As the AI wave reshapes the technology sector, Goldman Sachs' top TMT trader Peter Callahan has outlined 10 critical observations and 10 key charts that highlight the most significant structural shifts driving tech investment today.
The core themes center on three transformative comparisons: AI agent traffic is surpassing human traffic, optical components are outperforming memory chips, and open-source models are achieving superior performance at less than 8% of the cost of their closed-source rivals. These dynamics reveal a deep evolution in the investment logic for AI infrastructure.
In his August 9th report, Callahan maintains a constructive view on the tech sector for August and beyond, supported by cleaner positioning, reduced technical pressure, reasonable valuations, and strong underlying fundamentals. The Nasdaq 100 rallied 5% last week, with the VIX closing near 14, a new low for the year, as discussions around the Fed's policy path faded after the non-farm payroll data.
At the industry level, the latest earnings season has widened the sentiment gap between data infrastructure and development tool companies versus traditional SaaS application providers. The former are shifting from an AI "headwind" narrative to an AI "tailwind" narrative. Whether the latter can keep pace with this transition will be a key market focus moving forward. Meanwhile, persistent GPU shortages, non-human network traffic exceeding 50%, and rapidly expanding corporate AI budgets all point to an accelerating "inference economy."
Three Major Structural Themes: AI Agents, Optics, and Open-Source
AI agents are replacing humans as the internet's primary users. Cloudflare disclosed during its earnings call that non-human traffic has already surpassed human traffic and predicted that if current trends continue, non-human traffic will be 1,000 times greater than human traffic in five years, making humans a "rounding error" on the internet. Third-party data confirms this, showing that bot traffic now accounts for 53% of all web traffic, with human traffic falling below 50%.
Corporate AI budgets are expanding, but ROI divergence is emerging. Booking reported that its AI investments have already achieved positive returns, with related costs still at low single-digit percentages of overall technology spending. MercadoLibre spent approximately $80 million on AI this quarter and disclosed that its customer service headcount dropped from 10,000 to 7,000, even as its business volume grew threefold, with 90% of customer interactions now handled without human intervention.
Open-source models are challenging the closed-source ecosystem at a fraction of the cost. Pinterest revealed in its earnings report that after post-training an open-source model on its proprietary data, it achieved performance surpassing that of a closed-source third-party model, with each transaction costing less than 8% of the comparable closed-source model. This data directly challenges the business model of closed-source AI and provides a compelling real-world example in the ongoing "open-source vs. closed-source" debate.
A recent Goldman Sachs sales and trading report characterizes the current market as "returning to normal, not booming," describing the slowdown in earnings growth as a "structural change, not a directional reversal." Goldman Sachs partner Louis Miller concluded with a market adage: "The trend is your friend until it bends in the end."
1. Disney - Breakout or Breakdown?
Disney delivered a strong quarter with an acceleration in domestic theme park attendance, but debate over earnings revision potential persists. Technically, DIS stock has been in a downward channel for years and is now at a critical juncture, with investors closely watching whether it will break out or resume its decline.
2. Optics vs. Storage - The Quarter's Most Notable Sector Divergence
This is one of the most frequent topics Callahan has received from clients this week. Both optical stocks (blue line) and memory stocks (red line) experienced roughly 30% corrections in July, but since August, optical stocks have outperformed memory stocks by over 20 percentage points. Specifically, Coherent has risen for seven consecutive sessions, gaining 73% during that period. Lumentum is up 43%, Fabrinet 36%, Corning 34%, and Ciena 24% over the same period. In contrast, NAND, DRAM, and HDD storage stocks have generally seen gains of only 0% to 20%. This divergence reflects a market repricing of compute demand for the AI inference phase, with the demand for optical interconnects growing faster than expected.
3. Block - Can the Stock Price Catch Up to Fundamentals?
Payment service provider Block's stock price (black line) has been trading in the low $80 range, while its gross profit (blue line) has continued to rise, creating a growing divergence. Goldman Sachs trader Christian DeGrasse's chart poses a central question: can the stock price eventually converge with fundamentals and achieve an upward recovery?
4. HP - 40% Gain in 40 Days, Key Resistance Ahead
HP stock has quietly rallied about 40% in 40 trading days and is now retesting a key resistance level around $30, which has acted as a ceiling multiple times over the past year. A successful breakout will be a critical signal for the nature of this rebound.
5. SaaS Observation - A Breakout Moment Above the 200-Day Moving Average
While market conversations focus on identifying "inference economy" beneficiaries like Snowflake, Cloudflare, Palantir, CrowdStrike, and Palo Alto Networks, Callahan notes that ServiceNow has closed above its 200-day moving average for the first time in over a year, a signal that is hard to ignore. This aligns with his core view on SaaS/application companies: as Intuit, Adobe, and Salesforce release earnings outside of the main quarterly cycle, the market will closely watch if these companies can also build their own AI "tailwind" narratives. Notably, Atlassian has rebounded about 160% from its year-to-date low, but it is still down roughly 6% for the year—a contradictory data point that encapsulates the current state of the SaaS sector.
6. NVIDIA - A Belated Catch-Up Rally?
NVIDIA rose 11.5% last week, its best weekly performance since May 2025. Callahan believes the catalyst path is becoming more favorable, with upcoming earnings, the Vera Rubin architecture launch, and a busy industry conference season forming potential drivers for a NVDA catch-up rally. The chart shows the relative performance of NVIDIA versus the Philadelphia Semiconductor Index (SOX), with NVIDIA's relative strength being re-established. GPU supply remains tight. Akamai stated on its earnings call, "GPU demand is still exceptionally strong, and all of our GPU capacity is fully sold out," echoing comments from Microsoft and Amazon.
7. Spotify - The Divergence Between Positive Sentiment and Downward Revisions
Investor sentiment around Spotify's AI music initiatives is quite positive, but the chart reveals a concerning reality: over the past two years, the company's operating profit forecasts (orange line) have been consistently revised downward, while the stock price (white line) has remained elevated on sentiment. Callahan's assessment is straightforward: "Same ol' story." The divergence between sentiment and fundamentals is a long-standing core issue for this stock.
8. Atlassian - Two Consecutive 30% Single-Day Surges, Still Negative for the Year
Atlassian has seen its stock surge over 30% on the day after each of its last two earnings reports, driving a cumulative rebound of about 160% from its year-to-date low. However, the chart reveals a dramatic fact: TEAM is still in negative territory for the year. This underscores both the depth of the prior decline and the need for investors to carefully assess the margin of safety when chasing the stock.
9. UBER - The GARP Thesis Emerges
Uber is again showing a pattern of "beating estimates, then selling off on the day," but the stock has risen about 15% over the past two weeks. Investor conversations are increasingly referencing GARP (Growth at a Reasonable Price) and value investment logic. At roughly 13x forward free cash flow with mid-to-high single-digit to double-digit revenue growth, the valuation is attracting institutional interest.
10. Q2 Earnings - Information Technology Leads the S&P 500
The Information Technology sector led the S&P 500 in Q2 earnings season, with 82% of constituent companies beating earnings estimates and 72% beating revenue estimates. This data provides the most direct fundamental support for Callahan's constructive view on the tech sector.

