Executive Summary

For AI software stocks to move off the bottom, it is not enough to ask whether share prices have already fallen far enough. The more important question is whether the market has seen enough evidence that these companies can still create monetizable value in the AI era.

This article argues that the process requires two resets and one valuation threshold. The first is a value reset. Software companies need to prove that they still control core records, execution rights, and outcome responsibility, and that they can convert these capabilities into revenue, profit, and per-share value. The second is a shareholder reset. Shareholders who originally believed in the old growth story need to gradually exit, allowing new shareholders with different expectations and lower cost bases to take their place. The final condition is the valuation threshold. Even if a company is moving in the right direction, the stock price still needs to provide enough margin of safety to attract new capital.

Therefore, the key for AI software stocks to move off the bottom is not only about reducing selling pressure. It is the simultaneous emergence of new reasons for customers to pay, the absorption of old shareholder selling pressure, a reasonable price, and verifiable financial evidence. Only when AI usage turns into paid revenue, new revenue offsets pressure on the original business, and earnings expectations and per-share value stop being revised downward can the market begin to reprice the company on a new value basis.

Introduction

AI is creating two different forms of trust in the investment market.

For semiconductors, memory, data centers, networking equipment, and power infrastructure, the market can still understand how AI creates demand through capital spending, orders, supply constraints, and revenue growth.

For software companies, the question is much more complicated. Many software companies still have revenue, customers, and cash flow, and they continue to introduce AI features. Yet the market is no longer certain whether these companies can sustain their previous growth, pricing power, and valuations.

After share prices have fallen sharply, a natural question follows. What conditions do AI software stocks need to meet before they can truly move off the bottom?

Two Resets and One Valuation Threshold

Software stocks are not only facing a price correction. Their original value basis is also being questioned. At the same time, shareholders who believed in the old story will not all leave on the same day. Even after share prices have fallen sharply, the market still needs time to form a new ownership base.

For AI software stocks to move off the bottom, two resets may be needed.

• Value reset: The company needs to prove again that it still controls monetizable value in the AI era.
• Shareholder reset: Shareholders who believed in the old story gradually exit, while new shareholders take their place at lower prices and with different expectations.

Even if both resets begin to take shape, the stock still needs to pass one valuation threshold.

• Valuation threshold: Does the current price provide enough compensation for the uncertainty that remains?

The value reset asks whether the company deserves to be re-evaluated. The shareholder reset affects when the market may stop facing repeated pressure. The valuation threshold asks whether the stock is cheap enough today. These are three different layers of judgment.

The First Reset Is About Company Value

A value reset does not mean that a company has added AI features, or that management has started using new product language. A real value reset begins when the reason customers pay starts to change.

In the past, customers may have paid for per-user licenses, product features, and operating efficiency. In the future, software value may come less from the interface and individual features that users see, and more from the systems behind the enterprise that allow transactions, records, and responsibility to hold together. These systems include at least three capabilities.

The first is core records that enterprises can rely on

AI needs to know who the customer is, which contract version is currently valid, how accounting and transactions are recorded, and where a workflow stands. These are not ordinary text data. They are the records that enterprises actually use for operations, audits, filings, and decisions.

The second is execution rights

AI can make recommendations, but whether it can modify records, move funds, submit filings, approve spending, sign contracts, or initiate workflows is a different question. Real enterprise actions require identity verification, authorization scopes, internal controls, and audit trails.

The third is outcome responsibility

When AI makes a wrong judgment, who verifies the result, fixes the error, and carries the legal, financial, and professional responsibility? In work where the cost of mistakes is low, users may be willing to rely directly on general-purpose AI tools. But when the work involves taxes, funds, contracts, engineering safety, healthcare, or regulatory responsibility, customers are not only buying code and efficiency. They are also buying confidence in the result.

For this reason, the new value of some mature software companies may come from the combination of core records, execution rights, and outcome responsibility. But having these conditions does not mean that the value reset is already complete. To judge whether a value reset is beginning to take shape, it may be useful to ask four questions in sequence.

1. Necessity: After AI appears, does this system still need to exist?
2. Control: Does the company still control core records, execution rights, and outcome responsibility?
3. Value capture: Can the company convert this control position into revenue, profit, and pricing power?
4. Transmission to shareholders: Do these profits ultimately become per-share value, rather than being consumed by stock-based compensation, expensive acquisitions, or poorly timed buybacks?

These four questions are related, but they need to be judged separately. A software system can continue to exist while losing the user entry point. A company can control core records without being able to raise prices. A company can also create revenue and profit without translating that value into per-share value for shareholders.

When AI agents control the main entry point, existing software may continue to preserve core records and execution records, but users may log in less often, and the number of paid accounts may decline. Pricing may shift from high-priced per-user licenses toward lower-priced data access, transaction-based pricing, or usage-based pricing.

For this reason, being necessary and being able to create high economic value should not be treated as the same thing.

The New Story Has to Be Built by the Company

When the old software story stops working, a company cannot simply wait for the market to understand it again. A new story also does not emerge naturally just because an AI feature is added next to the existing product. The company may need to make new decisions.

• Is it willing to let AI products replace existing features?
• Is it willing to accept a decline in some per-user license revenue and shift toward usage-based, transaction-based, or outcome-based pricing?
• Can it move from being an operating tool to becoming a system of record and execution?
• Which capabilities can be handled in cooperation with foundation models, and which core records, execution rights, and model capabilities must remain within its own systems?
• Is it willing to disrupt its existing products before competitors do so first?

These choices involve more than technology. They also involve existing revenue, organizational power, and capital allocation. For this reason, a new narrative is not something a company simply says. It is something the company gradually builds through product choices, revenue structure, and capital allocation.

What matters, then, is not whether the company has launched AI features. It is why customers are willing to pay, and whether the company can retain the value created by that payment.

The Second Reset Is About the Shareholder Base

Even if a company begins to find new value, the stock may not move off the bottom immediately. Existing shareholders may have bought the stock based on growth in per-user licenses, subscription revenue, and high valuations. When these assumptions are challenged, they are not just facing one weak quarter. The original reason for owning the company is starting to break down. Some shareholders may leave immediately, some may wait for a rebound, and others may continue to believe that the company can return to its previous growth path.

As a result, even after the share price has fallen sharply, the market may still face continued selling pressure from high-cost shareholders. Each rebound may become an opportunity for existing shareholders to reduce their losses. Each new negative development may also cause another group of shareholders to give up waiting.

The market needs time for the stock to move from the hands of investors who believed in the old story to those who are willing to reassess the company under new conditions. New shareholders have lower cost bases, and their expectations for growth and valuation may also be more conservative. They do not necessarily need to believe that the company will return to the past. They only need to judge whether the current price already reflects enough uncertainty.

How to Observe the Shareholder Reset

A shareholder reset cannot be confirmed directly. It can only be inferred from volume and price reactions. Heavy trading volume alone is not enough, because the same short-term capital may trade repeatedly, and panic selling can also cause volume to spike. What matters is whether the price still falls as quickly as before after heavy selling pressure appears.

These clues cannot prove on their own that the bottom is complete. They can only help assess whether selling pressure is declining. For example, new negative news no longer causes the same degree of decline. When the stock tests its lows again, volume gradually shrinks. After breaking below the previous low, the stock can return to its earlier range. Rebounds come with higher volume, while pullbacks come with lower volume. When peers or the broader market decline, the stock no longer continues to lead the downside. Uncertainty remains, but the market no longer assigns a lower price.

These patterns may suggest that there are fewer shareholders eager to exit. But they can only indicate that sellers may be gradually exhausting themselves. They cannot prove that the company has found new value, and they cannot prove that the current price is cheap enough.

One Valuation Threshold

Even if a company’s new direction is worth recognizing, the current price may still not offer enough investment value.

A company may control core records, execution rights, and outcome responsibility. It may also have launched AI products with real usage. But if the stock price has already priced in the success of the new business, a company moving in the right direction still does not mean the stock price is reasonable.

The opposite can also be true. A company’s share price may have fallen significantly, and its valuation may be below historical levels. But if the original business model is still losing value, a lower valuation does not necessarily mean safety.

For this reason, the valuation threshold needs to be judged separately.

• Does the current price reflect short-term disappointment or permanent value decline?
• Even if the new business only succeeds partially, is the company still worth the current price?
• Has the market already priced in a successful transition?
• Are stock-based compensation, acquisitions, and capital allocation diluting per-share value?
• If the judgment is wrong, does the current price still leave enough margin of safety?

The valuation threshold is not only about whether the stock is cheap. It also affects whether the stock can attract new capital. Even if the company is moving in the right direction and old selling pressure is gradually declining, new investors may still have little reason to raise their bids if the price has already priced in a successful transition.

Understanding a company’s direction may only complete the industry analysis. Only when the price also provides enough margin of safety does the analysis become a more complete investment judgment.

Four States of Moving Off the Bottom

Using the value reset and the shareholder reset as the two main axes, today’s AI software stocks can be broadly grouped into four states. The valuation threshold is a separate condition that needs to be checked in each state.

Table 1. The Process Software Companies and the Market Are Going Through

The valuation threshold does not determine which state a company belongs to. It determines whether new capital is willing to enter at that point. Even if the value reset and the shareholder reset have both begun to take shape, the stock may still not be attractive enough if the price has already priced in a successful transition.

What Happens When the Two Resets Do Not Move Together

The following examples are not meant to determine the investment value of individual companies, nor to place any company permanently into one category. They are meant to show how the value reset and the shareholder reset can unfold at different speeds.

Intuit: The value reset may move ahead of market repricing

Intuit controls financial records trusted by both enterprises and individuals, tax and accounting rules, filing processes, and customer trust in the accuracy of results. In theory, this gives it a better chance than software that depends mainly on interfaces and features to turn AI into a new professional service and execution platform.

If AI agents move into transaction classification, tax preparation, accounting workflows, and financial advice, Intuit’s value may not be limited to its original software interface. The more important question is whether it can turn its existing data, rules, processes, and trust into new reasons for customers to continue paying.

But recognizing the company’s direction does not mean the current price is necessarily reasonable. It also does not mean that the original shareholder base has finished adjusting. The market is still reassessing Intuit’s past growth expectations, pricing power, and reasonable valuation. Higher trading volume may suggest that shareholder turnover is taking place, but it cannot prove that the process is complete.

This type of case shows that a company may have already begun to reposition its own value. However, the adjustment in price, valuation, and shareholder structure may not yet be complete. The industry direction may be right, while the investment conditions may still need more time to form.

DocuSign: Old expectations may be revised down first, while new value still needs to be proven

DocuSign’s original market expectations, which were built on high growth in electronic signatures, have fallen significantly. After the digital demand that was pulled forward during the pandemic began to fade, the market no longer assigned a high valuation simply because electronic signatures were becoming more widely adopted. The company has also been trying to move toward a fuller contract lifecycle management platform, shifting from a single signing tool to contract data, approval workflows, performance tracking, and post-signature management.

This direction is not without value. Contracts contain enterprise transactions, responsibilities, pricing, terms, and execution obligations. If these elements can be structured and embedded into daily workflows, they could indeed become important data assets. The question is whether the market can gain enough confidence that DocuSign can build sufficient control points, pricing power, and profit in this new area.

A contract database could become an operating platform that customers must rely on. It could also be a new defensive layer introduced after electronic signature growth slowed. The key issue is not only whether the new products are growing, but whether they can become core workflows that are difficult for customers to replace and that the company can effectively monetize.

Therefore, even if the share price and historical valuation multiples have fallen sharply, this does not directly prove that the company has completed a value reset. It also does not, by itself, prove that the current price provides enough margin of safety. Lower market expectations only reduce the threshold for disappointment. They do not automatically create a new reason for growth.

This type of case shows that price and market expectations can fall first. But if the company has not yet found a new reason for customers to pay, the share price may become more stable while still lacking the momentum for repricing.

When Does the Market Begin to Reprice

When the value reset, shareholder reset, and valuation threshold gradually begin to align, a stock moves closer to the conditions needed to move off the bottom. But true repricing still requires new evidence that changes the market’s expectations for the future.

A company launching AI features only proves that it is participating in the technology transition. It does not prove that the value reset is complete. What the market really needs to see is AI usage turning into paid revenue, new pricing models offsetting the decline in per-user license revenue, customers becoming willing to entrust more important core records, execution rights, and outcome responsibility to the platform, and new revenue ultimately flowing into profit and per-share value.

Only when this evidence gradually accumulates can the market begin to believe that the company’s value reset is not just a temporary story.

This change will also gradually appear in market behavior.

• Analysts stop continuously revising down revenue and earnings expectations.
• The company begins to stabilize or raise its financial outlook.
• AI revenue is no longer just a small-scale experiment.
• New revenue can offset pressure on the original business.
• Better financial results begin to attract new capital.
• The market becomes willing to evaluate the company under new growth and risk conditions.

At that point, the market is doing more than stopping the price from moving lower. It is beginning to reassess the company on a new value basis. Therefore, the real signal that a stock is moving off the bottom is not simply that there are fewer sellers. It is that company evidence, earnings expectations, and new buyer bids begin to move in the same direction.

Conclusion

For AI software stocks to move off the bottom, companies need more than a new reason for customers to pay. The selling pressure left by old shareholders also needs to be gradually absorbed, and the current price still needs to be attractive enough for new investors.

The value reset creates new reasons to pay. The shareholder reset reduces old selling pressure. The valuation threshold determines whether new capital is willing to enter.

But what truly leads the market to reprice a company is the gradual conversion of these conditions into verifiable financial evidence. When AI usage begins to generate revenue, new revenue can offset pressure on the original business, and earnings expectations and per-share value stop being revised downward, new buyers have a reason to raise the price they are willing to pay.

Therefore, the key for AI software stocks to move off the bottom is not only that the share price has fallen far enough, nor only that sellers are becoming fewer. What matters is whether the company can prove a new value basis, whether the market has absorbed the disappointment after the old story broke down, and whether new capital has begun to value the company based on a different future.

Note: AI tools were used both to refine clarity and flow in writing, and as part of the research methodology (semantic analysis). All interpretations and perspectives expressed are entirely my own.