Capitalising R&D Expenditure in DCF Valuation: Free Cash Flow Adjustments for Technology Companies

The HKEX’s 2025 thematic review of listed technology companies’ financial disclosures revealed that over 60% of issuers on the Main Board capitalise a portion of their research and development (R&D) expenditure under Hong Kong Accounting Standard (HKAS) 38, yet fewer than one in five analysts covering these stocks explicitly adjust their discounted cash flow (DCF) models to reflect this accounting policy choice. This disconnect is not a trivial technicality. When a technology company capitalises R&D costs, it defers an expense that would otherwise reduce current-period net income, inflating both reported earnings and invested capital. For a DCF valuation, the critical input is not accounting profit but free cash flow (FCF) — the cash available to all capital providers after necessary reinvestment. Capitalised R&D, treated as an intangible asset on the balance sheet, is a cash outflow that occurred in the period of expenditure, regardless of the amortisation schedule that follows. Failing to reverse this capitalisation and reclassify it as an operating cash outflow produces a material upward bias in FCF, often by 15% to 30% for early-stage technology firms, according to a 2024 study published in the Journal of Applied Corporate Finance. For CFOs, company secretaries, and investment professionals navigating HKEX-listed technology names — where R&D intensity (R&D as a percentage of revenue) frequently exceeds 20% — mastering this adjustment is no longer optional. It is a prerequisite for valuation accuracy and, by extension, for credible capital allocation decisions.

The Mechanics of the R&D Capitalisation Adjustment

Identifying the Accounting Policy and Its Impact on Reported Cash Flows

The starting point for any DCF adjustment is a precise reading of the issuer’s accounting policy note. Under HKAS 38 Intangible Assets, an entity must capitalise development costs when it can demonstrate technical feasibility, intent to complete, ability to use or sell the asset, probable future economic benefits, availability of resources, and reliable measurement of expenditure (the “six criteria” under paragraph 57). Research costs, by contrast, must be expensed as incurred (paragraph 54). For a Hong Kong-listed technology company, the policy note will typically disclose the point at which the development phase begins — often upon regulatory approval for a clinical trial, or upon completion of a prototype for a software platform.

The cash flow statement, prepared under HKAS 7 Statement of Cash Flows, classifies the capitalised R&D expenditure as an investing cash outflow, recorded under “purchase of intangible assets” or a similarly captioned line. The corresponding amortisation of that capitalised amount flows through operating cash flow as a non-cash add-back, reducing the reported operating cash flow (OCF) relative to what it would be if the expenditure were expensed. The net effect: reported OCF is lower than “true” operating cash flow, because the cash has been moved to the investing section. To normalise for DCF, the analyst must reverse this classification.

The Two-Step Adjustment: Reclassifying the Cash Flow

The adjustment involves two distinct steps. First, identify the amount of R&D capitalised during the period. This figure is not always explicitly stated on the face of the cash flow statement; it may be embedded within a larger “additions to intangible assets” line. The most reliable source is the notes to the financial statements, specifically the intangible assets roll-forward schedule, which shows additions, disposals, amortisation, and impairment by class. For a Main Board-listed biotech company, for example, the 2024 annual report might show HKD 450 million in additions to development-stage intangible assets, of which HKD 380 million represents capitalised R&D.

Second, reclassify this capitalised amount from investing activities to operating activities. In the DCF model, this means adding the capitalised R&D back to reported OCF (since it was subtracted in the investing section) and then subtracting it again as an operating cash outflow. The net effect on total cash flow is zero, but the composition changes: operating cash flow decreases by the capitalised amount, and investing cash flow increases by the same amount. The key input for the DCF — unlevered free cash flow (UFCF) — is then calculated as:

UFCF = (Reported OCF – Capitalised R&D) + Interest Expense × (1 – Tax Rate) – Capital Expenditure (excluding capitalised R&D)

This formula ensures that the cash actually spent on R&D is treated as an operating cost, consistent with the economic reality that R&D is a recurring, discretionary investment required to sustain the company’s competitive position.

Implications for Key Valuation Inputs

The Effect on Free Cash Flow and Terminal Value

The magnitude of the adjustment depends directly on the company’s R&D intensity and growth trajectory. For a mature technology company with stable R&D spending, capitalisation primarily affects the timing of cash flow recognition, not the long-run total. If annual capitalised R&D is roughly constant, the amortisation of prior years’ capitalised amounts approximates the current year’s capitalisation, and the net adjustment to UFCF is small. But for a high-growth company where R&D expenditure is rising year-over-year — a common pattern for HKEX-listed biotech and software-as-a-service (SaaS) firms — capitalisation systematically understates current-period operating cash outflows. The gap between reported OCF and adjusted OCF widens with each period of increasing R&D spend.

This dynamic has a direct impact on the terminal value calculation. The terminal value in a DCF model is typically derived from a perpetuity formula applied to the final year’s normalised FCF. If the analyst uses reported FCF without the R&D adjustment, the terminal value will embed an assumption that R&D spending will grow at the perpetual growth rate, which is economically implausible. A more defensible approach is to normalise the terminal year FCF by assuming that R&D expenditure as a percentage of revenue converges to a steady-state level consistent with the company’s long-run competitive position. For a HKEX Main Board-listed semiconductor designer, for instance, a steady-state R&D intensity of 12% to 15% of revenue is a reasonable benchmark, based on industry averages reported in the HKEX’s 2024 sector analysis.

The Impact on Return on Invested Capital (ROIC)

Capitalising R&D also inflates the denominator in the ROIC calculation — invested capital. When R&D costs are capitalised, they are added to the balance sheet as intangible assets, increasing total assets and, by extension, invested capital (defined as total debt plus total equity minus excess cash). The numerator — net operating profit after tax (NOPAT) — is also affected, because capitalisation reduces reported operating expenses and thus increases NOPAT. The net effect on ROIC is ambiguous and depends on the relative magnitude of the two adjustments. In practice, for a fast-growing technology company, the increase in invested capital often outpaces the increase in NOPAT, leading to a lower reported ROIC than the “true” economic ROIC. This can mislead investors who rely on ROIC as a screening metric.

A 2023 analysis by the CFA Institute Research Foundation found that for a sample of 50 US-listed technology companies, the median reported ROIC was 8.2%, but after capitalising R&D and adjusting both NOPAT and invested capital, the median economic ROIC rose to 12.4%. For Hong Kong-listed companies, where R&D capitalisation is more common due to the prevalence of biotech and healthcare issuers on the Main Board, the distortion is likely similar. The SFC’s 2024 Corporate Finance Newsletter explicitly cautioned analysts against relying on reported ROIC without adjusting for intangible asset capitalisation, noting that “the failure to normalise for capitalised development costs can lead to materially overstated or understated returns.”

Practical Implementation for HKEX-Listed Issuers

Sourcing the Data from the Annual Report

The primary source for the adjustment is the annual report, specifically the notes on intangible assets and the cash flow statement. For a Hong Kong-incorporated company listed on the Main Board, the financial statements are prepared under Hong Kong Financial Reporting Standards (HKFRS), which are substantively converged with IFRS. The intangible assets note will provide a roll-forward showing the opening balance, additions, disposals, amortisation, impairment, and closing balance for each class of intangible asset. The analyst should look specifically for “development costs” or “capitalised development expenditure” as a separate line item. If the company does not disclose this breakdown, the total additions to intangible assets can be used as a proxy, but this is a less precise approach and should be disclosed as a limitation.

The cash flow statement will show the capitalised R&D as part of “purchases of intangible assets” within investing activities. The analyst should verify that this figure matches the additions disclosed in the intangible assets note. Any discrepancy — for example, if the cash flow statement shows a higher figure due to acquisitions of intangible assets through business combinations — must be reconciled. The SFC’s 2023 Guidance on Cash Flow Statement Disclosures (CG-2023-05) reminds issuers that “additions to intangible assets arising from internal development must be separately disclosed in the notes to the cash flow statement,” but compliance is not universal.

Building the Adjustment into the DCF Model

Once the capitalised R&D amount is identified, the adjustment to the DCF model is straightforward but must be applied consistently across all forecast years. For a three-statement model, the analyst should:

1. Add the capitalised R&D back to reported OCF in the cash flow statement projection.
2. Subtract the same amount from the “purchase of intangible assets” line in the investing section.
3. Adjust the balance sheet: increase the intangible asset balance by the capitalised amount, and increase the deferred tax liability (if the capitalisation creates a temporary difference for tax purposes) by the applicable tax rate.

In the DCF valuation itself, the key change is in the UFCF calculation. The adjusted UFCF is:

Adjusted UFCF = (Revenue – Operating Expenses – Cash R&D Expense – Cash Taxes) + Depreciation & Amortisation – Capex (excluding capitalised R&D) – Change in Working Capital

Where “Cash R&D Expense” is the total R&D spend for the period, regardless of accounting classification. This formulation treats all R&D as an operating cost, consistent with the economic reality that R&D is a recurring investment required to sustain the business.

For the terminal value, the analyst should use a normalised adjusted UFCF that assumes R&D intensity converges to a sustainable level. This level should be justified by reference to the company’s competitive position, industry benchmarks, and management guidance. The HKEX’s 2025 Listing Committee Report on Technology Company Disclosures (LC-2025-03) noted that “many issuers provide forward-looking guidance on R&D expenditure as a percentage of revenue, which can serve as a useful input for analysts.”

Common Pitfalls and How to Avoid Them

Double-Counting Amortisation

The most frequent error in this adjustment is double-counting the amortisation of previously capitalised R&D. When the analyst adds back the current year’s capitalised R&D to OCF, they must also ensure that the amortisation of prior years’ capitalised amounts is not being subtracted again. In a standard DCF model, depreciation and amortisation (D&A) are added back to net income to arrive at OCF. If the analyst then adds back capitalised R&D, they are effectively adding back the current year’s expenditure twice — once through the D&A add-back (for the amortisation of prior years’ capitalised amounts) and once through the explicit R&D adjustment. The correct approach is to use the adjusted UFCF formula above, which starts from revenue and deducts cash R&D expense directly, bypassing the need to add back D&A entirely. Alternatively, the analyst can start from reported OCF and add back only the incremental capitalised R&D — i.e., the difference between the current year’s capitalisation and the current year’s amortisation of prior years’ capitalised amounts. This incremental approach is computationally simpler but requires careful tracking of the amortisation schedule.

Ignoring Tax Effects

Capitalising R&D creates a deferred tax liability because the tax deduction is typically taken when the expenditure is incurred, not when it is amortised for book purposes. This timing difference means that the company pays less tax in the current period than its book income suggests. In the DCF model, the analyst must adjust the tax expense to reflect the cash tax paid, not the book tax expense. The simplest method is to use the effective cash tax rate — total cash taxes paid divided by pre-tax income — rather than the statutory or effective book tax rate. For a Hong Kong-listed company with a 16.5% profits tax rate, the cash tax rate may be significantly lower during a period of rising R&D spend, because the tax deduction is front-loaded. The 2024 HKMA Circular on Tax-Related Disclosures in Financial Statements (C24-2024-08) reminds issuers to disclose the reconciliation between the statutory tax rate and the effective tax rate, which provides the data needed for this adjustment.

Over-Adjusting for Non-R&D Intangibles

Not all additions to intangible assets represent capitalised R&D. A company may acquire intangible assets through business combinations (e.g., customer relationships, trademarks, or patents) or through direct purchases from third parties. These acquisitions are genuine investing activities and should not be reclassified as operating cash flows. The analyst must distinguish between internally generated intangibles (capitalised R&D) and externally acquired intangibles. The intangible assets note in the annual report typically separates these categories. If the disclosure is insufficient, the analyst should cross-reference the business combination note to identify acquired intangibles. The SFC’s 2022 Guidance on Intangible Asset Disclosures (CG-2022-07) states that “issuers must disclose the nature of additions to intangible assets, distinguishing between internally generated and externally acquired amounts.”

Actionable Takeaways

1. For each HKEX-listed technology company in your coverage, extract the capitalised R&D amount from the intangible assets roll-forward note and reclassify it as an operating cash outflow in your DCF model to eliminate the upward bias in reported free cash flow.
2. Normalise the terminal year free cash flow by assuming that R&D intensity converges to a steady-state level consistent with industry benchmarks — 12% to 15% of revenue for semiconductor firms and 20% to 25% for biotech, based on HKEX sector data — rather than extrapolating the current growth rate.
3. Adjust the tax expense in your DCF model to reflect the cash tax rate, which will be lower than the book tax rate during periods of rising R&D spend due to the front-loaded tax deduction for capitalised development costs.
4. Verify that the intangible assets note distinguishes between internally generated and externally acquired intangibles; if it does not, request the breakdown from the issuer’s investor relations team or note the limitation in your valuation report.
5. Avoid the double-counting error by using the adjusted UFCF formula that starts from revenue and deducts cash R&D expense directly, rather than adding back both D&A and capitalised R&D to reported OCF.