Intangibles, Provisions & the Relevance-Reliability Trade-off
Financial Performance & Management Control - TUHH Institute of Management Accounting & Simulation, Hamburg · part of my Technology Management MBA · study notes for revision.
The last chapter dealt with the assets you can kick: plant, machines, inventories, and the depreciation and impairment machinery around them. This chapter goes to the awkward corners of the balance sheet. First, the things a company owns but cannot touch - software, patents, brands, a footballer’s registration. Second, the things it owes but cannot yet pin a firm date or amount on - lawsuits, restructurings, warranty claims. Both force accounting to answer the same uncomfortable question: how much do we trust an estimate before we let it into the statements?
IFRS answers with a compromise. It wants the statements to be relevant (show what actually drives value) and reliable (show only what can be measured with confidence). For intangibles that produces a strict gate that keeps most self-made value out; for provisions it produces an expected-value logic that lets uncertain obligations in, as long as they are probable and estimable. Hold that tension in your head and every rule below stops looking arbitrary.
1 · Why intangibles matter - the market sees what the balance sheet misses
Section titled “1 · Why intangibles matter - the market sees what the balance sheet misses”Start with a puzzle from the Fortune 500 ranking of 1999. Two companies, ranked on five measures. One is a classic car maker, the other a software house.
| Rank (Fortune 500, 1999) | Old-economy car maker (GM) | Software house (Microsoft) |
|---|---|---|
| Sales | 1 | 109 |
| Assets | 12 | 126 |
| Net assets | 28 | 24 |
| Profits | 29 | 11 |
| Market value | 42 | 1 |
The car maker is number one in sales and sits on a mountain of assets, yet the stock market ranks it only 42nd. The software house is 109th in sales and 126th in assets, yet the market values it above every other company. What the market is paying for - code, know-how, network effects, people - barely appears in the books.
The same gap shows up whenever one company buys another. Compare the price paid with the book value of the acquired equity:
| Acquisition (late 1990s, in DM) | Purchase price | Book equity | Market-to-book |
|---|---|---|---|
| Telecoms group buys a US mobile operator (Telekom / Voicestream) | 106 bn | 10 bn | about 1,000 % |
| Mobile-network takeover (Vodafone / Mannesmann) | 350 bn | 46 bn | about 760 % |
| Car-maker merger (Daimler-Benz / Chrysler) | 190 bn | 51 bn | about 370 % |
| Steel merger (Thyssen / Krupp-Hoesch) | 36 bn | 12 bn | about 300 % |
Buyers paid three to ten times what the books said the equity was worth. The conclusion the slides draw is blunt: accounting is still built around tangible assets, even though the intangible ones - brands, know-how, customers, licences, talent - seem to be what the market actually finds value-relevant. The rest of this chapter explains why the rules are so cautious anyway.
2 · What counts as an intangible - the IAS 38 definition
Section titled “2 · What counts as an intangible - the IAS 38 definition”Under IAS 38 an intangible asset is an identifiable, non-monetary asset without physical substance. All three words carry weight:
Plenty of “untouchable” things are deliberately handled by other standards, so IAS 38 does not cover them: intangibles made for customers (that is inventory, IAS 2), leases (IAS 17), monetary and financial assets (IAS 39), goodwill (IFRS 3) and exploration rights (IFRS 6). When you meet goodwill later in this chapter, remember it lives outside IAS 38.
3 · Recognition - bought vs self-made, research vs development
Section titled “3 · Recognition - bought vs self-made, research vs development”The general entry test is the one from the tip box: the company must probably benefit from the item and its cost must be reliably measurable. For a purchased intangible (a licence, a patent, a player registration) that is easy - there is an invoice. The trouble starts with self-made intangibles, and here IAS 38 splits the life of an idea into two phases:
The hurdles in IAS 38.57 boil down to three questions the company must be able to answer with evidence:
- Can we build it? Technical feasibility of completing the intangible must be demonstrable.
- Will we finish it? There must be the intention and the financial and other resources to complete it and then use or sell it.
- Is anyone going to want it? A market for the output, or its internal usefulness, must be demonstrable, so that the future benefits are probable.
If any of these fails, the development spend is expensed like research. And some self-made items are banned outright, however valuable they are: internally generated brands, customer lists, training expenditure and internally generated goodwill never become assets. The reasoning is reliability, not relevance - nobody doubts a strong brand is worth something, but nobody can measure the cost of building it in a way an auditor could verify.
4 · After recognition - the cost model or the revaluation model
Section titled “4 · After recognition - the cost model or the revaluation model”- Carry at cost less amortisation over the useful life, less any impairment
- Indefinite useful life (no foreseeable end to the benefit) means no amortisation, impairment test only
- Same machinery as for tangible assets in chapter 5
- Allowed only where an active market exists for the intangible (IAS 38.72 onwards), which is unusual
- Carried at fair value; the uplift goes to a revaluation surplus in equity, not through profit - the “neutral” version
- Still amortised and impairment-tested from the revalued amount
5 · A real policy note - how a pharma group applies IAS 38
Section titled “5 · A real policy note - how a pharma group applies IAS 38”The notes to Bayer’s 2006 consolidated statements show what the rules look like in practice, and the honesty is instructive:
- Research = costs of current or planned investigations aimed at new scientific or technical knowledge. Never capitalised. Development = costs of applying research findings or specialist knowledge to production, processes, services or goods before commercial production starts. Capitalised if, and only if, narrowly defined conditions are met, essentially that the future economic benefits are sufficiently certain to cover the development cost.
- The punchline: because development projects depend on regulatory approval and other uncertainties, the conditions are normally not satisfied before approval is received. So almost all pharmaceutical development is expensed too.
- What counts as R&D spend: direct personnel and material costs plus related overheads of the labs and engineering departments, experimental and pilot facilities (including depreciation of the buildings used), clinical research, regular fees for using third-party patents, taxes on research facilities, and filing fees for self-generated patents that are not capitalised.
- One area is capitalised: the application-development phase of in-house software, amortised over the software’s useful life from the date it is placed in service.
6 · Discussion cases - a player, a protein and a brand
Section titled “6 · Discussion cases - a player, a protein and a brand”Three short cases from the slides test whether the recognition logic has sunk in:
- Not the person - the club buys the registration right, a contractual, identifiable, non-monetary right
- Cost is reliable (the transfer fee), benefit probable → intangible asset
- Amortised over the 3-year contract, one third per year; impaired if injury or loss of form destroys the value
- Scientists are convinced it matters, but further basic research and development are still needed before any drug exists
- This is the research phase → expense, no asset
- Highly relevant, hopelessly unreliable - reliability wins
- Because the brand was bought in a business combination it can be measured → recognised at 6 m, even though B itself could never show it
- The unexplained 4 m is goodwill (IFRS 3): impairment-tested, not amortised
- The rules trade relevance for reliability: bought intangibles in, self-made ones mostly out
- HGB is criticised for its many explicit options; IAS 38 hides implicit options - management decides when research becomes development and whether the 38.57 tests are met
That last point is the one examiners like. Replacing explicit choices with judgement calls does not remove discretion; it moves it into the definitions. Two identical software projects can produce different balance sheets depending on how early management declares the development phase open.
7 · Provisions - IAS 37
Section titled “7 · Provisions - IAS 37”Flip to the other side of the balance sheet. A provision is a liability of uncertain timing or amount (IAS 37.10). It is still a liability - a present obligation to a third party expected to drain resources - just one where you do not yet know exactly when or how much. Recognise it when all three conditions hold:
- There is a present obligation from a past event. Its existence may be uncertain, but it must be more likely than not (IAS 37.15).
- An outflow of resources to settle it is probable.
- A reliable estimate of the amount can be made (IAS 37.14). If everything else holds but no reliable estimate is possible, the item goes to the notes instead.
The slides compress this into a probability grid. Read the row (is there an obligation?) against the column (will it cost anything?):
| Obligation ↓ · Outflow → | Outflow probable | Outflow possible | Outflow remote |
|---|---|---|---|
| Probable (more likely than not) | Provision on the balance sheet (notes only if no reliable estimate) | Notes: contingent liability | nothing |
| Possible | Notes: contingent liability | Notes: contingent liability | nothing |
| Remote | nothing | nothing | nothing |
Two boundary rules follow directly from the liability definition. No third party, no provision: because a liability needs someone to be obliged to, IFRS refuses provisions for purely internal future expenses, a clear break with HGB, which allows provisions for certain internal expenses such as deferred maintenance. And pensions are not here: obligations under defined-benefit pension plans are regulated separately in IAS 19.
8 · Provision or not? The seven test cases
Section titled “8 · Provision or not? The seven test cases”| Situation | Verdict | Why |
|---|---|---|
| Bank credit | Liability, not a provision | Amount and timing are fixed by contract; nothing uncertain |
| Product-liability litigation, neutral experts expect no compensation | No provision | Outflow not probable; disclose as contingent liability if it is at least possible |
| Product-liability litigation, neutral experts expect probable compensation | Provision | Present obligation, probable outflow, amount can be estimated |
| General overhaul every four years, next one in 2017 | No provision | No third party is owed anything; the company could simply skip it. HGB thinks differently |
| Restructuring needed, detailed plan exists and has been communicated to employees | Provision | The announcement creates a valid expectation in others, so a constructive obligation exists |
| Management merely intends to restructure, no plan | No provision | An intention is not an obligation; management can still change its mind |
| Defined-benefit pension plan | IAS 19, not IAS 37 | Pension obligations have their own standard |
9 · Measuring a provision - best estimate means expected value
Section titled “9 · Measuring a provision - best estimate means expected value”IAS 37.36 asks for the best estimate of the amount needed to settle the obligation, which the slides equate with the expected value: weight every outcome by its probability and add up. If the time value of money is material, discount the result to a present value. The two slide examples (worked through in full below) give 60 where HGB would book 50, and 75 where HGB would book 100. Notice that the expected value is neither the most likely outcome nor the worst case. IFRS is aiming for neutrality; HGB is aiming for prudence. Same facts, different numbers, purely because of the philosophy behind the standard.
10 · Contingent liabilities - the footnote zone
Section titled “10 · Contingent liabilities - the footnote zone”When an obligation is only possible, or an outflow is not probable, the item is a contingent liability and is described in the notes rather than booked. Bayer’s 2005 notes are the slide’s example: contingent liabilities of 177 million, all from liabilities assumed on behalf of third parties, described as potential future obligations whose existence is uncertain at the balance-sheet date. The note then walks through the Cipro litigation: 39 putative class actions, one individual suit and one consumer-group suit (dismissed) filed in the United States since July 2000, alleging that a 1997 patent settlement with a generics maker breached antitrust law by delaying generic ciprofloxacin, with claims for triple damages. The company points out that the patent later survived re-examination and court challenges and has since expired. Everything a reader needs to judge the risk is there, but no provision was recognised, because compensation was not judged probable.
11 · Lessons learned - the trade-off in one place
Section titled “11 · Lessons learned - the trade-off in one place”- Recognise an item when the benefits are probable and it can be measured reliably. That test cuts out internally generated goodwill, research, brands and customer lists: relevance vs reliability in its purest form. The same trade-off sits inside development costs: capitalising them is more relevant, expensing them is more reliable, and IAS 38 sits uneasily in between.
- Amortisation methods are the same as for tangibles: straight-line, reducing balance and usage-based. Straight-line dominates in practice, except where reducing balance is chosen to accelerate tax deductions.
- Assets sometimes lose value in ways systematic amortisation cannot capture. Then they are written down, usually to value in use based on discounted cash flows.
Worked example
Section titled “Worked example”(a) The software project - three balance sheets from one set of facts
Section titled “(a) The software project - three balance sheets from one set of facts”Facts from the slides: research costs 1,000 and development costs 2,000 in period 01, the IAS 38.57 conditions are satisfied, the fair value on an active market at 31 December 01 is 2,200, and use starts in period 02 (so no amortisation yet). Deferred taxes ignored.
(b) Two provisions measured as expected values
Section titled “(b) Two provisions measured as expected values”Rule
best estimate = Σ pᵢ × amountᵢ (present value if the time value of money is material)
Case 1
0.2 × 100 + 0.8 × 50 = 20 + 40 = 60 → IAS 37 books 60; HGB books 50, the single most likely outcome
Case 2
0.5 × 100 + 0.5 × 50 = 50 + 25 = 75 → IAS 37 books 75; HGB books 100, a coin toss so prudence takes the higher figure
(c) Splitting a 20 m acquisition price
Section titled “(c) Splitting a 20 m acquisition price”Case corner
Section titled “Case corner”The session closes by pointing all of this at a real annual report, Borussia Dortmund’s, with five questions: how are players and their value handled and is that meaningful; how is player value depreciated; what other significant assets does the club have and how far do they appear on the balance sheet; what does the P&L reveal about the business model and where is extra information found; and what is generally odd about accounting for a sports club? My sketch of the answers:
Players as intangible assets. The club’s intangible assets consist of purchased player registrations plus some software. A registration is capitalised at cost, which includes the transfer fee and the directly attributable adviser costs, and is then amortised straight-line over the term of the individual player’s contract. If a player’s value collapses (a serious injury, a loss of form, a contract that will not be renewed) the registration is written down; in the 2015/16 season write-downs on intangible assets were large enough to push the first half-year into a net loss even though revenue had grown. Amortising over the contract term is defensible - the right expires when the contract does - but it makes the book value track the calendar, not the player’s form or market price.
What the balance sheet cannot see. The rules apply the IAS 38 gate ruthlessly. A player who came through the club’s own academy cost no transfer fee, so there is no reliable cost and no asset, however valuable he is. The same goes for the brand and the fan base: internally generated, therefore banned, even though the club itself names brand reach as its only non-financial performance indicator and admits it cannot be measured. So the balance sheet shows the bought squad and the stadium, while the things that actually generate the cash are largely missing. Two clubs with identical squads can show wildly different asset totals depending on how many players were bought rather than raised.
What the P&L reveals. The revenue lines tell the business model better than the balance sheet does: TV marketing (broadcasting rights), advertising (sponsoring), match operations (matchday income), merchandising, conference and catering, and transfer deals. Transfer income is the interesting one: it is lumpy, tied to sporting success and squad planning, and it is the only place where the value of a home-grown player ever becomes visible, on the day he is sold. The general oddity of accounting for a football club is exactly this: the accounting is technically correct and yet the balance sheet is a poor guide to what the club is worth, because its value sits in people, a brand and an audience that the reliability test keeps outside.
Key terms
Section titled “Key terms”| Term | What it means in plain words |
|---|---|
| Intangible asset | An identifiable, non-monetary asset without physical substance, such as a licence, patent, software or a player registration (IAS 38) |
| Identifiable | Separable from the business, or arising from a contract or legal right; the test that keeps goodwill out of IAS 38 |
| Research phase | Original, planned investigation aimed at new knowledge; always expensed |
| Development phase | Applying knowledge to a plan or design before commercial production; capitalised only when the IAS 38.57 criteria are met |
| IAS 38.57 criteria | Technical feasibility, intention and resources to complete, and a demonstrable market or usefulness |
| Internally generated brand / customer list / goodwill | Self-made intangibles that IAS 38 explicitly refuses to recognise, whatever they are worth |
| Cost model | Cost less amortisation less impairment; impairment-only for indefinite-life intangibles |
| Revaluation model | Fair value carrying amount, allowed only with an active market; uplift goes to a revaluation surplus in equity |
| Goodwill | What an acquirer pays above the fair value of identifiable net assets; IFRS 3, impairment-tested, not amortised |
| Provision | A liability of uncertain timing or amount (IAS 37.10), recognised when the obligation is more likely than not, the outflow probable and the amount estimable |
| Constructive obligation | An obligation created by a communicated plan or established practice that gives others a valid expectation, such as an announced restructuring |
| Contingent liability | A possible obligation, or a present one that is not probable or not estimable; disclosed in the notes, not booked |
| Best estimate / expected value | Probability-weighted average of the outcomes, discounted if the time value of money matters |
| Relevance vs reliability | The trade-off behind every rule here: useful information versus verifiable information |
Test yourself
Section titled “Test yourself”- A car maker ranks first in sales and twelfth in assets but only 42nd in market value, while a software firm ranks 109th in sales and first in market value. What does this say about the balance sheet, and what evidence from large acquisitions points the same way?
- A company spends 1,000 on research and 2,000 on development of a software product in year 01. The IAS 38.57 conditions are met and an active market prices the product at 2,200 at year-end. State the intangible asset and the net income under HGB, the IFRS cost model and the IFRS revaluation model.
- A lawsuit will cost either 100 (probability 0.2) or 50 (probability 0.8). What provision does IAS 37 require, what would HGB show, and how does the answer change if the probabilities are 0.5 and 0.5?
- Provision or not under IFRS: (a) a general overhaul due in four years, (b) a restructuring plan already announced to staff, (c) a lawsuit where independent experts expect no payout, (d) a defined-benefit pension plan.
- Company A pays 20 m for company B, whose net tangible assets are worth 10 m and whose brand is valued at 6 m. How is the brand treated in A’s consolidated statements, what is the remaining 4 m, and why could B never have shown that brand itself?
- Why does a club that buys a player for 40 m on a 3-year contract show an asset, while an equally good academy graduate is not on the balance sheet? Name one consequence for comparing clubs, and one option that HGB makes explicit but IAS 38 leaves to management judgement.
Revision summary
Section titled “Revision summary”Next: Financial Statement Analysis → - turning the statements into a judgement about the business.