Anthony Hopkins starred in a movie called "The Father", in which the protagonist, suffering from Alzheimer's disease, loses his grasp on the concept of time, embarking on a disorienting journey through time: Are events happening now, a decade ago, or never at all? Why does someone who just appeared a second ago say "long time no see" when we meet again? Is time real, or merely a human construct? Pondering this angle, one realizes that "The Father" is not just about Alzheimer's but more fundamentally about the essence of existence (though perhaps I'm overthinking it).

More than two years ago, when I first watched this movie, I immediately thought of an old friend – the A-share software industry. And now, I find myself thinking of it again. I've studied the software industry far longer than the internet industry; in industry jargon, I'm a "computer industry analyst." To clarify, the "computer industry" in A-shares is essentially synonymous with the software industry, as there are few hardware companies listed. Even if there were, they've been usurped by the neighboring communication and electronic component industries, leaving only the inseparable software companies.

Every software industry analyst harbors aspirations of shaping the world with their insights, much like every software company aspires to upend Microsoft and outcompete IBM. I recall when I first joined the industry over a decade ago, three concepts dominated the A-share software landscape:

The first was "Core Electronics, High-end Communication Chips, and Basic Software Products" (dubbed "Core, High, and Basic"), of which the first two had little to do with software, but the latter was software's turf. Since 2009, at least fifty A-share software companies have claimed at some point to be developing operating systems, databases, middleware, or office software – all part of the "basic software" coveted by capital markets. Second was "Cloud Computing," encompassing public and private clouds, spanning IaaS/PaaS/SaaS layers. As most A-share software firms lacked the capabilities for infrastructure and platform layers, they repeatedly emphasized: the application layer is the future, and we will devour the trillion-yuan cloud computing market! (When they first discussed cloud computing, Alibaba Cloud hadn't even launched, let alone Tencent Cloud.) Third was "Information Security," including basic, universal security technology products and tailored solutions for industries (especially local governments). These companies flourished after 2010, with over thirty listing on A-shares. Later, "security" and "controllability" merged into the "Anke" concept, combined with "autonomy," forming the iron triangle of A-share software company concepts.

Besides these three themes, several other popular concepts occasionally surfaced. For instance, software outsourcing/business process outsourcing once created a crop of high-performing ChiNext stocks, but outsourcing was deemed too low-end for Indians, earning only 30-40x P/E ratios. Then there were various industry application solutions, essentially system integration plus a mishmash of custom development. Yet, financial statements don't lie; people gradually realized these "solutions" boasted only 20-30% gross margins, less profitable than being an agent. Thus, such businesses commanded 30-40x P/E ratios, rarely featured in broker research reports.

The most impressive, commanding 70-80x or even 100x P/E ratios, were companies mastering "chokehold technologies," leading technological trends, or even choking others. I recall brokers proposing the "Wave-Qiming-ChinaSoft-Teamsun" concept, alleging they held such technologies. Later, UFIDA Network (then UFIDA Software), Aerospace Information, Taiji Computers, Venustech, and ZKTech were added. Entry criteria: "irrefutable chokehold capabilities" – if no one can prove you can't choke others, you can, end of discussion.

No computer analyst held internet companies in high regard – consumer internet firms were seen as lowly money-grubbing enterprises. Baidu garnered some respect due to its "technological edge," while Tencent and Alibaba, "lacking core technologies," were dismissed.

Theoretically, studying the software industry requires strong technical fundamentals, at least a basic understanding of programming and cutting-edge computer technology. Oddly, few software industry analysts come from a software background and seem unwelcome. They know too much, "only industry logic, not capital market logic." Ultimately, the A-share software industry (and any industry) is a grand show, requiring a dazzling performance before the curtains close. Those who refuse to dance and shout "this is a show" are not only obnoxious but also face public scorn.

This reaffirms Andersen's greatness as a fairytale writer. "The Emperor's New Clothes" is just a fairy tale; in reality, those who cry out "the emperor has no clothes" fare worse. So, what's the endgame of consistently rejecting truth? Alzheimer's disease, becoming "trapped in time."

If you've read 2010 and 2023 computer industry strategy reports from major brokers, you'll marvel at their similarities:

Core, High, and Basic; autonomy and controllability; import substitution for key software; SaaS's rise in the cloud era; To B business's potential, with key industry budgets set to expand tenfold; the Internet of Everything, from IoT to connected cars and homes; information security's leap from hundreds of billions to trillions of markets.

Over time, concept packaging changes, but the concepts remain. Industry application solutions became "Industrial Internet" post-2018; "IoT" is rarely mentioned now, its replacement unclear. New concepts emerge, like AI, which swept A-share TMT sectors in 2023, revitalizing a languishing segment.

AI hype highlights A-share software firms' dilemma: not at the forefront of tech innovation, not major beneficiaries; not even representative of domestic software. Internet giants and Huawei's software departments are stronger, more advanced, and representative of productivity. A few exceptions exist, but most A-share software firms are inconsequential, even institutional investors admit, excluding them from "core assets."

Fundamentally, A-share software firms fall into five categories:

System integrators, essentially moving boxes for overseas IT giants (now for domestic ones too). Software and service outsourcing firms, developing and maintaining systems for overseas (now domestic) IT giants. Regional/industry-specific solution providers, relying on local or state-owned enterprise connections. SaaS and PaaS firms, with modest profitability and cash flow, facing steep competition from internet giants. Comprehensive software firms, a handful nearing size limits, struggling to grow market value.

The first three unlikely core assets; the fourth has potential, tied to Industrial Internet; the fifth, for cyclical concept and valuation hype. Analysts bet on "Cloud + Industrial Internet," envisioning Tencent/Alibaba-scale To B firms.

This vision shattered post-2021: A-share Industrial Internet firms' profits lagged when Tencent/Alibaba thrived; valuations tanked when they struggled. Amid Alibaba Cloud and Tencent CSIG layoffs, discussing Industrial Internet's future seems untimely.

The To B bubble's burst left the A-share software industry with little lifeline. Without Generative AI's 2022 emergence, the industry might've collapsed. But for software firms, the AI hype came and went fast; focus shifted to electronics and communication (chip computing power and servers). While supporting tech is right, the emphasis is on "hard tech," tightly integrated with high-end manufacturing. "Software," by definition, intersects limitedly with manufacturing (excluding MES solutions), at least per mainstream views.

Thus, the A-share software industry has reverted to over a decade ago. No, it's been stuck there, never progressing. I know analysts covering A-share software; once, I asked for new insights every few months. Now, answers are repetitive. So, I asked, "Is it 2024, 2019, 2014, or 2010?"

"What's the difference?" came the earnest reply.

"The Moon Over the Meadow" was a favorite childhood song; I've grown fond of it again, particularly the chorus:

"My heart is filled with melancholy, not for the crescent moon. But for today's village, still singing songs of yesteryear. Oh, moon of my hometown! Your crescent sadness pierces my heart."

The A-share software industry is that village trapped in time, forever singing the same songs. In truth, there are many such villages. I'm sure you have more answers, but space limits me. Those who understand, understand.