R&D: Expensed, So Profit Looks Worse
Research and development is spending on creating new products and processes. In most jurisdictions it is expensed as incurred rather than capitalised as an asset, so a company investing heavily reports lower profit than an otherwise identical one that has stopped investing.
How it works
Research and development sits with the overheads, below gross profit. It is spending on things that do not exist yet, charged against revenue from things that already do.
The accounting treatment is the whole story. Buy a machine and it becomes an asset depreciated over years. Spend the same amount inventing something and it is charged in full this year, with nothing on the balance sheet to show for it.
The reason is evidential rather than economic. Accounting standards require reasonable certainty of future benefit before something becomes an asset, and most research does not produce anything at all.
What it does to the profit figure
A company spending heavily reports a lower profit. In the illustration, 700 of operating income becomes 1,200 with the 500 of research spending added back — the same business, described two ways.
Which creates an obvious and well-used lever. Halving the spend raises reported profit immediately and costs nothing until the products that were not developed fail to arrive, several years later.
A falling ratio is a decision, not a drift. Spending as a share of revenue declining year after year at a company that depends on new products is management choosing present earnings over future ones. It is one of the few forward-looking signals available in a set of historical accounts.
Unlike depreciation, this is real money going out. Adding it back to estimate underlying profitability is legitimate; treating it as a non-cash charge is not, because the cash genuinely left.
In practice: comparing companies
The ratio is only comparable within an industry. Twenty per cent of revenue is ordinary for software, three per cent is normal for an industrial business, and a retailer may spend nothing at all.
Software development is the significant exception. Once a project reaches technical feasibility, part of the cost can be capitalised — which moves spending off the income statement and onto the balance sheet.
Which makes two similar companies non-comparable without checking. One expensing everything reports lower profit than one capitalising aggressively, with no operational difference between them.
Tax credits reduce the real cost. Many countries refund a meaningful share of qualifying spending, so the gross figure on the income statement overstates what the company actually paid.
The adjustment worth making is one addition. Operating income plus research spending gives what the business would earn if it stopped investing — which is not a forecast, and it is the right denominator for asking whether the investment is affordable.
One consequence of the accounting reaches further than the income statement: it distorts the balance sheet too. A company that has spent a decade building products carries none of that work as an asset, so its book equity is a fraction of what it has actually created. Compare it with a company that bought the same capability through an acquisition, and the second one carries goodwill on its balance sheet for exactly the thing the first one built itself.
Which makes book-value comparisons between builders and buyers close to meaningless. Return on equity looks spectacular at the company with no assets recorded, and ordinary at the one that paid cash for the same position. Neither figure is describing operating quality — they are describing how the capability was acquired, and the accounting rules do the rest.
There is one more asymmetry worth naming: the spending is disclosed and the results are not. Companies report what they spent and almost never what it produced, so the only available assessment is indirect — revenue from products launched in the last three years, where any company gives that figure, is a far better measure than the input.
What research and development is not
It is not an asset, in most cases. It is charged in full this year.
It is not a non-cash charge. The money genuinely leaves.
It is not comparable across industries. Normal levels vary enormously.
And it is not evidence of innovation. It measures spending, not results.
When it fails as a signal
High spending is not the same as productive spending. The line records money going out and says nothing about what came back, and companies with the largest budgets are not reliably the most inventive.
A second failure is comparing gross figures across tax regimes. Credits vary by country and can change the net cost substantially.
A third is missing capitalisation. A company moving development costs to the balance sheet shows falling research expense and rising profit with no change in behaviour.
A fourth is treating a cut as efficiency. It raises this year’s profit by definition, and the effect takes years to appear.
And a fifth is adding it back and then valuing the result. The underlying figure is a comparison tool, not a sustainable earnings estimate — a business that stops investing usually stops earning too.
The original data
Of the 31,760 trading and investing videos in this site’s corpus, 0 have “research and development” in the
title and 0 have “income statement”. “Financial statements” returns 3 at a median of 1,065,893 views, “cash
flow” returns 17 at a median of 67,134, and “valuation” returns 9 at a median of 17,868. The counts are in
research/corpus-coverage.json, produced by site/measure_corpus.py.
The figures in the diagrams on this page are illustrative. The 700 becoming 1,200 when 500 of spending is added back is the arithmetic that explains why so many growing technology companies report thin profits while generating substantial cash.
The check worth running is the ratio across five years, beside revenue growth. Spending falling as a share of revenue while revenue growth also falls is the pattern to take seriously — it is a company harvesting rather than building, and the accounts will look better each year until they do not.
Related
Operating expenses is the wider category this belongs to. Operating income is the figure it is subtracted from. And growth investing is the approach that has to read this line carefully.
Adding this line back before looking at profit changed how I read technology companies entirely. A business spending twenty per cent of revenue on development and breaking even is not a break-even business - it is a profitable one choosing to spend the profit, and the accounting cannot show you the difference.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.