an outflow of 158.
Set this beside Example 1 and the two are structurally identical. The employer promise is bounded by the contribution, so nothing more complicated is required.
Funded status: the balance sheet number
Under both IFRS and US GAAP, everything that is not explicitly structured as a defined contribution plan is accounted for as a defined benefit plan. That sweeps in OPEB and even informal post-employment arrangements. The balance sheet carries one number, the funded status.
Fair value of plan assets is the assets held by the plan, such as bonds, stocks, cash and derivatives, held exclusively to pay benefits and measured at the price that would be received in an orderly sale, using quoted market prices where available. Two features of plan assets matter for valuation later: they are the property of the plan and not of the sponsor, so once contributed they cannot be withdrawn, and they are legally isolated from the sponsor in bankruptcy.
Pension obligation is the present value, without deducting any plan assets, of the expected future payments required to settle the obligation arising from employee service in the current and prior periods.
A negative funded status means the plan is underfunded and is reported as a net pension liability. A positive funded status means the plan is overfunded and is reported as a net pension asset. This is one of the rare places where accounting standards allow a net rather than a gross presentation. The netting stops at the plan boundary, however: separate plans cannot be netted against each other, so a company that runs an overfunded pension plan and an unfunded OPEB plan will report both a net pension asset and a net pension liability. In practice the amounts may not be discrete balance sheet lines at all, appearing instead inside other non-current assets or other non-current liabilities with detail in the notes.
The discount rate applied to the obligation is the yield on investment grade corporate bonds denominated in the same currency as the benefits, or on government bonds where no liquid corporate bond market exists. Estimating the obligation requires a long list of actuarial assumptions covering salary growth, retirement dates and mortality, which is why IAS 19 encourages sponsors to engage a qualified actuary.
Recognising the employer contribution as the expense would breach accrual accounting, because the contribution is not the cost of providing benefits in that period. Contributions need not be made in the period the service is delivered. A sponsor may make no contribution for several years while the plan still meets its payments. Over those same years employees keep earning service, which raises the obligation, and retirement dates draw nearer, which raises the obligation again as the discount unwinds. The expense must track those movements, not the cash.
The IFRS decomposition
Under IFRS the periodic pension cost has three components, two in profit or loss and one in other comprehensive income.
- Service cost, in two parts. Current service cost is the increase in the pension obligation caused by employee service in the current period. Since a typical formula pays a percentage of final salary for each year of service, every additional year of service raises future payments. An actuary computes it using the projected unit credit method, the detail of which sits below the level presented to investment analysts. Past service cost arises when a plan amendment changes the obligation relating to service in prior periods. Under IFRS both are operating expenses in profit or loss, grouped with other compensation in the relevant functional category, so service cost for the sales force is expensed within selling, general and administrative expense.
- Net interest expense or income, the accretion of the obligation with the passage of time. It is the net pension liability or net pension asset at the beginning of the period multiplied by the discount rate. Under IFRS it sits below the operating income line, with other financing costs.
- Remeasurement, which has two parts: the difference between the actual return on plan assets and the amount already assumed in the net interest calculation, and actuarial gains and losses. Actuarial gains and losses arise from changes in assumptions such as the salary growth rate, the discount rate or mortality. An assumption change that increases the obligation is an actuarial loss; one that decreases it is an actuarial gain. Under IFRS remeasurements go to other comprehensive income and never touch earnings.
Two things are conspicuously missing from the expense. It contains no employer contribution and no benefit payment. It is a non-cash accrual driven by the change in the net pension liability or asset. Contributions appear on the statement of cash flows, normally in operating activities, which is also where the non-cash expense is added back under the indirect method. So for cash flow purposes defined benefit plans look like defined contribution plans: contributions are operating outflows.
Benefit payments from the plan to retirees do not appear on the accounts of the sponsor at all. The plan is a separate entity with its own statements, and in any case a benefit payment reduces plan assets and the obligation by the same amount, leaving funded status untouched.
At the start of 20X1 Workflow Corporation sets up a defined benefit pension plan for staff who qualify. The promised amount is 1 percent of salary in the final 12 months before retirement, multiplied by years of service, paid in cash. Recipients may take it as one sum on retiring or as monthly instalments. Workflow puts JPY 710 million into the plan at the start of 20X1, and the trustees place it mainly in fixed income and equity securities. The company reports under IFRS.
Income statement. Operating expense of 5 million for service cost. There is no interest component because the plan obligation at the beginning of the year is zero.
Statement of stockholders equity. Remeasurements of −21.3 million. The actual return on plan assets was −3% × 710 = −21.3 million, against an assumed amount of zero, so the whole of it is a remeasurement recognised in other comprehensive income.
Balance sheet. Cash falls by 710 million on the contribution. The net pension asset is 683.7 million: the opening funded status of 710 reduced by the 21.3 loss on plan assets and by the 5 of service cost, since 710 − 21.3 − 5 = 683.7. Equivalently, plan assets of 710 − 21.3 = 688.7 less an obligation of 5 gives 683.7.
Statement of cash flows. Cash flow from operating activities of −710 million, the contribution to plan assets.
Income statement. Operating expense of 9 million for service cost, and net interest income of 18.3 million, being the opening net pension asset of 913 multiplied by the discount rate of 2 percent. Because the plan is in surplus, the net interest line is income rather than expense, and under IFRS it sits below operating income.
Statement of stockholders equity. Remeasurements of 30.3 million. The actual return on plan assets was 5% × 1,010 = 50.5 million. The amount already captured through the interest calculation is 2% × 1,010 = 20.2 million. The difference, 50.5 − 20.2 = 30.3 million, is the remeasurement.
Balance sheet. Net pension asset of 952.6 million. Take the opening 913, add net interest income of 18.3, add the remeasurement of 30.3 and deduct service cost of 9: 913 + 18.3 + 30.3 − 9 = 952.6. Checking from the components, plan assets end at 1,010 + 50.5 − 5 = 1,055.5 and the obligation ends at 97 + 9 + 1.94 − 5 = 102.94, and 1,055.5 − 102.94 gives the same 952.6. Benefits paid of 5 cancel out of both sides, exactly as expected.
Statement of cash flows. No impact, because no contributions were made.
Where US GAAP differs
US GAAP matches IFRS on the balance sheet and on the statement of cash flows. The income statement and other comprehensive income are materially different, and the expense is split into five components rather than three.
- Current service cost, computed on the projected unit credit method as under IFRS and typically reported as an operating expense.
- Interest cost, the discount rate multiplied by the pension obligation at the beginning of the year. This is a gross interest expense rather than a net figure, and it is normally presented within interest expense below operating income.
- Expected return on plan assets, management estimate of a return rate applied to opening plan assets measured at fair value. It is not netted directly against interest cost in the manner of IFRS net interest, but recognised separately as an offset within earnings. The expected rate is usually derived from historical returns on asset classes and therefore depends on the asset allocation of the plan.
- Amortisation of past service cost. Past service cost goes first to other comprehensive income in the period the change occurs, then is amortised into the income statement over the average service lives of the affected employees.
- Amortisation of net gains or losses. Actuarial gains and losses, together with the difference between the expected and the actual return on plan assets, may go straight to profit or loss, but the more common choice is other comprehensive income with subsequent amortisation into earnings under the corridor approach.
The corridor approach compares the net cumulative unrecognised gains and losses at the start of the reporting period with the pension obligation and the fair value of plan assets at that date. If the cumulative unrecognised amount exceeds 10 percent of the greater of those two figures, the excess is amortised over the expected average remaining working lives of participating employees and included in periodic pension cost in earnings. The word corridor refers to that 10 percent band, and only the amount outside it must be amortised. The purpose is to smooth earnings against large swings in estimates or in plan asset values.
| IFRS component | IFRS recognition | US GAAP component | US GAAP recognition |
|---|---|---|---|
| Service costs | Recognised in profit or loss. | Current service costs | Recognised in profit or loss. |
| Past service costs | Recognised in OCI and subsequently amortised to profit or loss over the service life of employees. | ||
| Net interest income or expense | Recognised in profit or loss as net pension liability or asset multiplied by the discount rate. | Interest expense on the pension obligation | Recognised in profit or loss. |
| Expected return on plan assets | Recognised in profit or loss as plan assets multiplied by the expected return. | ||
| Remeasurements: net return on plan assets and actuarial gains and losses | Recognised in OCI and not in profit or loss. Net return on plan assets equals actual return less plan assets multiplied by the interest rate. Actuarial gains and losses are changes in the obligation arising from changes in actuarial assumptions. | Actuarial gains and losses including differences between actual and expected returns on plan assets | Recognised immediately in profit or loss or, more commonly, recognised in OCI and subsequently amortised to profit or loss using the corridor or a faster method. The asset difference equals actual return less plan assets multiplied by the expected return. |
Two comparisons are worth committing to memory. IFRS applies the discount rate to plan assets inside net interest, whereas US GAAP applies management estimate of the expected return. Management can therefore flatter US GAAP earnings by raising the expected return assumption, a lever that does not exist under IFRS. And IFRS parks remeasurements permanently in other comprehensive income, whereas US GAAP recycles them into earnings over time through the corridor, so US GAAP earnings carry an echo of past actuarial surprises that IFRS earnings do not.
What the disclosures give you
Defined contribution disclosure requirements are minimal. IAS 19 requires only the amount recognised as an expense, typically given in a note headed Employee Compensation, Post-Employment Benefits or similar. Shell plc, the Amsterdam-based integrated oil company, limits its defined contribution disclosure to the amounts recognised on the income statement for each of the last three years.
Defined benefit disclosure, including OPEB, is another matter. For sponsors of large plans this is often the longest note in the accounts. Under IAS 19 management must explain the characteristics of the plans and the risks attached to them, identify and explain the amounts in the financial statements arising from the plans, and describe how the plans may affect the amount, timing and uncertainty of future cash flows.
Shell plc provides retirement benefits in most of the countries where it operates, through funded and unfunded defined benefit plans and through defined contribution plans, with the most significant pension plans in the Netherlands, the United Kingdom and the United States, and OPEB comprising retirement health care and life insurance in certain countries. The narrative shows how varied the position can be inside one group. The principal Dutch plan, a funded career-averaged arrangement paying annuities, reported a surplus of $1,756 million at 31 December 2021 against a surplus of $405 million a year earlier; it was closed to employees hired or rehired after 1 July 2013 but remains open for ongoing accrual by existing active members, who represent 26 percent of the overall Dutch defined benefit liability against 31 percent in 2020. The three largest UK plans, funded final salary arrangements, together reported a surplus of $3,807 million at 31 December 2021 against a deficit of $76 million in 2020, after netting unfunded plans of $473 million reported within non-current liabilities. All three UK plans are closed to new hires, two remain open for ongoing accrual, and active members account for 20 percent of the overall UK defined liability against 23 percent in 2020. Crucially, those three plans are separate and independent and cannot be netted against one another. In the United States the principal funded final average pay plan reported a surplus of $182 million at 31 December 2021 against a deficit of $1,846 million in 2020, alongside an unfunded defined benefit plan with a deficit of $1,129 million against $1,475 million in 2020, with active members representing 24 percent of the funded plan liability against 25 percent in 2020. The unfunded US OPEB plans, which provide medical, dental, vision and life insurance benefits and share costs between the company and retirees, reported a deficit of $4,067 million at 31 December 2021 against $4,497 million in 2020, with the medical plan closed to employees hired or rehired on or after 1 January 2017.
| Item | 31 December 2021 | 31 December 2020 |
|---|---|---|
| Pension obligations | (107,336) | (115,792) |
| Plan assets | 104,495 | 102,678 |
| Effect of asset ceilings | (13) | (17) |
| Surplus (deficit) | (2,854) | (13,131) |
| Recognised as non-current assets | 8,471 | 2,474 |
| Recognised as non-current liabilities: pensions | (6,458) | (10,237) |
| Recognised as non-current liabilities: OPEB | (4,867) | (5,368) |
| Total recognised | (2,854) | (13,131) |
The aggregate deficit narrowed from 13,131 to 2,854 over the year. Note that a group-level net figure of 2,854 conceals a non-current asset of 8,471 and non-current liabilities of 11,325, because separate plans cannot be netted.
| Asset class | 31 December 2021 | 31 December 2020 |
|---|---|---|
| Equities | 32% | 33% |
| Debt securities | 57% | 57% |
| Real estate | 7% | 6% |
| Investment funds | 3% | 3% |
| Cash | 1% | 1% |
| Total | 100% | 100% |
Contributions to defined benefit pension plans were estimated at $900 million for 2022.
| Actuarial input | Value at 31 December 2021 | Sensitivity range | Change in the obligation |
|---|---|---|---|
| Growth rate applied to pensions already being paid | 2.0% | −1% to +1% | (9,908) to 12,171 |
| Discount rate applied to the pension plans | 2.0% | −1% to +1% | 18,954 to (14,599) |
| Assumed inflation | 2.1% | −1% to +1% | (10,691) to 13,325 |
| Life expectancy, men currently aged 60 | 87 | −1 year to +1 year | (1,946) to 1,937 |
| Life expectancy, women currently aged 60 | 89 | −1 year to +1 year | (1,863) to 1,972 |
The discount rate row runs in the opposite direction to the others: a one percentage point fall in the discount rate raises the obligation by 18,954, which is larger than any other single sensitivity in the table.
Two analytical habits follow. Track the disclosed assumptions over time and against peers. Management looking to flatter the numbers can use a high discount rate, a low salary growth rate, low life expectancy, a low inflation rate and, for medical OPEB, a low health care cost growth rate, all of which shrink the obligation and the expense. Under US GAAP there is one further lever: raising the expected return on plan assets.
Modelling a defined contribution plan is easy and normally done implicitly. Because such expenses are usually structured as a percentage of salaries and paid in cash, they share drivers with short-term benefits and the rest of operating expenses, so forecasting selling, general and administrative expense also forecasts the defined contribution expense for staff in those functions. Cash flow matches the recognised expense, and the only balance sheet effect is an accrued liability already captured by working capital ratios.
Defined benefit plans, including OPEB, need four items forecast explicitly: service cost, net interest expense or income, remeasurements, and employer plan contributions. Those four generate the income statement charge, the net pension asset or liability on the balance sheet, and the contribution outflow on the statement of cash flows.
There is a materiality threshold that saves a great deal of wasted effort. Where a company has small, well funded plans, meaning the net pension liability is no more than 5 percent of equity market capitalisation, and particularly where those plans are closed or frozen, detailed forecasting is not worth the time. The plan is not part of the investment case.
Two effects in a valuation
A valuation has to reflect the funded status and the future service cost.
The funded status is handled asymmetrically, and the asymmetry is not simply conservatism. An underfunded plan is treated as debt in the enterprise value calculation and in the bridge from enterprise value to equity value. An overfunded plan is ignored. The justification lies in the legal structure described earlier. The sponsor of an underfunded plan is obliged to pay the benefits whatever the shortfall, so the deficit is a genuine claim ranking alongside debt. But a surplus is an asset only in a nominal sense, because plan assets cannot be withdrawn and handed to shareholders or lenders. They exist solely to pay benefits. Several data providers follow this convention and include net pension liabilities in their debt and enterprise value figures.
Future service costs are a separate matter, because they are not inside the funded status at all. They are compensation the employee earns in future periods in place of short-term benefits. Treat them as you would treat share-based compensation: even though service cost is not a cash expense, deduct it from free cash flow, which in practice means leaving it expensed and not adding it back to EBIT when building free cash flow. This applies unless the plans are frozen, in which case no further benefits accrue from service and there is nothing to deduct.
Net interest expense or income should be excluded from the discounted cash flow model. It is the unwinding of the discount on an obligation already stated at present value. The valuation is being performed on a present value basis, and the present value of an underfunded plan has already been dealt with by deducting the net pension liability from enterprise value. Including net interest as well would count the same obligation twice.
Pulling the two halves of the lesson together
The two forms of compensation in this lesson look very different on the page and behave the same way in a valuation. Both are non-cash charges that a naive discounted cash flow model would ignore. Both represent genuine transfers of value away from existing shareholders. And in both cases the recommended treatment is the same: leave the expense in when computing free cash flow, and deal separately with the balance sheet consequence, dilution for share-based awards and the funded status deficit for defined benefit plans. The instinct to add back every non-cash charge is the single most common error in this material.