Software Sector – YTD Performance Snapshot

When Agents Replace Software

36 listed software companies. An average YTD loss of 38 %. But the stock market carnage is only the symptom. The cause: AI agents can already automate 44 % of all cognitive work hours, the very work SaaS was built to organize. This is the data behind the repricing.

−37.7 %
Average YTD Loss (36 companies)
−53.7 %
Average Δ from 52-Week High
2 / 36
Stocks Positive on the Day
77.30
Average P/E (non-negative earners)
Prof. Dr. Tobias Blask
Prof. Dr. Tobias Blask
Founder, svrn_alpha
April 9, 2026

What Happened?

April 9, 2026. Across 36 tracked software names, only two stocks (Trade Desk and Zoom) managed a green day. The rest? A sea of red. Snowflake lost 12 % in a single session. Wix shed 10 %. Cloudflare dropped 9.5 %. In a market already down sharply for the year, these are not corrections. This is a structural repricing.

The chart below tells the YTD story: Atlassian has lost nearly 64 % since January 1st. Asana is down 58 %. monday.com, Figma, Duolingo, HubSpot: every single one has shed between a third and two thirds of its market value in less than four months. The average loss across the 36 names in this analysis stands at –37.7 % YTD. The median is –39.7 %.

This is not a rotation. This is a reset of what software is worth.

What unites these companies? They were the defining stocks of the 2020/2021 era of zero-interest-rate exuberance: high-growth, high-multiple SaaS names that commanded premium valuations on the promise of future cash flows. Now, with higher-for-longer rates, tariff uncertainty rattling risk appetites, and AI threatening the competitive moats of established software franchises, that promise is being marked down. Aggressively.

But this is not just a market story. Two McKinsey reports (the Global Tech Agenda 2026, a survey of 600+ technology leaders, and the MGI landmark study on AI agents and workforce transformation) provide the structural evidence for what the stock market is pricing in. Enterprises are not just cutting SaaS budgets. They are preparing for a world where AI agents perform 44 % of all cognitive work, the very work these software companies were built to mediate.

The Structural Shift: McKinsey's Evidence

McKinsey's February 2026 survey reveals a tectonic shift in enterprise technology spending. AI has now surpassed both cybersecurity and infrastructure modernization as the #1 investment priority for the next two years. Half of all companies identify AI as their top investment area. Among top performers, 28 % plan to increase technology budgets by more than 10 % in 2026 alone, compared to just 3 % of other companies.

The critical point for software investors: those dollars are not flowing into more SaaS licenses. They are flowing into agentic AI systems that autonomously plan, decide, and act across workflows. Precisely the workflows that Atlassian, monday.com, Workday, ServiceNow, and HubSpot were built to manage with human-in-the-loop subscription software.

The McKinsey finding that reframes everything: Top-performing companies are shifting from buying software tools to building intelligence-driven enterprises. Nearly half now operate product and platform models where cross-functional teams replace traditional SaaS-mediated handoffs. Technology's center of gravity has shifted from cost center to value creator, and that value is captured internally, not via third-party subscriptions.

Four Forces Destroying the Old Model

1
From SaaS Subscriptions to Agentic Automation
McKinsey documents that leading companies are scaling agentic AI systems that replace entire workflows. Not augmenting existing SaaS tools, but bypassing them. When an AI agent can coordinate a project, assign tasks, and track deadlines autonomously, the value proposition of Atlassian's Jira or monday.com's boards becomes existentially threatened. The market is pricing this displacement risk now.
2
Insourcing Over Outsourcing
Top performers are bringing strategic technology expertise in-house at nearly twice the rate of other organizations. Half are investing in reskilling their own workforces. This is a direct threat to the "seat-based licensing" model: companies that build internal capabilities need fewer external software tools. They replace SaaS middleware with proprietary AI-powered systems tailored to their specific workflows.
3
Velocity, Not Efficiency
McKinsey's top performers have shifted spending from efficiency (cutting costs via vendor renegotiation) to velocity (accelerating how work gets done). The most popular efficiency lever is no longer renegotiating licenses. It's making teams more productive through AI. When the productivity gains come from AI agents rather than SaaS platforms, the subscription software layer becomes a cost to eliminate, not a tool to optimize.
4
Technology Is Now Strategy, Not a Line Item
Nearly two-thirds of top-performing companies say their CIOs are very involved in crafting enterprise strategy. Among these top performers, almost half now cocreate strategy iteratively throughout the year (up from 18 % in the prior survey), while the all-company figure stands at 29 %. When technology becomes the strategic core of the enterprise, companies stop buying off-the-shelf SaaS and start building bespoke intelligence layers. This is the fundamental threat the market is pricing in.

The Deeper Force: Why This Is Structural

The McKinsey Tech Agenda shows how enterprises are redirecting budgets. A second McKinsey report, the McKinsey Global Institute's "Agents, Robots, and Us" study from late 2025, reveals why. The scale of what AI agents can already do makes the traditional SaaS business model structurally obsolete for a significant share of enterprise workflows.

44 %
US cognitive work hours automatable by AI agents today
$2.9 T
Economic value unlockable in the US by 2030 (midpoint)
27 %
Share of work hours automated by 2030 (midpoint adoption)
77 %
Of the $2.9T value attributed to agents (not robots)
6.8×
Growth of AI fluency demand in job postings (2 years)
75 %
Knowledge workers already using AI tools informally

Consider the implications for the 36 names in this analysis. MGI finds that AI agents can already automate work occupying 44 % of all US work hours in the cognitive domain: the scheduling, tracking, reporting, coordinating, and decision-making tasks that Atlassian, monday.com, Workday, HubSpot, Salesforce, and ServiceNow built their businesses around. In the midpoint adoption scenario, 27 % of all work hours will be automated by 2030, unlocking $2.9 trillion of economic value, of which 77 % comes from agents alone.

The SaaS business model was built for a world where humans use software tools. The new world is one where AI agents are the workers. And they need different infrastructure.

The numbers are already showing up in the real economy. MGI documents that AI fluency (the ability to work alongside AI agents) has seen demand grow nearly sevenfold in two years, faster than any other skill in US job postings. About 75 % of knowledge workers already use AI tools, even when their companies have not formally deployed them. Hiring has reportedly slowed for entry-level programmers and analysts, precisely the roles whose workflow software like GitLab, MongoDB, and Atlassian was designed to support.

The Case Study Evidence

MGI's case studies quantify what this looks like in practice. A large utility's AI agents now handle 40 % of all customer support calls, resolving over 80 % without human involvement, cutting cost per call by 50 %. A global pharma company reduced touch time for medical writing by 60 %. A regional bank's AI agents achieved up to 70 % code accuracy, with developers collaborating with 15 to 20 agents each, cutting required human hours by up to 50 %. In a global tech company's sales operation, AI agents projected 7 to 12 % annual revenue increases while saving 30 to 50 % of time across sales roles.

The critical insight: In every case study, the productivity gains came from redesigning entire workflows around AI agents, not from bolting AI features onto existing SaaS tools. The organizations that captured value eliminated software-mediated handoffs entirely. This is why 90 % of companies have invested in AI, but fewer than 40 % report measurable gains: most allocate the vast majority of their budget to technology and almost nothing to rethinking their processes and upskilling their people. The budget split is precisely backwards. The winners invert it.

MGI analyzed 190 business processes across the US economy and found that about 60 % of potential productivity gains are concentrated in sector-specific domain workflows. These are exactly the workflows that vertical SaaS companies (Veeva in life sciences, Guidewire in insurance, Doximity in healthcare) were built to serve. When AI agents can execute these workflows end-to-end, the SaaS layer becomes a cost to eliminate, not a platform to optimize.

Key Metrics at a Glance

−63.8 %
Worst YTD: Atlassian (TEAM)
−2.7 %
Best YTD: Zoom (ZM)
−39.7 %
Median YTD Loss
−86.7 %
Furthest from 52w High: Figma (FIG)
+0.90 %
Best Day: Zoom (ZM)
−12.0 %
Worst Day: Snowflake (SNOW)

YTD Performance: All 36 Names

Year-to-Date Return (%) · Sorted by Performance · As of April 9, 2026
The scale of the drawdown is extraordinary. Not a single company is above its 200-day moving average. 33 of 36 stocks are below all five key moving averages tracked (RS Rank, 1M, 20SMA, 50SMA, 200SMA). The average distance from 52-week highs stands at –53.7 %. On average, these stocks would need to more than double just to recover their prior peaks.

Key Observations

1
Profitability Offers No Protection
Intuit (P/E 23×), Adobe (13×), ADP (19×), SAP (23×): all solidly profitable, all down 25 to 46 % YTD. The sell-off is not discriminating between unprofitable growth names and established earners. Multiple compression is indiscriminate when the macro narrative shifts.
2
The Unprofitable Cohort Is Getting Destroyed
Atlassian, Asana, Figma, GitLab, Klaviyo, MongoDB, Snowflake, Elastic, Braze, Samsara, Zeta, Cloudflare: all carry "n/a" P/Es, meaning they are either unprofitable or the ratio is not meaningful. On average, this cohort has lost significantly more than profit-generating peers. When risk appetite dries up, no-earnings names are the first to reprice.
3
Valuation Leaders Are Not Outliers
AppLovin at $126.7B market cap is down 45 %. ServiceNow at $93.9B is down 42 %. Salesforce at $157.1B is down 36 %. SAP at $208.2B is down 32 %. Even the largest, most liquid names in enterprise software are not providing shelter. There is nowhere to hide in this sector.
4
Zoom and Twilio: The Relative Survivors
With YTD losses of just –2.7 % (ZM) and –13.1 % (TWLO), Zoom and Twilio are the standouts, and both show green signals across their RS indicators. These are names that already went through their brutal re-rating in 2021/2022 post-pandemic. With expectations already reset and valuations compressed, they have less to fall and are being treated differently by the market today.

Full Dataset: 36 Software Names

All data as of April 9, 2026. P/E ratios where available; "—" indicates not applicable or not meaningful (negative or near-zero earnings). RS indicators show position relative to key moving averages: above, below.

Ticker Company Price 1D % Mkt Cap P/E YTD % Δ 52w High RS · 1M · 20 · 50 · 200
TEAMAtlassian $58.96 −7.64 % $15.5B −63.76 % −75.72 %
ASANAsana $5.73 −5.37 % $1.4B −58.39 % −69.84 %
MNDYmonday.com $62.62 −4.56 % $3.2B 27.98 −57.81 % −80.24 %
DOCSDoximity $21.16 −3.20 % $3.9B 17.78 −52.55 % −72.34 %
FIGFigma $19.06 −5.43 % $9.9B −49.96 % −86.66 %
DUOLDuolingo $90.16 −0.99 % $4.2B 10.52 −48.98 % −83.45 %
HUBSHubSpot $205.46 −5.74 % $10.8B 238.54 −48.84 % −69.90 %
GTLBGitLab $19.67 −7.83 % $3.3B −48.21 % −63.63 %
WDAYWorkday $113.03 −5.15 % $29.0B 43.78 −47.48 % −59.05 %
TTDTrade Desk $20.41 +0.77 % $9.7B 22.72 −46.57 % −77.68 %
INTUIntuit $358.39 −7.71 % $99.2B 23.30 −45.75 % −55.96 %
KVYOKlaviyo $17.81 −1.55 % $5.4B −45.23 % −52.87 %
MDBMongoDB $231.65 −7.39 % $18.6B −45.17 % −47.91 %
APPAppLovin $375.55 −4.00 % $126.7B 16.13 −45.04 % −49.63 %
NOWServiceNow $89.26 −8.42 % $93.9B 53.47 −41.70 % −57.79 %
APPSDigital Turbine $3.01 −2.59 % $361M −40.28 % −63.65 %
SNOWSnowflake $131.96 −12.02 % $45.6B −40.02 % −52.98 %
ESTCElastic $45.31 −7.00 % $4.7B −40.01 % −52.84 %
BRZEBraze $20.80 −3.12 % $2.4B −39.43 % −44.78 %
CVLTCommvault $76.79 −2.89 % $3.4B 39.77 −39.36 % −61.74 %
GWREGuidewire $124.37 −9.13 % $10.5B 56.77 −37.97 % −54.38 %
WIXWIX $66.80 −10.20 % $3.7B 76.07 −36.08 % −65.07 %
CRMSalesforce $170.29 −3.21 % $157.1B 21.87 −35.68 % −42.48 %
ADBEAdobe $229.40 −4.11 % $92.7B 13.38 −34.58 % −45.76 %
DOCUDocuSign $45.17 −1.18 % $8.8B 30.64 −34.16 % −52.29 %
CLBTCellebrite $12.15 −9.57 % $3.0B 39.06 −32.65 % −40.82 %
SAPSAP $164.23 −2.93 % $208.2B 23.11 −32.28 % −47.58 %
VEEVVeeva $157.19 −5.60 % $25.7B 28.90 −29.64 % −49.38 %
ZETAZeta $15.20 −3.31 % $3.7B −26.21 % −38.96 %
IOTSamsara $26.71 −8.84 % $15.5B −25.60 % −44.83 %
ADSKAutodesk $221.03 −8.15 % $46.6B 42.28 −25.33 % −32.84 %
ADPADP $195.51 −2.62 % $78.7B 18.80 −23.93 % −40.73 %
DDOGDatadog $109.44 −6.06 % $38.7B 333.28 −20.26 % −45.74 %
TWLOTwilio $124.10 −4.26 % $18.8B 586.07 −13.13 % −14.94 %
NETCloudflare $191.56 −9.46 % $67.3B −3.64 % −26.32 %
ZMZoom $83.98 +0.90 % $24.7B 13.59 −2.67 % −13.94 %

Average P/E (profitable companies only, n=23): 77.30  ·  April 9, 2026

What This Means for Investors

The AI Disruption Discount

The McKinsey data provides the analytical framework for what the market is expressing through prices. This is not panic selling. It is the rational repricing of an entire business model that is losing its structural advantages.

Workflow automation tools (Asana, monday.com, Workday) face headwinds from agentic AI systems that McKinsey's top-performing companies are already deploying at scale. Systems that autonomously plan, decide, and act across the very workflows these SaaS tools were built to manage. Developer tools (GitLab, MongoDB, Atlassian) face pressure from AI-native coding environments. Marketing clouds (HubSpot, Braze, Klaviyo) are being challenged by enterprises building their own intelligence layers rather than buying off-the-shelf solutions.

The market is asking a hard question: if McKinsey's top performers are rewiring their entire operating models around AI and insourcing strategic technology capabilities, who is left to renew these SaaS subscriptions at current prices?

The Budget Paradox

Here is the paradox that makes this different from 2022: enterprise tech budgets are growing, not shrinking. Half of companies plan to increase technology spending by more than 4 % in 2026. Among top performers, 28 % are increasing budgets by more than 10 %. But this money is not flowing into traditional software. It is flowing into AI infrastructure, proprietary agentic systems, and insourced capabilities. The total addressable market for SaaS may be growing in nominal terms while the share captured by subscription software companies shrinks. That is the worst possible setup for stocks priced on revenue multiples.

The 52-Week High Problem

The average distance from 52-week highs across this cohort is –53.7 %. That is not a correction. That is a regime change. For context: a stock down 50 % from its high needs to rise 100 % to recover. The technical picture across these names is uniformly bearish. All 36 companies are below their 50-day and 200-day moving averages, and 33 of 36 show no green signals in any of the five tracked RS indicators.

The Survivors' Playbook

Zoom and Twilio stand apart. Both were already eviscerated in the 2022 tech wreck (Zoom lost 90 % from its pandemic peak, Twilio lost similarly). Today, expectations are calibrated to a lower base. Zoom has a P/E of 13.6×, below many "value" stocks. Twilio is showing three green RS signals. These are not recovery stories yet, but they are the names least likely to disappoint from here, precisely because the market already priced in their structural challenges years ago.

The McKinsey framework suggests a survivor profile: companies that become infrastructure for AI rather than being displaced by AI. Datadog (observability for AI systems), Cloudflare (infrastructure for agentic workloads), and Twilio (communications APIs that AI agents need) may fit this pattern. The competitive advantage of the next era will be architectural and sovereign, not based on vendor relationships or subscription lock-in. The market has not yet made this distinction, which is either an inefficiency or a sign that the repricing has further to go.

The data does not yet signal a bottom. When 33 of 36 large-cap software names are below all five key moving averages simultaneously, the setup for a sustained recovery is not in place. Unlike 2022, where the selloff was driven by rates and eventually reversed when multiples stabilized, the structural pressure this time is real and accelerating. MGI documents $2.9 trillion of economic value being unlocked through AI agents by 2030. That value does not flow through SaaS subscriptions. It flows through redesigned workflows where agents replace the software-mediated handoffs that these companies monetize.

The Core Question

For investors, the question is no longer whether traditional SaaS is being disrupted. MGI's data makes that unambiguous. The question is how much of that disruption is already priced in at –38 % average YTD. If 44 % of cognitive work hours are automatable and only 27 % will be automated by 2030 (midpoint), the disruption is still in its early innings. The market may be right on direction but wrong on timing.

What comes after SaaS is not another generation of software tools. It is an operating layer where autonomous agents execute real work, where humans set direction and agents compound results, and where the institution that controls its own infrastructure controls its own future. The companies and institutions that understand this will not just survive the repricing. They will define what replaces it.

What Does the Post-SaaS Operating Layer Look Like?

We are building it. Sovereign, model-agnostic, designed for institutions that want to control their own infrastructure and their own future.

See the Architecture