Investors appear to have priced in a 32.6% lasting productivity boost for software engineering from AI, according to a working paper by economists at UC Berkeley and the London School of Economics. That is a measure of market expectations, not proof that engineers are already getting 32.6% more done.
Watch Desk analysis
What happened
The researchers analysed how company share prices responded to AI-related news between November 2022 and December 2025, looking at whether firms with larger software engineering payrolls gained more when an AI stock index rose. They used an economic model to estimate the productivity gain investors expect.
The findings appear in the working paper The Macroeconomic Effect of AI: Sizing the Software Engineering Channel. The Register reports that co-author Chen Lian described the method as translating that relationship between AI news and share prices into anticipated productivity gains.
The researchers also estimate that AI-related news over the period corresponds to a present-value GDP increase equivalent to a permanent 3.61% rise in the level of GDP. They caution that market expectations may prove too optimistic or too pessimistic.
Why it matters
This is an attempt to measure what investors think AI might do before its effects are fully visible in productivity statistics. That makes the estimate interesting, but also means it is a forecast embedded in market movements, not a tally of gains delivered in workplaces.
The researchers’ figure sits in the same broad territory as reported task-level productivity gains from coding tools, but those gains can meet bottlenecks elsewhere. The account notes one obvious candidate: code reviews may struggle to keep pace if AI increases the flow of commits. Faster code generation is useful; faster queues to review it are less so.
Our read
The 32.6% figure is best read as a snapshot of investor expectations, not a productivity receipt. Its value is that it makes those expectations measurable and open to comparison with what happens next. The real test will be whether businesses turn faster coding tasks into lasting gains across the whole development process, rather than simply producing more work for the review queue.
What to watch
- Whether later productivity data supports the market-implied estimate.
- Whether software engineering employment and output change alongside AI adoption.
- How code review and other workflow bottlenecks affect the gains.
- Whether the researchers’ method produces useful estimates for other parts of the economy.
Discussion spark: If AI speeds up coding but creates a larger review backlog, should companies count that as a productivity gain?
Sources and evidence
- Investors are pricing in a 32.6% AI productivity boost for software engineers – news.lavx.hu (29 September 2026, 12:33 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.