Every few years, the market convinces itself that a foundational sector of the economy is functionally obsolete. In 2020, it was energy (terminal decline). In 2022, it was physical retail (e-commerce singularity). Today, if you spend any time on the financial industry corners of the internet (X, Reddit), you will be led to believe that the Enterprise SaaS model is dead. A walking zombie toward a cliff edge of deflationary AI pricing and vibe-coded obsolescence.

The consensus trade right now is crowded and loud: Long tangibles, short software. The logic is that Nvidia and the hyperscalers (the “Fragile Fifty Percent” driving the S&P) are the only game in town, while software seats are about to be vaporized by autonomous agents written by teenagers in San Francisco.

I am taking the other side of that trade.

But before I tell you why, I think we are on the precipice of a massive value realization for the application layer, let’s give the bears their day in court. To properly underwrite software, you have to stare the Death of SaaS narrative in the face.

The Prosecution: Why Your Subscription is Going to Zero

If I were building the bear case for software today, I wouldn’t just wave my hands at “AI”. I would look at the unit economics. Here is the strongest version of the argument that the SaaS business model is structurally broken:

1. Differentiation Trends to Zero: AI makes code cheap. The “moat” of features is gone; it’s a knife fight for every renewal.
2. The “Agentic” bypass: Value accrues to the agentic layer sitting on top of the System of Record (SOR). The database becomes a dumb pipe.
3. The Vibe Coding Startups: AI-native startups, unburdened by technical debt, can spin up bespoke CRM or ERP wrappers for a fraction of the cost.
4. Deflationary Seat Counts: As agents do more work, human seats decline. If you sell per-seat license, your revenue model is mathematically doomed.
5. The outcome Trap: Legacy SaaS struggles to transition from “selling seats” to “selling outcomes”. It’s a DNA problem they can’t solve.
6. Pricing Power Erosion: Without differentiation and lock-in, the 10% annual price hike is history.
7. Margin Deterioration: AI revenue is structurally lower margin due to high inference costs.
8. The CAC Spike: LLMs kill organic search traffic, causing Customer Acquisition Costs to skyrocket.
9. Talent Inflation: Competition for scarce AItalent explodes OpEx.
10. Investor Sentiment: The sector gets de-rated permanently as “ex-growth”.

It is a terrifying list. It is also, in my view, almost entirely wrong about how large enterprises actually function.

The Defense: Why the SF Teen Won’t Replace Your ERP

The bear case relies on a fundamental misunderstanding of what a CIO actually buys. Your Enterprise Resource Planning is the central nervous system of your organization.

They aren’t buying “code”. They are buying accountability, compliance, and liability insurance.

1. The “Internal IT Shop” Fallacy: The bear case assumes that because it is possible to build a bespoke Jira wrapper with an LLM for $20, a Fortune 500 CIO will actually do it. They won’t. Any of the CIO surveys, which there are plenty for investors to access, show there is zero appetite for turning revenue-generating engineering teams into internal IT shops tasked with maintaining vibe-coded slop products. If Workday breaks, the CIO calls Workday. If an internal app breaks, the CIO gets fired. Enterprises outsource liability. They will happily pay a premium to Salesforce just to have someone to sue if the database corrupts.
   
2. The Compliance Moat: there is a reason Anthropic uses Workday and OpenAI uses Slack. Even the creators of these models are not vibe coding their own payroll systems. Why? Governance. Microsoft, ServiceNow, and Salesforce have spent two decades building FedRAMP, SOC2, and HIPAA compliance moats. A demo built on Lovable (no shade) cannot pass a bank’s procurement security review. The Compliance Moat is the most underpriced asset in software right now, in my opinion.
   
3. The Coming Wave:  The market’s current obsession with “stalling” LLM benchmarks (just ask Ben Affleck) is a category error. It confused training gains with agentic utility. J.P. Morgan’s latest analysis projects that based on current trajectories, agentic models will reach human-level performance by this Spring. This matters because we are transitioning from Phase 1 (infrastructure) to Phase 4 (Applications). History shows that value creation in tech cycles moves in waves. From builders to users. The hardware trade is crowded; the application trade has been taken to the woodshed. If the ‘Agentic Breakout’ hits in a few months from now, the re-rating in software won’t be linear; it will be violent.
   
4. Debunking the Seat Count Apocalypse: The most persistent bear argument is that AI kills seats. The data says otherwise. According to J.P. Morgan’s research, unemployment rate in sectors most exposed to AI disruption are actually lower than in insulated sectors. Furthermore, firms that have adopted AI are seeing labor productivity gains of ~30%. This supports the Jevons Paradox: As the cost of a task drops, the demand for that task rises. AI doesn’t kill seats; it makes the output of the seat more valuable.

Exhibit A: The Earnings Reality Check

If you want to know if software is dying, don’t check a viral anonymous Twitter account. Look at the earnings prints from this week. Or at least open the Quartr app. If the thesis were true, seat counts should be imploding. They aren’t.

ServiceNow: Bears have argued that AI would bypass the ITSM layer. The reality? NOW just posted a $3.47B in subscription revenue (+21% YoY) and a massive 25% surge in cRPO. Most critically, renewal rates ticked up to 98%. AI embedded directly into the stack is a churn reducer. When the product gets smarter, switching gets harder. The market is also mispricing the recent Armis acquisition. By integrating Armis’ cyber-asset intelligence, NOW isn’t just “managing IT”; they are building the map required for autonomous agents to operate safely. They are building the control tower for the autonomous enterprise.[1]

Microsoft: The death of CRM battle cry might be the loudest among the bears. The most linear argument has been that AI eats CRM, i.e. we won’t need Salesforce or Microsoft’s Dynamics because an agent will manage data in a lake. The reality is, Microsoft Dynamics 365 revenue accelerated to 19% on a year-on-year basis in Q4 last year – according to this week’s earnings release. If AI were replacing the application layer, Dynamics would be shrinking. Instead, it is capturing share. Look at the Microsoft 365 Commercial numbers: Seats grew at 6%, but Cloud Revenue grew 17%. The growth isn’t coming from more humans; it’s coming from more value per human. Copilot is driving a massive expansion in Average Revenue Per User (ARPU).[2]

The Verdict

The bearish narrativeon software won’t vanish overnight. It will be discounted away, quarter by quarter, as the catastrophic downside revisions fail to materialize.

We are watching a classic hype cycle trough. The tourists have left the software trade, chasing the hardware momentum. But the hardware is just the shovel; the software is the gold mine.

We prefer incumbents. The “boring” companies with the distribution, the data, and the balance sheets to survive the transition to outcome-based pricing. The bear case is seductive, but as a friend of mine recently put it to me: “You can’t replace the integrated stack of a global bank with a vibe-coded LLM wrapper. It is genuinely the lowest EV activity you can do.” SaaS isn’t dead. It’s just getting smart.

---

[1] All data provided by ServiceNow 10-K Annual Report as of 1/29/2026.

[2] All data provided by Microsoft’s 10-Q First Quarter Report as of 1/29/2026.

IMPORTANT LEGAL DISCLOSURES

CURRENT MARKET DATA IS AS OF 12/31/2025. OPINIONS AND PREDICTIONS ARE AS OF 12/31/2025 AND ARE SUBJECT TO CHANGE AT ANY TIME BASED ON MARKET AND OTHER CONDITIONS. NO PREDICTIONS OR FORECASTS CAN BE GUARANTEED. INFORMATION CONTAINED HEREIN HAS BEEN OBTAINED FROM SOURCES BELIEVED TO BE RELIABLE BUT IS NOT GUARANTEED.

THIS PRESENTATION (THE “PRESENTATION”) HAS BEEN PREPARED SOLELY FOR INFORMATION PURPOSES AND IS NOT INTENDED TO BE AN OFFER OR SOLICITATION AND IS BEING FURNISHED SOLELY FOR USE BY CLIENTS AND PROSPECTIVE CLIENTS IN CONSIDERING GFG CAPITAL, LLC (“GFG CAPITAL” OR THE “COMPANY”) AS THEIR INVESTMENT ADVISER. DO NOT USE THE FOREGOING AS THE SOLE BASIS OF INVESTMENT DECISIONS. ALL SOURCES DEEMED RELIABLE HOWEVER GFG CAPITAL ASSUMES NO RESPONSIBILITY FOR ANY INACCURACIES. THE OPINIONS CONTAINED HEREIN ARE NOT RECOMMENDATIONS.

THIS MATERIAL DOES NOT CONSTITUTE A RECOMMENDATION TO BUY OR SELL ANY SPECIFIC SECURITY, PAST PERFORMANCE IS NOT INDICATIVE OF FUTURE RESULTS. INVESTING INVOLVES RISK, INCLUDING THE POSSIBLE LOSS OF A PRINCIPAL INVESTMENT.

INDEX PERFORMANCE IS PRESENTED FOR ILLUSTRATIVE PURPOSES ONLY. DIRECT INVESTMENT CANNOT BE MADE INTO AN INDEX. INVESTMENT IN EQUITIES INVOLVES MORE RISK THAN OTHER SECURITIES AND MAY HAVE THE POTENTIAL FOR HIGHER RETURNS AND GREATER LOSSES. BONDS HAVE INTEREST RATE RISK AND CREDIT RISK. AS INTEREST RATES RISE, EXISTING BOND PRICES FALL AND CAN CAUSE THE VALUE OF AN INVESTMENT TO DECLINE. CHANGES IN INTEREST RATES GENERALLY HAVE A GREATER EFFECT ON BONDS WITH LONGER MATURITIES THAN ON THOSE WITH SHORTER MATURITIES. CREDIT RISK REFERES TO THE POSSIBLITY THAT THE ISSUER OF THE BOND WILL NOT BE ABLE TO MAKE PRINCIPAL AND/OR INTEREST PAYMENTS.

THE INFORMATION CONTAINED HEREIN HAS BEEN PREPARED TO ASSIST INTERESTED PARTIES IN MAKING THEIR OWN EVALUATION OF GFG CAPITAL AND DOES NOT PURPORT TO CONTAIN ALL OF THE INFORMATION THAT A PROSPECTIVE CLIENT MAY DESIRE. IN ALL CASES, INTERESTED PARTIES SHOULD CONDUCT THEIR OWN INVESTIGATION AND ANALYSIS OF GFG CAPITAL AND THE DATA SET FORTH IN THIS PRESENTATION. FOR A FULL DESCRIPTION OF GFG CAPITAL’S ADVISORY SERVICES AND FEES, PLEASE REFER TO OUR FORM ADV PART 2 DISCLOSURE BROCHURE AVAILABLE BY REQUEST OR AT THE FOLLOWING WEBSITE: HTTP://WWW.ADVISERINFO.SEC.GOV/.

ALL COMMUNICATIONS, INQUIRIES AND REQUESTS FOR INFORMATION RELATING TO THIS PRESENTATION SHOULD BE ADDRESSED TO GFG CAPITAL AT 305-810-6500.