Goldman Sachs: Software stocks' rebound unlikely to signal a reversal, sustained growth may wait until 2027

Goldman Sachs pointed out in the latest release of the "Revisiting the Moat" series, the fifth report, that although the software sector has experienced two rebounds this year, the fundamental improvement truly driven by AI is likely to be a matter of 2027, and the current market is more suitable for selecting individual stocks rather than betting on the entire sector.

Since the beginning of this year, the median stock price drop for application software companies covered by Goldman Sachs has been 38%. There have been two decent rebounds: a rise of about 14% from the end of February to early March, and a rise of about 22% from mid-April to mid-May. However, Goldman Sachs believes that these rebounds do not yet have the conditions to evolve into a sustained market trend.

The core judgment of the report is that the software sector will most likely maintain a trading range in the coming months, but individual stock differentiation will be very significant. The first-quarter report has already confirmed this point, with some companies experiencing daily price fluctuations of more than 20%, making stock selection ability the key to obtaining excess returns.

Green shoots are emerging, but monetization is still far away

Goldman Sachs observed that the management of leading software companies have already realized that the competitive landscape is intensifying and have begun to make organizational adjustments. Klaviyo introduced a co-CEO in December 2025, allowing the founder to focus on product development; Workday's founder resumed the position of CEO in February this year; Adobe announced the CEO succession plan in March.

There has also been progress on the product front. ServiceNow has launched an IT Service Management (ITSM) level automation product and already has 50 pilot customers, while Salesforce has released a "headless CRM" architecture. These are positive signals, but there is still a distance from being reflected in the financial data.

Goldman Sachs' industry research shows that software companies are facing a 1% to 4% elongation of the sales cycle, with customers tending to sign contracts with shorter terms and being more aggressive in pricing when renewing. Even companies that have an early layout in AI are also pointing to 2027 for more substantial revenue contributions. Salesforce launched Agentforce in September 2024, and it took a year to see an inflection point in new annualized revenue, stating that it would take another 12 to 18 months for substantial revenue improvement.

AI revenue must be incremental, not substitutionary

Goldman Sachs believes that the key to a real shift in sector sentiment is that AI-generated revenue must be an addition to the company's overall growth, not just a reshuffling within the internal structure. If AI expands the overall profit pool, it will be reflected in core indicators such as backlog of orders, billings growth, and gross profit rate acceleration.

ServiceNow has revised its AI-related annual contract value target from $1 billion to $1.5 billion for 2026, but its organic growth guidance for the remaining performance obligations suggests a deceleration of about 250 basis points. Goldman Sachs pointed out that only when investors confirm that AI revenue is making a net contribution to overall growth will the relevant disclosures be interpreted positively.

The report referenced the history of Meta as a reference: Meta's stock price continued to be under pressure after its IPO in May 2012 until October 2013 when mobile advertising revenue exceeded expectations, and the stock price truly reversed. Goldman Sachs believes that the software industry is currently in a similar "proof period".

Historical pattern: Mean reversion exists, but differentiation is obvious

Goldman Sachs traced back the performance data of 154 industry sectors since 2001. Among all sectors that have experienced significant declines, the sector with the poorest performance has a median return of 4% in the following year, a median return of 20% in three years, and a positive return probability of about 52% to 54%.

It is worth noting that sectors that fell due to cyclical factors rebounded stronger—with a median return of 45% in three years; whereas sectors that fell due to structural concerns had a median return of only 16% in three years, but still had a more than half probability of achieving positive returns.

Opportunities at the individual stock level

Goldman Sachs believes that Oracle and Cloudflare on the infrastructure side are expected to benefit from the increased demand for inference, with Oracle being in a favorable position in the upcoming financial report. On the application side, Microsoft and ServiceNow are favored, as they are most capable of pushing the AI proxy layer from pilot to monetization.

In contrast, Intuit faces new competitive threats in