Residual Income Valuation for Intangible-Heavy Firms

Applying the residual income model to firms heavy in intangible assets—tech companies, pharmaceuticals, biotech—requires first fixing a major accounting distortion. Intangible investments like R&D, software development, and clinical trials are expensed, not capitalized, so book value of equity is artificially depressed. This understates true economic book value and inflates ROE as a ratio. Before calculating residual income, the analyst must restate both book equity and earnings to reflect a capitalized intangible asset base, otherwise the valuation will be misleading.

The expensing problem and its impact on valuation

Under current accounting standards (ASC 730 for research and development, ASC 430 for software), most intangible investments hit the income statement as period expenses rather than balance-sheet assets. This made sense in an industrial economy where R&D was a small cost of doing business. In a tech, pharma, or biotech firm, it’s absurd: a company’s principal asset—the intellectual property created by R&D spending—is invisible on the balance sheet.

The consequences for residual income valuation are serious:

1. Book value is understated. A tech firm may have 20–30 years of capitalized R&D sitting nowhere on the balance sheet, yet it is a real economic asset generating current and future profits.
2. ROE is overstated. Because earnings are depressed by R&D expenses and book value is depressed by missing capitalized R&D, the ratio (net income ÷ book equity) is higher than the true economic return on capital deployed.
3. Residual income is distorted. If ROE is too high and cost of equity is fixed, the spread appears wider than reality, inflating the estimated intrinsic value.

Example: A software firm reports $100 million net income after $200 million R&D expense, on book equity of $500 million. Reported ROE = 20%. But suppose true economic earnings (R&D capitalized) are $250 million, and true economic book equity is $1.5 billion. True economic ROE = 16.7%. The firm looks far more profitable on reported figures than it truly is.

Restatement framework: Capitalizing R&D

The fix is to restate the historical income statement and balance sheet as if R&D (or other expensed intangibles) had been capitalized, amortized, and accumulated. Here’s the mechanics:

Step 1: Estimate amortization period

Intangible assets lose value over time. R&D might amortize over 5–10 years (a drug that loses exclusivity in 7 years, or software that is obsolete in 8 years). Software development might amortize over 3–5 years. This is not a known fact; it’s an assumption. Shorter amortization periods reflect higher obsolescence risk.

Step 2: Build a capitalized R&D asset stock

Reconstruct historical R&D spending back 5–10 years. For each year’s R&D outlay, track it as an accumulated asset, decaying at the assumed amortization rate. The ending intangible asset balance is:

$$Intangible\ Asset = \sum_{i=1}^{n} R&D_i \times (1 - \text{decay rate})^{n-i}$$

where n is the current year and decay rates reflect the assumed useful life.

Step 3: Adjust reported equity and earnings

• Add the intangible asset stock to reported book equity.
• Subtract the annual amortization of capitalized R&D from reported net income.

The adjusted ROE is:

$$Adjusted\ ROE = \frac{Reported\ NI - Amortization\ of\ R&D}{Reported\ BV + Capitalized\ R&D\ Stock}$$

Step 4: Recalculate the residual income spread

With adjusted earnings and adjusted book equity, recalculate the ROE-minus-cost-of-equity spread. This is the spread to use in the RI model going forward.

Worked illustration: A biotech firm

Suppose a biotech firm reports:

• Net income: $50 million (after $150 million R&D expense)
• Book value: $200 million
• Reported ROE: 25%
• Cost of equity: 12%
• Reported spread: +13 percentage points

Assume R&D has a 7-year useful life. Reconstruct R&D spending for the past 7 years (say, $120M per year). The capitalized R&D stock is approximately $600 million (at mid-point of a 7-year accumulation). Annual amortization is $150M ÷ 7 = $21.4 million.

Adjusted figures:

• Adjusted net income: $50M + $21.4M = $71.4M (add back amortization, which is not a cash charge but an intangible depreciation)
• Adjusted book equity: $200M + $600M = $800M
• Adjusted ROE: $71.4M ÷ $800M = 8.9%
• Adjusted spread: 8.9% − 12% = −3.1 percentage points

The biotech is destroying value on an economic basis, not creating it, even though reported ROE is 25%. This is a profound difference and will reshape the valuation dramatically.

Sensitivity to amortization life

The restatement is highly sensitive to the assumed amortization period. Shorter amortization (e.g., 5 years instead of 7) yields a larger intangible asset stock and higher required amortization charge. Longer amortization (10 years) yields a smaller charge and higher adjusted ROE.

Industry benchmarking and peer analysis help here. How long do peers’ products and technologies typically generate excess returns? What is the patent life? In pharma, patent cliffs are well-defined (often 20 years from filing, but market exclusivity is much shorter). In biotech, the useful life of IP is more uncertain.

Rather than pick a single amortization life, run the RI model across a range (e.g., 5–10 years for R&D, 3–7 years for software). This shows how sensitive the valuation is to this assumption. If intrinsic value is robust across the range, confidence in the result is higher. If it flips from value creation to destruction with a small amortization-life change, the valuation is fragile.

Challenges with forward-looking intangibles

Restatement addresses past R&D. But a forward-looking residual income model must also handle future R&D spending. There are two approaches:

Approach 1: Capitalize forward R&D in the forecast

Project future R&D spending (as a percent of revenue or an absolute dollar amount), then treat it as a capitalized intangible asset added to book value each year, with parallel amortization offsets to earnings. This keeps the intangible asset stock in the forecast consistent with the historical restatement and ensures that growing R&D spending results in a growing intangible asset base, not an implausible jump in book value depletion.

Approach 2: Use economic earnings

Define earnings as economic earnings, free of accounting noise. Start with operating cash flow (which includes R&D as a cash outflow), adjust for changes in intangible assets and other long-term working capital, and use that as a basis for residual income. This is more complex but avoids the need to forecast amortization schedules.

Most practitioners use Approach 1 because it’s more transparent and keeps the RI model aligned with the historical restatement.

Special case: Software and SaaS companies

Software companies and SaaS (software as a service) firms present a variant problem. Development costs are often expensed under ASC 430, but the software asset (the code, the platform) is perpetual or very long-lived. Traditional amortization periods (5–7 years) can be too aggressive.

For a SaaS firm with stable recurring revenue and modest product iteration, a 10–15 year amortization period may be more appropriate. The software platform doesn’t become obsolete quickly; it evolves incrementally.

Additionally, SaaS firms sometimes capitalize certain implementation and cloud-hosting costs, creating a hodgepodge of capitalized and expensed development costs on the balance sheet. The analyst should segregate true capitalized software assets from everything else and then estimate what would have been capitalized had all development been capitalized.

Interaction with intangible assets already on the balance sheet

Some firms do capitalize certain intangibles—goodwill from acquisitions, separately identified acquired intellectual property, capitalized software. These are already on the balance sheet and included in book value. Do not restate them twice.

The restatement should focus on expensed intangibles: R&D, internally developed software, and organizational and training costs. Acquired intangibles (goodwill, acquired patents, trade names) are already capitalized and should be left in place.

A careful analyst will look at the notes to the financial statements, identify what is capitalized, and adjust only the gap—the material expensed amounts that should be capitalized for an RI valuation.

Terminal value implications

When calculating terminal value in a multi-stage RI model, the assumption about long-term R&D spending is critical. If the firm is assumed to grow at 3% perpetually and is not expected to increase R&D intensity, then the intangible asset stock should stabilize (amortization roughly equals new R&D capitalization). If R&D is expected to grow faster than the business, the intangible asset stock will expand, and the RI forecast should reflect that explicitly.

See also