Chapter 075　What Is IR in the Age of AI?

IR is usually read as investor communications. That reading carries two misconceptions. The first is that IR is a craft of presentation. The second is that its success can be measured in the share price. Vol. VIII, Ch. 074 examined the work done on the investor’s side. This chapter examines the design work done on the company’s side. Future Value has moved to the center of any explanation of enterprise value, and disclosed information is now compared across hundreds of companies in moments. What should an enterprise put outside itself, at what grain, and at what frequency? We redesign IR as the practice of conveying Future Value.

1 The question — why it arises now

IR exists to reduce information asymmetry. People inside a company know more than people outside it. When the gap grows too wide, capital does not come. When it does come, it demands a higher return in proportion to the uncertainty. The cost of capital rises. IR was installed as a permanent function around the disclosure regime to narrow the gap. The design rested on one assumption. Most of enterprise value can be explained by the assets that appear on the financial statements. There are factories, inventory, and receivables. Profit comes out of them, and cash flow after that. If that holds, the job of IR is to explain the statements accurately. Explain the numbers, state the causes of variance, give the outlook for the next period. That is standard IR. As this volume has shown so far, the assumption no longer holds. The explanatory power of the financial statements has come apart from enterprise value. Brand, people, trust, innovation, the capacity to absorb AI — every one of them sits outside the statements. The indicators treated in Vol. VII, Ch. 065 and Vol. VII, Ch. 066 measure the size of that gap. They do not describe its contents. Beyond that gap lies Future Value. Future Value is not future profit. It is not the discounted present value of future cash flows. It is the capability to create value that does not yet exist. By definition, it cannot travel in the format of a report of results. It deals with what has not yet become a result. Then AI arrived. Disclosed material is no longer read only by human beings. Reading across hundreds of companies, matching each against its own filings from earlier years, and detecting shifts in wording can now be done at low cost. IR material becomes comparison material the moment it is written. So the question has moved. It is not “how do we explain this clearly.” It is “what do we leave outside the company, in a form that can be verified?” The first is a problem of expression. The second is a problem of design. IR has to move to the second.

2 Conventional answers and their limits

Three answers about IR circulate widely. Each is partly right. None can carry Future Value. The first answer: “IR is the activity of guiding the share price to a proper level” This is the most common answer. Our share price is low because we are not understood. So we thicken the explanation, add meetings, and polish the material. The instinct is natural, and it works to a degree. But set that as the objective, and IR becomes a function of the short-term share price. If the price falls, the explanation was insufficient. If it rises, the explanation worked. The share price moves on countless factors: the market as a whole, interest rates, supply and demand, other companies’ results. The portion IR can move is a small fraction of it. The objective also carries a side effect. The moment the share price becomes the goal, IR begins to select the stories that move the price. Stories that move the price are stories verified over short intervals. IR shortens its own horizon. The second answer: “IR is the activity of explaining clearly to investors” The second position values the quality of transmission. More diagrams, less jargon, a better narrative line. This is a sound direction. Information that is not understood may as well not exist. The limit is that clarity cannot substitute for verifiability. Talk about the future sounds more certain the more clearly it is told. But how something sounds and how likely it is are two different quantities. The pursuit of clarity often turns into the pursuit of assertion. And once AI joins the readership, the relative value of clarity falls. A summary of a difficult document is available instantly. What remains is what the document said, and what became of what it said. The third answer: “IR is the activity of complying correctly with the disclosure regime” The third position is the legal one. Disclose the prescribed items, at the prescribed time, on fair terms. In most markets, giving material information to some parties first is restricted by rule. This discipline is a precondition of capital markets and must never be loosened. But the regime is a floor. It requires only the information a market needs to function at a minimum. As Vol. IV, Ch. 039 argued, standardized disclosure raises the floor of corporate practice; it has no power to create a ceiling. Compliance is a premise, not a design. The structural bias the three answers share All three start from what has already happened. Conventional IR therefore carries three biases as a matter of structure. First, a bias toward reporting results. Recall how time is allocated at a results briefing. In most cases the bulk of it goes to explaining the period that has closed. The future is touched on briefly at the end, in the form of a medium-term plan. This allocation was not chosen deliberately. It happens because the only things that can be explained lie in the past. Second, a bias toward good news. IR material is written by the company itself. The writer always has an incentive to describe the projects that advanced and to leave out the ones that stalled. No bad faith is required. Write what is easy to write, and the bias appears on its own. The material ends up brighter than the company is. Third, a structure driven by quarterly explanation. When results are demanded on a short cycle, explanatory resources are pulled toward that cycle. Much of an IR team’s time disappears into variance analysis of the period just closed. Designing a long-term conception into verifiable form is always deferred. None of the three is a problem of individual competence. They are problems of design. A device built to report results cannot carry what has not become a result. The device itself has to be redesigned.

3 Redefinition — IR is the design that makes Future

Value verifiable and selects capital Future Value Theory redefines IR as follows. IR is the practice of continuously converting an enterprise’s Future Value Creation Capability into a form that outsiders can verify. Through that process, it is the design by which the enterprise gathers capital matched to its own time horizon. The definition has two parts. We take them in order.

3.1 The purpose, redefined — not to raise the share price

The purpose of IR is not to raise the share price. There are two purposes. The first is to obtain a proper valuation. Proper does not mean high. A valuation above the reality is a liability to the company. Excess expectation is corrected sooner or later. What is lost in the correction is not only the share price. Confidence in the words of the executive team is lost with it. Restoring that confidence takes far longer than restoring the price. The second is to gather shareholders matched to the company’s own time horizon. This is the function of IR that is least often discussed. Capital has time horizons. Capital that participates in daily price formation, capital that asks for results in a few years, capital that can wait a decade. No one of them is superior. Markets work because horizons of many lengths coexist. The problem is that companies frequently accept capital without ever communicating their own horizon. A firm declares a ten-year build of a new capability while its material is filled with quarterly variance. Capital that watches quarters gathers. To meet the demands of the capital it has gathered, management shortens its horizon further. Shareholder composition is not the result of management. It is often the cause of it. IR is therefore not the activity of receiving capital. It is the activity of selecting capital. Try to be understood by every investor, and you end up aligned with the shortest horizon in the room. Deciding who does not need to understand you is where IR design begins.

3.2 IR is not communications; it is the design of capital

IR is not a subordinate function of the communications department. Set out the Future Value Chain again. Purpose → Learning → Redefinition → Creation → Enterprise Value Enterprise Value comes last. IR appears to sit outside the final link of the chain. In practice it does not. The character of the capital IR gathers governs whether investment in Learning is approved. It governs whether Redefinition can be started at all. IR looks like the exit of the chain. It is the entrance to the next turn of it. First Principle 3 states the point. Capital Exists to Create Possibility. Capital exists to create possibility, not merely to maximize return. Gathering capital that cannot create possibility is not a success of IR.

3.3 What IR accumulates is Trust Capital

What is IR actually accumulating? Not the share price. Trust Capital. Future Capital = Financial × Human × Learning × Trust × AI × Knowledge × Ecosystem × Purpose In the Future Capital Equation, Trust stands beside Financial as one of the eight forms of Future Capital. This is multiplication, not addition. Let Trust approach zero, and the other seven terms cannot lift Future Capital however large they are. An abundance of financial capital cannot compensate for absent purpose, and advanced AI cannot compensate for absent trust. As Vol. VII, Ch. 068 argued, trust is the state in which verification can be skipped. In capital markets that state takes a concrete form. Investors do not have to verify the future the executive team describes from zero, every time. A company in that state is discounted less for the same conception. First Principle 8 names the property that matters most to IR. Trust Compounds Faster Than Capital. Trust compounds faster than capital and becomes the last durable advantage. Compounding means that time is on your side. It also carries an asymmetry. What grows by compounding is lost at once when it breaks. Trust accumulates in years and collapses in days. IR design is the deliberate, long-horizon accumulation of an asymmetric asset.

3.4 Make the VFI the language of IR

What should a company actually say when it conveys Future Value? Here the VURA Future Index (VFI) becomes usable. The VFI assesses an organization’s capacity to create future value rather than its current value, making visible what conventional financial statements cannot. As Vol. III, Ch. 026 set out, the equation the VFI corresponds to is this one. FVCC = Purpose × Learning × Redefinition × AI Integration × Ecosystem × Capital Allocation × Trust Seven terms, multiplied. The relationship is multiplicative: weakness in any single capability weakens the whole, so purpose, AI, and capital are each individually insufficient, and Future Value emerges only when all seven reinforce one another. Vol. III, Ch. 026 designed the index as an instrument of self-assessment. What we propose here is an extension of it. Use the seven terms of the FVCC Formula as the structural axes of the IR material itself. The consequence is not small. Conventional IR material is organized by business segment. A segment is a category that reflects the business structure of the past. Speak about the future in the categories of the past, and the range of what can be said is bound by the past. Organized by the seven terms of the FVCC Formula, the material becomes a discussion of capability. One caution. Following the third design principle of the VFI, the seven terms must never be combined into a composite score. IR is the setting where pressure to produce a single number is strongest. But the property of a product cannot be reproduced by a total or an average. What is disclosed is not a score. It is the current position of each of the seven terms, and the direction in which each is being renewed.

4 Structure — designing IR that conveys Future Value

Now bring the definition down to practice. Three things must be designed: what to disclose, at what grain, and at what frequency. Then we set out how to speak about the future in verifiable form.

4.1 What to disclose — divide along the three layers

Divide disclosure into three, following the three nested layers of value. For the first layer, Financial Value, follow the regime. Revenue, profit, cash flow. Here accuracy and comparability are everything. For the second layer, Enterprise Value (the middle layer of value), show the actual state. Customer relationships, the composition of the workforce, brand, the condition of trust. This territory overlaps with the non-financial disclosure frameworks. Follow the standard, then add the firm-specific elements the standard does not pick up. For the third layer, Future Value, show facts about capability and capital allocation. This is the center of the chapter. What should be conveyed is not the conception. It is the facts that support the conception. Specifically: how the composition of capital allocation changed from the prior year. How many projects were approved for which no payback period could be stated. Which definition of the business was discarded during the year. How many decisions were withdrawn because an assumption had been updated. Every one of these is a fact the company already holds internally. Each is hard to observe from outside, and none of them can be dressed up. The force of a document about the future is decided not by its description of the future, but by the density of facts of this kind.

4.2 At what grain — down to the smallest unit that can carry a

checkpoint There is one criterion. Go down to the smallest unit an outsider can verify. “We will engage in earnest with a new domain” cannot be verified. “Next year we will allocate this share of capital allocation to this domain” can be verified. “We will roll out AI across the company” cannot be verified. “This is where the time AI returned to us went” can be verified. Lowering the grain hurts. The lower you go, the more clearly a miss shows. But disclosure in which a miss never shows is disclosure in which a hit never shows either. A declaration that is never verified never gets the chance to build trust. Lowering the grain is not the same as releasing competitive secrets. There is no need to write the contents of a technology or the details of a deal under negotiation. What is written is the allocation fact — what the company bet on — and the observation point by which its progress is measured. The substance of the bet can stay concealed while the fact of the bet is made verifiable.

4.3 At what frequency — three cycles with different time

constants Match frequency to the speed at which the object changes. The cycles must not be collapsed into one. Financial results are disclosed on the cycle the regime prescribes. There is no choice here. Elements belonging to Enterprise Value, such as the customer base or the composition of the workforce, need a half-yearly or annual cycle. Measure them quarterly and what you are measuring is noise. Future Value Creation Capability needs an annual cycle. As Vol. III, Ch. 026 argued, capability does not move within a quarter. Measure something that does not move frequently enough, and you will read measurement error as a change in capability. High-frequency measurement also invites short-termism without exception. There is one exception. When a checkpoint set in advance is reached. This is disclosed by event, not by cycle. Reached, or not reached. If not, why. This irregular disclosure is the spine of IR that speaks about the future.

4.4 Three promises that make the future verifiable

The future is uncertain. How can something uncertain be made verifiable? The answer is to make verifiable not the future itself but the process of moving toward it. Three things are promised in advance. The first promise — publish the milestones in advance. If you say a new capability will be acquired over five years, publish at the same time what should be observable after one year for that to be on track. A standard built after the event is not explanation; it is retrofit. Only a standard set in advance works as a checkpoint. The second promise — state the assumptions. Every conception rests on assumptions. The market will move this way. The technology will advance at this rate. Regulation will stay within this range. Write them down. Written assumptions let everyone argue about what caused a miss. A conception with no stated assumptions can offer nothing beyond “our outlook was optimistic” when it misses. Stating assumptions has a second benefit. Management itself notices sooner when an assumption is going obsolete. The third promise — promise in advance to explain the misses. This is the most important of the three. Most companies fall silent about the items they did not hit. Silence is the most expensive choice available, because a shortfall is easy to observe from outside. The choice not to mention it is read as evidence of an organization that cannot handle failure. All three promises reduce the company’s own degrees of freedom. Reducing them is the point. Only a declaration that has lowered its own degrees of freedom can serve as collateral for trust. A declaration that can mean anything carries no collateral at all.

5 What it looks like in practice — how AI changes IR

We have set out the design. Now, what AI does to the practice. The change arrives from two sides: the company’s and the investor’s.

5.1 The company’s side — faster documents, and the trap in them

The first thing that happens inside the company is that producing material gets faster. Checking consistency against past documents, drafting variance analysis, preparing anticipated questions and answers, producing versions in several languages. AI performs all of this to a high standard. Much of the IR team’s working time is released. Here is the trap. As Vol. III, Ch. 026 argued, what to watch is where the time AI returned actually went. If the freed hours go into producing more material, Future Time has not increased. Page counts rise, briefings multiply, meetings multiply. The quality of the content is unchanged. If anything, consistency gets harder to hold as volume grows. Returned time should go into designing the three promises of 4.4. Where to place the checkpoints. Which assumptions to write down. How to explain a miss. AI cannot stand in for any of it. AI Optimizes. Humans Define. AI optimizes; humans define value, purpose, and direction. Deciding what to promise is a choice that carries responsibility.

5.2 The investor’s side — cross-sectional comparison gets

sharper On the investor’s side, the nature of analysis changes. Reading several years of material from several hundred companies to track shifts in wording was not economically realistic. It is now. Line up the same item across peers, match it against earlier text, detect the phrase that has disappeared. The cost has fallen sharply. Three things follow. First, the gap between words and reality becomes computable. How did a conception announced three years ago disappear from this year’s material? Which project stopped being mentioned? Work that used to depend on human memory is now performed mechanically. The quiet retreat stops being quiet. Second, the value of rhetoric falls. Highly abstract phrasing is treated as zero information in a comparison, because every company writes the same thing. Quantified checkpoints and stated assumptions stand out instead. What creates a difference is not skillful language but verifiable language. Third, consistency of disclosure becomes an asset. A company that publishes the same items, in the same framework, at the same grain, year after year, withstands time-series comparison. A company that changes its framework frequently is read as a company refusing comparison. The change of framework is itself a signal. In one line: AI lowers the efficiency of performance in IR and raises the efficiency of accumulation. The gain available from a single well-staged presentation shrinks. The gain from holding the same discipline for years grows. The property that Trust Capital compounds faster than capital works more strongly in the Age of AI, not less.

5.3 The IR that must not be done, and what it costs

Three practices should be avoided. Each works in the short term and is expensive over time. First, staging the numbers. Introducing a new indicator with a changed definition. Cutting out a convenient comparison period. Adding more adjusted profit concepts. None of these is illegal. That is precisely why they are widespread. The cost arrives in two stages. The first stage is loss of comparability. When definitions change every year, time-series analysis becomes impossible. A company that cannot be analyzed is discounted in proportion to the uncertainty. The second stage is loss of confidence. If a change of definition ever coincides with deteriorating results, every indicator afterward is doubted. The cost of one piece of staging does not stay with that one piece. Second, excess story. This is the state in which the narrative outruns the reality. A grand conception fills the first half of the material, and the facts that would support it are absent from the second. The form attracts attention in the short term. The attention it attracts brings capital with a short horizon. The cost is expectation running ahead of the company. As Vol. IV, Ch. 034 argued, enterprise value is a function of expectation. Expectation beyond the reality is corrected sooner or later. And the correction happens even when the reality has not deteriorated. A company that told too much loses its valuation without having failed at anything. Third, delaying disclosure of inconvenient information. This is the heaviest of the three. Carrying bad news over to the next briefing. Softening the wording, folding it in among other items. Even where the choice is permitted under the disclosure regime, it must not be made. The cost is the asymmetry of trust itself. Bad news that arrives late is judged not on its content but on the fact that it was late. Afterward, every piece of good news the company issues carries a reservation. As Vol. VII, Ch. 068 argued, trust collapses in days and takes years to restore. What do the three have in common? Each one draws down Trust Capital, a long-term asset, in order to raise a short-term valuation. Set out the Value Equation. Value = Purpose × Trust × Capability × Time Trust stands immediately after Purpose. This is multiplication. Because the relationship is multiplicative, value without purpose has no direction, without trust cannot spread through society, without capability cannot be realized, and without time cannot endure. Cutting Trust discounts Purpose, Capability, and Time together.

5.4 What it looks like in the executive meeting

Bring the abstraction back to the room. Picture the preparation of a results briefing. One business is running behind plan. The executive responsible proposes to lower the grain of the description by one step. Nobody is trying to lie. The judgment is only that there is no need to emphasize the point right now. That judgment injures no one on the day. But a company that had already published its checkpoints cannot make it. Design means creating a state in which you do not have to rely on the judgment of the day. Put in place, in advance, a mechanism that holds the discipline even on the day management loses its nerve. That is the design of IR.

6 Questions for the executive

The argument, in one line. IR is not an activity that moves the share price. It is the design that continuously converts Future Value Creation Capability into verifiable form and, through that discipline, gathers capital matched to the enterprise’s own time horizon. The share price is a result. A proper valuation appears as the outcome of accumulated verification. And shareholder composition is an accurate image of what the company disclosed, and how. Three questions to close. Each can be answered at your next board meeting. Question 1 — Of the statements in your most recent results material, what share can still be verified three years from now? In most companies the share is startlingly low. Explanation of the past and abstract declarations of direction fill nearly all of it. Statements that cannot be verified build no Trust Capital. Begin by counting the share. Question 2 — Does your shareholder composition match your own Future Horizon? If you hold up a ten-year conception while most of your shareholders judge on a horizon of a few quarters, that mismatch will show up in management decisions. It shows up as long-term investment proposals dropped on the grounds that they are hard to explain. The mismatch is not the investors’ responsibility. It is the result of the company’s own design — of whom it addressed, and what it put out. Question 3 — Among the plans you published three years ago, which shortfalls remain unexplained? This is the hardest of the three to answer. But in a period when cross-sectional comparison by AI is the norm, the gap will be observed. Speak of a new future while leaving an unexplained gap behind you, and that future is discounted before you finish speaking. Explaining the gap first is not a retreat. It is the fastest way to build Trust Capital. None of the three questions asks how to communicate. All three ask how to build so that you can be verified. AI compares disclosures. Investors allocate capital. Enterprises create the future. And trust runs through the whole of it. IR is not the craft of speaking about the future. It is the design that leaves responsibility attached to having spoken. Words that leave no responsibility do not accumulate, however skillfully they are made. Only words that leave responsibility grow faster than capital.

In brief

IR is the practice of continuously converting Future Value Creation Capability into a form outsiders can verify.
The purpose is not the share price. It is to obtain a proper valuation and to select capital matched to the company’s own time horizon.
Shareholder composition is not the result of management; it is often the cause of it.
What IR accumulates is Trust Capital. It builds over years and collapses in days.

Key concepts

Future Value Creation Capability / VURA Future Index (VFI) / Future Capital / Future Horizon

The chain of ideas

Purpose → verifiable disclosure → Trust → capital matched in time horizon → investment in Learning

Related first principles

Principle 8 — Trust Compounds Faster Than Capital. Principle 3 — Capital Exists to Create Possibility. Principle 4 — AI Optimizes. Humans Define.

Related chapters

Vol. VIII, Ch. 074 “What Are Investors Looking At?” — the process by which disclosure is read, taken first
Vol. VII, Ch. 068 “Does Trust Become Enterprise Value?” — why trust works as capital, argued there
Vol. III, Ch. 026 “What Is the VURA Future Index (VFI)?” — the source of the seven axes used to structure IR
Vol. VIII, Ch. 077 “What Is Capital Strategy in the Age of AI?” — how to match the time horizon of the capital gathered

Papers and companion volumes

Kadowaki, N. (2026a). Future Value Theory: A Management Framework for Enterprise, Capital, and Society in the Age of AI. VURA Working Paper Series. SSRN: https://ssrn.com/abstract=7120980 / Zenodo: https://doi.org/10.5281/zenodo. 21255662
Kadowaki, N. (2026b). Enterprise Redefinition: Toward an Enterprise Evolution Theory for the Age of AI. VURA Working Paper Series. (Published on Zenodo; under review at SSRN)
100 Questions on Management in the Age of AI, #080 “What Do Investors Look At in the Age of AI?” / #023 “Who Decides Enterprise Value in the Age of AI?”

Chapter 075 What Is IR in the Age of AI?