Primer: Corporate Venture Capital Funds

We are often asked how corporate venture capital funds are structured – and how they differ from the mainstream venture capital funds we usually discuss in these primers. It is a fair question, and an underserved one. Corporate money is everywhere in the venture market in a limited partner capacity, but the dedicated literature on how corporate venture funds are actually built is surprisingly thin. This primer is our attempt to fill that gap.

First, a scope note, because the term “corporate venture capital” (CVC) gets used loosely. In this post, we are speaking of venture programs that a corporation itself conceives, organizes, capitalizes and controls – at least at inception. The corporation directs the strategy, houses (or closely tethers) the team, and is the dominant or exclusive source of capital, at least at first. We are not speaking of an independent, third-party venture fund that happens to have a corporation as an anchor limited partner. In that case, the corporation is a customer of the venture industry. Our subject here is the corporation becoming a participant in the venture industry – standing up its own investing operation, with its own name on the door.

A second scope note: Much of what gets announced as a corporate “fund” is not, legally speaking, a fund at all. When a corporation issues a press release announcing a “$500 million fund” to invest in some strategic area, the word “fund” is often doing a lot of work. Frequently, it describes an internal budget allocation – a commitment to write checks off the corporate balance sheet through a corporate development or venturing group – rather than a separate legal vehicle with committed capital, a general partner and a partnership agreement. Both models are legitimate. But they are structurally very different, and the difference matters, as we will see.

The corporate venture phenomenon is large by deal participation and small by fund count. By most industry counts, there are now more than 3,000 active corporate investors globally – a record – and rounds with corporate participation have represented roughly 45% or more of total US venture deal value over the past decade (though a much smaller share of deal count, since corporates cluster in larger, later rounds). And yet, dedicated corporate-sponsored fund vehicles remain, in our experience, a rarity – a very small fraction of the venture funds formed in any given year. Most corporate venture activity still runs off the balance sheet. The corporations that take the further step of forming a true fund are making a distinct, interesting set of structuring decisions. Those decisions are our subject.

At a high level, there are three questions that determine almost everything else about a corporate venture program:

• Whose money is in the fund? Is it 100% corporate balance sheet capital, or does the fund include executives, employees, selected outside strategic investors or, at the far end, institutional investors raised in the open market?
• How are the investment professionals compensated? Is it through salary and bonus like other corporate employees, some synthetic proxy for carried interest or true carried interest like a venture capital firm?
• How far from the corporate mother ship does the team sit? Are they employees in the corporate headquarters on corporate benefit plans or a separately branded, separately housed management organization that merely bears the corporation’s name and mandate?

These questions are related, but they are not the same. And, importantly, the answers are not fixed. A corporation may start at one point on each spectrum and migrate over time – often in response to exactly the pressures this primer describes. Structures are positions, not destinies. We will take each spectrum in turn. But first, we will provide some terminology and context on why corporations do this at all.

A note on terminology

A few terms recur throughout this post, and it is worth pausing to define them because readers of this primer may come from the corporate world, the venture world or both.

Balance sheet investing means the corporation (or a subsidiary) buys minority stakes in startups directly, using ordinary corporate funds, with no separate fund vehicle. The “program” is an internal team and a budget.

A fund, by contrast, is a separate legal vehicle – almost always a limited partnership or limited liability company – with investors (limited partners, or LPs) who commit capital and a general partner (GP) entity that controls the vehicle and is affiliated with a management company that employs the team. In a captive corporate fund, the corporation may be the only LP, and the GP and management company may be wholly owned subsidiaries. For more on how the GP and management company fit together generally, see our primer on structuring the general partner and management company.

A fund may have committed capital – binding capital commitments drawn down over time – with a fixed term (often 10 years, plus extensions), or it may be evergreen, meaning the sponsor replenishes capital periodically without a fixed fund life. Captive corporate funds are frequently evergreen or annually budgeted; funds with outside investors are almost always committed-capital vehicles with fixed terms.

A management fee is the periodic fee (classically, around 2% to 2.5% annually of committed capital during the investment period) that pays the team’s salaries and operating costs. In captive structures, the fee is often replaced with a cost-based budget. Carried interest (“carry”) is the share of investment profits – classically, 20% – allocated to the manager. We covered carry mechanics in depth in our primer on carried interest in private equity and venture capital funds, and we will lean on those concepts below. Synthetic carry (also called phantom or shadow carry) is a contractual bonus program designed to mimic carried interest economics without an actual profits interest in a fund – an important tool in the corporate setting, discussed below.

Finally, deal flow means the stream of investment opportunities a program sees – how many, at what quality and how early. Deal flow will turn out to be the hidden variable that connects all three of our spectrums.

Why corporations run venture programs at all

Corporations are the original strategic investors. The core rationale has been consistent for decades: A venture program is a window on the technologies, business models and talent that will reshape the corporation’s industry – ideally, seen years before those forces show up in the corporation’s income statement. The program lets the corporation identify, learn from, partner with, pilot with and sometimes ultimately acquire exciting young companies before its competitors even know they exist. A secondary (and sometimes coequal) rationale is financial return: Startups are an asset class, and a well-run program can be genuinely profitable. Most programs describe themselves as some blend of “strategic” and “financial,” and the exact blend chosen has structural consequences throughout this primer.

Done well, this strategy has paid off in spades over decades. Johnson & Johnson’s development corporation has been investing since the early 1970s. Intel Capital, founded in 1991, became one of the most prolific technology investors in history, with storied exits ranging across the software industry. The current artificial intelligence wave has minted a new generation of prolific corporate investors deploying capital at a scale that rivals the largest institutional firms.

Done poorly, the strategy has an equally long and well-documented failure mode. Corporate venturing has come in waves – a conglomerate-era wave in the 1960s and 1970s, a 1980s wave, a dot-com wave in the late 1990s (when new corporate units launched by the dozens each year), a post-2010 wave, and the current AI-driven wave – and each boom has been followed by a bust in which corporations shuttered programs, marked down portfolios and exited the field, often at precisely the moment patient capital would have been most valuable. Traditional venture capitalists have long been skeptical of corporate money for exactly this reason, deriding “tourist” corporate investors who arrive in bull markets and vanish in bear markets, leaving portfolio companies without expected follow-on support.

The academic research is instructive and worth internalizing because it previews the rest of this primer. Landmark work by Paul Gompers and Josh Lerner found that corporate venture investments perform at least as well as independent venture investments when there is genuine strategic overlap between the parent and the portfolio – but that corporate programs without a clear strategic focus are dramatically less stable, frequently shutting down after only a handful of investments. More recent research found that the median lifespan of a corporate venture unit has been only around four years and that nearly half of programs actively invest for three years or less – a striking figure when set against the 10-year-plus horizons of institutional venture funds. The same literature consistently identifies compensation and autonomy – the second and third spectrums below – as leading causes of early program death.

The lesson, which will echo through everything that follows, is that the corporations that succeed at this treat venture investing as a durable, structurally distinct discipline. The ones that fail treat it as a corporate initiative like any other.

The first structural question: Does the corporation even need a fund?

Before we reach the ownership spectrum, there is a threshold question: Why would a corporation investing entirely its own money bother to form a fund at all? If the corporation is the only investor, a partnership agreement between the corporation and its own subsidiary can look like ceremony. Plenty of respected programs – including some of the largest in the world – invest straight off the balance sheet, and some corporations run both models side by side (a formal fund vehicle for one strategy, direct balance sheet checks for another). But there are real reasons the fund structure keeps getting chosen, and they are worth spelling out because they explain features of the market you will see repeatedly.

Commitment and credibility. A fund with stated committed capital signals durability. Founders, co-investors and the venture community generally have long memories of corporate programs that disappeared mid-downturn or failed to support portfolio companies in follow-on rounds. A balance sheet program is, structurally, an annual budget line that a new chief financial officer can cancel; a capitalized fund is a multiyear commitment that survives budget cycles. Sophisticated founders and their institutional co-investors underwrite this difference.

Speed and delegated authority. Startup financings move on venture timelines, not corporate approval timelines. Industry surveys of corporate venture units consistently identify speed, corporate prioritization and bureaucratic decision-making as their biggest operating problems. A fund with an investment committee charter and delegated authority up to defined check sizes can commit in days. A program that must route each deal through corporate capital-approval processes often cannot – and the best founders, who have choices, notice.

Compensation plumbing. If the corporation ever wants to pay true carried interest – and, as we discuss below, it may eventually have no choice – it needs a vehicle with a waterfall. Carry is a profits interest in an entity; there must be an entity.

Ring-fencing and governance hygiene. A separate vehicle isolates liabilities, cleanly houses board seats and indemnification for fund personnel serving on portfolio company boards, and creates a defined perimeter for conflicts management and information walls (more on those below).

Measurement. A fund produces fund accounting – vintages, internal rates of return, distributions to paid-in capital – which imposes a discipline on performance conversations that strategic programs often lack. It is much harder to declare victory on “strategic value” alone when there is an auditable net return figure.

Accounting and tax. The choice between direct balance sheet holdings and a fund vehicle can meaningfully affect how the portfolio runs through the parent’s financial statements (including fair-value marks and earnings volatility) and how gains are characterized and shared. These are specialist questions – the corporation’s auditors and tax advisors should be in the room early – but they are frequently among the actual deciding factors, and structure should be chosen with them in view rather than retrofitted.

Optionality. Perhaps most important for this primer: A fund can evolve. A fund can later admit executives, employees, friendly outside strategics or institutional investors. A balance sheet program cannot admit anyone; it must first become a fund. Nearly every migration story we describe below begins with, or requires, this conversion.

With that foundation, we can turn to the first spectrum.

Spectrum one: Whose money is in the fund?

Think of ownership as a spectrum running from fully internal to fully external, with several distinct, recognizable stops along the way. Each stop has its own legal architecture, and – critically – each stop changes the duties owed, the documents required and the degree of freedom the corporation retains.

Stop one: 100% corporate capital. At this end, the fund is a wholly captive vehicle: The corporation (or a treasury subsidiary) is the sole limited partner, and wholly owned subsidiaries serve as general partner and management company. The partnership agreement can be short because it is essentially an intercompany document; the real “constitution” of the program is the investment committee charter and investment policy adopted alongside it. These vehicles may be evergreen or organized in successive committed funds. A committed, numbered fund series – Fund I, Fund II, Fund III – is itself a market signal, even with a single LP: It tells founders the capital is real, dedicated and multiyear. A well-known public example of the fully realized version is BMW’s venture unit, which spent roughly its first five years investing off the corporate balance sheet and then, in 2016, converted to a conventional GP/LP fund structure with the parent as the sole limited partner. Hallmarks of the model include successive dedicated funds, an autonomous investment committee, offices in the venture markets rather than corporate headquarters and a stated policy that portfolio companies are never required to do business with the parent. That is single-LP capital run with independent-firm mechanics, and it previews where the second and third spectrums are headed.

Stop two: Corporate capital plus insiders. The next stop is still, for all practical purposes, internal: The balance sheet remains the overwhelming source of capital, but a handful of senior insiders – the CEO, other key executives, sometimes founders or board members – invest personally alongside the corporation. The motivations are alignment and engagement: Executives with personal capital in the fund pay a different quality of attention to it, and participation can function as a senior retention perk. Legally, this is usually straightforward. Even a wholly captive single-LP fund relies on a private fund exemptions under the US Investment Company Act – typically, 3(c)(1) or 3(c)(7) – but with the corporation as the only beneficial owner, the exemption analysis is trivial. Once insiders come in, that analysis takes on real content, though a small group of senior, accredited (often “knowledgeable employee”) participants is comfortably manageable within the standard exemptions that every venture fund uses, provided counsel is disciplined about who may participate, in what capacity and through what kind of holding arrangement.

Stop three: The employees’ securities company. Open participation more broadly – beyond a handful of executives to a wide population of employees – and the standard private fund exemptions run out of room because the most commonly used exemption caps a fund at 100 beneficial owners. There is a purpose-built, if niche, solution: the employees’ securities company, a special category under the Investment Company Act for investment vehicles owned by the employees (and, in defined cases, former employees and certain family members) of a single employer group. The regime requires an exemptive order from the SEC – an actual application-and-order process, with conditions tailored to protect employee investors – and, once granted, permits participation by hundreds of accredited employees, far beyond what an ordinary private fund could accept. In practice, these programs are typically limited to the employees themselves as individuals (estate-planning entities and outside money are generally off the table or tightly constrained), and eligibility is often restricted to senior or accredited populations. Employees’ securities companies are a fixture at large financial institutions and professional services firms – investment banks, asset managers, consulting, accounting and law firms have long maintained them as internal investment programs – and they represent a genuine, if specialized, stop on the corporate fund ownership spectrum. Standing one up is an art form of its own and well beyond the scope of this primer; the point for now is simply that a corporation can open its venture fund to a broad internal population, but doing so requires a distinct regulatory architecture, not just a longer signature block.

Stop four: Corporate majority plus curated outside strategics. Here we cross a meaningful line: the first genuinely outside dollar. The corporation remains the majority investor and the sponsor, but it invites a small, carefully selected group of other corporations to invest as limited partners – typically, players in the same ecosystem who share the strategic goal but do not compete with the sponsor (or, ideally, much with each other). We have seen, for example, a major manufacturer sponsor a fund whose other limited partners were a handful of its major suppliers: All of them wanted a window on emerging technologies in their shared value chain, none of them was strategically threatened by the others’ presence, and the manufacturer retained control of the vehicle and of who could ever be admitted. The consortium logic is powerful where an industry’s players are structurally noncompetitive with one another. The same dynamic explains why coalition-style energy investment platforms have been able to assemble dozens of utilities as strategic investors in a single program: Because utilities are largely territorial, they can share a deal-flow window without sharing customers. The best known of those platforms, however, are independent firms built around corporate LPs and thus, strictly speaking, fall outside this primer’s captive scope. Structurally, stop four is where the documents grow up: Once outside investors are in the vehicle, the sponsor owes them duties, the partnership agreement becomes a real negotiated instrument, and issues like reporting, conflicts, allocation of deals between the fund and the sponsor’s own balance sheet, information walls among the strategic LPs, and admission controls (who else may ever be let in) all need real answers. We return to these in the documents section below.

Stop five: The corporate-branded market fund. At the far end of the spectrum, the vehicle is a true fund, raising capital from many investors – institutions, funds of funds, family offices – marketed in the open market like any other venture fund but leveraging the brand name, team, track record and deal flow that arose from the mother ship corporation. The corporation typically remains the sponsor-anchor: often the largest single investor, with its name still on (or associated with) the door, and with negotiated strategic rights. The market has produced a steady stream of these. A large pharmaceutical company’s venture arm, formed in the 1980s, spun out in 2020 and promptly raised a $500 million first independent fund – oversubscribed at its hard cap – with the former parent as the largest investor alongside institutional asset managers, endowments, foundations, pensions and family offices, followed by an even larger second fund. A major telecommunications company’s venture unit, built inside the parent from 2011 with a string of marquee exits, achieved independence when a global private markets firm purchased a significant minority stake and the two anchors committed side by side to the successor fund. The former venture arms of a major enterprise software company and a global bank each became independent, multibillion-dollar firms that still trace their brand and origin story to the parent. In each case, the corporation converted a captive cost center into an asset: It monetized or leveraged the franchise, kept strategic access as anchor investor, and – not incidentally – solved the compensation problem because a market fund pays market economics.

Movement along the spectrum. Corporations do not always just pick a stop; they frequently travel. The most common direction of travel is outward – from captive toward external – and the single biggest engine of that migration is the one we take up next: the historical inability of corporations to retain first-rate investing talent without paying like a venture firm. When a corporation concludes that it must pay carried interest to keep its team, one natural reaction is to pivot the whole program toward a more traditional fund structure – outside capital, market economics – while leaving the corporation’s name attached and its strategic access intact. But the road runs both ways, and the decision is genuinely strategic rather than cosmetic. In one of the most watched corporate venture stories in recent memory, a major semiconductor company announced in early 2025 that it would spin off its storied, multibillion-dollar venture arm into a stand-alone firm – rebranded, free to raise external capital, with the parent as anchor investor. However, within months and under new chief executive leadership, the company reversed course, electing to keep the unit in-house, harvest the existing portfolio and invest more selectively. Same company, same team, two different answers within a single year. The spectrum is real, and where a corporation sits on it is a live strategic decision, revisited as leadership, balance sheets and markets change.

A necessary digression: Deal flow

Before we can talk sensibly about compensation – the second spectrum – we have to digress, briefly, on deal flow because it is the variable that determines how hard the compensation problem will be.

Some corporations simply attract venture deal flow by virtue of who they are. They are famous, admired and situated in a sector dense with startups whose founders grew up knowing the corporation’s name and affirmatively want its money, its platform or its blessing. Think of the prominent companies in crypto, developer infrastructure and AI: Founders in those ecosystems seek them out, unprompted, in volume. For a corporation like that, the venture program’s job is substantially processing – evaluating, selecting, negotiating, supporting and exiting the opportunities that arrive – rather than sourcing.

Other corporations get no such gift. A large industrial company may have an entirely legitimate, high-conviction strategic need – say, a window on emerging battery chemistry or advanced materials – in a space where there simply are not a million startups and where the founders who do exist are not culturally wired to cold-call an industrial conglomerate for capital. That corporation will not receive deal flow; it must go get it. And going to get it is precisely the skill of the professional venture capitalist: building networks over years, being present where companies are formed, earning a reputation that causes the best founders and the best co-investors to make room in competitive rounds. It is slow, relationship-driven, personal work – and it walks out the door at night.

One refinement worth stating plainly: Even for the deal-flow magnets, inbound volume is not the same as access to the best opportunities. The strongest founders choose their investors, and they choose based on reputation – for speed, for reliability in follow-ons, for behaving well in hard moments. So, every program has some stake in professional venture craft. But the difference in degree is enormous, and it drives everything in the next section. A program whose job is processing abundant inbound flow can be staffed and paid one way. A program whose job is manufacturing proprietary access in a startup-scarce sector must be staffed with genuine venture talent – and genuine venture talent knows exactly what it is worth.

Spectrum two: Compensating the team

Let us give away the punchline first because the history is the argument. Over the past 20, 30, 40 years, corporate venture programs have suffered chronic, well-documented turnover, and the fact pattern repeats with remarkable consistency: The corporation staffs a program, pays its people like corporate employees, and watches them leave after a vintage or two – to a venture firm across the street that pays carried interest or to found a fund of their own on the strength of the track record they built with the corporation’s money. The corporation becomes, in effect, a training academy for the venture industry. It absorbs the losses of the learning years and donates the seasoned professionals. The revolving door is not a corner case; researchers studying why the median corporate venture unit historically survived only about four years put incentive design near the top of the list, and industry surveys find that only a minority of corporate venture units – on the order of a third – offer any form of carried interest at all, even as those same units recruit the majority of their investment professionals from venture, private equity and other corporate venture firms where carry is the norm. Hiring from a carry-paying talent pool while offering a no-carry package is a structural leak. Eventually, it drains the tank.

And yet – interestingly and importantly – plenty of corporations have not experienced the revolving door and have retained venture teams happily for decades on salary and bonus. In our experience over a couple of decades, two things reconcile these observations, and both run through the deal flow digression above.

First, there is a real category of programs where sourcing is not the job. The corporation attracts abundant inbound flow; the team’s work is analysis, execution and portfolio support; and the professionals are content – often genuinely, durably content – with a strong salary and a bonus that serves as a rough proxy for success without being literal carried interest. Layer in the real attractions of corporate life (stability, benefits, lower fundraising stress, a powerful platform) and that model can hold indefinitely. There are also individual outliers: excellent network-builders who simply prefer the corporate environment and stay for reasons money does not capture. Those cases exist. They should not be planned on, but they exist.

But where the opposite conditions hold – the corporation does not attract its own flow and is relying on charming, well-networked individuals to manufacture access – the flight risk after a vintage or two is very high if the corporation does not begin to compensate like a venture firm. The individuals’ networks are portable. Their track record is portable. The market across the street prices both daily.

Why this is so hard inside a corporation. The difficulty is not stinginess; corporations pay large sums for talent all the time. The difficulty is one specific attribute of carried interest: It is unlimited by design. Carry is a percentage of investment profits, and investment profits have no ceiling. When you grant a venture team real carry, you are telling the CEO, on day one, that these new hires down the hall might someday outearn her – might, in a great vintage, be the highest-paid people in the entire enterprise. Some corporate cultures, compensation committees and disclosure postures can absorb that. Many cannot. Corporate compensation systems are built on benchmarking, internal equity, banding and predictability; carry violates all four on purpose. That is the collision, and there is no drafting trick that makes it disappear. If the corporation truly values the strategic benefits of an in-house program, lacks natural deal flow and therefore needs professional sourcing talent – that collision is the price of admission, and it has to be reconciled somehow rather than wished away.

The menu. In practice, the market has developed a graduated menu, and it is worth walking through it in order because each step trades a little corporate comfort for a little more retention.

• Salary and bonus. The team participates in the corporation’s ordinary annual compensation and bonus cycle, perhaps with long-term incentive awards in parent equity. Simple, culturally frictionless and – for processing-oriented programs with strong inbound flow – often sufficient.
• Success-linked bonus. Same architecture, but the bonus formula is explicitly tied to program outcomes – realized gains, portfolio marks, strategic milestones. A gesture toward alignment, still capped and committee controlled. Better, but no professional mistakes it for carry.
• Synthetic carry. Also called phantom or shadow carry, synthetic carry is a contractual bonus plan engineered to track what a carried interest would have paid – often computed deal by deal or on a notional fund waterfall – but paid (and taxed) as compensation. Industry surveys suggest that among corporate venture units organized as separate legal entities, a substantial majority now offer some synthetic carry arrangement, and it has become the standard middle answer. Understand what it is and is not. It is a serious retention tool that lets the corporation cap, vest, defer, condition and claw back the payout and keep the whole arrangement inside the compensation system. It is not economically or tax-equivalent to real carry. Payments are ordinary compensation income to the recipient, rather than receiving the capital-gains treatment that real carry may produce (see our carried interest primer on allocations of long-term capital gains, the three-year rule and related concepts). In addition, the plan must be carefully designed to comply with the deferred compensation tax rules, and every cap and condition the corporation adds to make the plan internally palatable subtracts from the very retention value it was built to provide. Sophisticated candidates model the difference precisely.
• Real carried interest. Real carry is an actual profits interest in an actual fund. As noted above, that requires the existence of an actual fund. Real carry has the full retention and alignment power of the venture model and potential capital-gains character for the recipients. In addition, in the right circumstances, individual real carry holders may access qualified small business stock benefits on fund gains that the corporate parent, as a corporation, could never claim on its own holdings – a quiet structural nuance worth flagging to your tax advisors. Real carry also has the full cultural cost described above, plus internal mechanics that need real design: vesting and forfeiture, clawbacks, allocation among the team, and treatment of departed members. (See our recent primers on planning for and handling senior partner transitions and departures; every issue there arrives inside a corporate program, too, with a parent company watching).

The cap trap. To make the numbers survivable in the executive suite, corporations granting real or synthetic carry are perpetually tempted to haircut it: capping the payout, taking a large share of the carry back to the corporate sponsor or setting the team’s percentage well below market. Some of that is fine and, indeed, customary: A corporate sponsor that is putting up all the capital, the brand and the deal flow can legitimately keep a meaningful piece of the economics, just as it might if it were seeding any manager. But there is a threshold below which the exercise defeats itself. Every dollar of carry clawed back to soothe the org chart is a dollar of retention spent. Take too much off the table and you have paid all the structural and political costs of a carry program while recreating the exact flight risk the program was built to solve. The team will do a vintage on your discounted economics, then go get whole ones. The art is landing in the middle zone, with the team keeping enough genuinely uncapped upside that the best people stay while the corporation keeps enough economics and control that the arrangement can be sold internally. The internal sales pitch, candidly, is usually some version of: “Yes, they may outearn the CEO – but they are not really ‘in’ the corporation anymore.” Which brings us directly to the third spectrum.

Spectrum three: How far from the mother ship?

Compensation and operations travel together. The way a corporation makes venture-scale pay politically survivable is, very often, physical and organizational distance – and the way it preserves strategic value across that distance is contract. Operations, too, run along a spectrum.

Fully in-house. At one end, the venture team consists of corporate employees in every respect: on the corporate benefit plans, in the corporate offices, on the corporate systems and policies, reviewed in the annual cycle, paid through the ordinary compensation process (perhaps with a success-flavored bonus), with an investment committee that includes senior corporate executives and a budget that lives inside a business unit or the corporate development function. For a corporation with strong inbound flow and a processing-oriented mandate, this works, and it has real advantages: The team is deeply plugged into the business units, where strategic value is actually created through pilots, commercial agreements and technical diligence drawn from thousands of in-house engineers. The costs are the familiar ones: slower decisions, corporate-cycle risk and a founder-facing perception problem. (Startups and their institutional co-investors can tell when they are dealing with a committee.)

The hybrid middle. Most real programs live somewhere in the middle: a dedicated, separately branded unit; delegated investment authority up to defined thresholds; a ring-fenced multiyear budget or a formal single-LP fund; team members still employed by the corporate group but managed outside the ordinary band structure; synthetic carry. This is the pragmatist’s zone, and thoughtfully built, it captures much of both worlds.

Arm’s length. At the far end, typically reached precisely because the corporation has concluded it must pay real carry and needs somewhere culturally survivable to put the people earning it, the corporation pushes the team out. The team gets its own office away from headquarters, usually in the venture markets. It operates through its own management company, running its own profit and loss on a management fee (budget- or percentage-based) paid by the fund. It receives real carried interest and develops its own brand identity, or operates under a licensed use of the corporation’s name. From the inside, the team now looks, feels and pays like a venture capital firm. But the corporation has not simply become a passive LP in somebody’s fund; it has kept its hands on the specific levers that preserve strategic value:

It controls admission. The corporation typically retains consent rights over which other limited partners may ever be admitted, enabling it to keep strategic competitors permanently out of the vehicle – out of its deal flow window, out of its information, off its cap tables. This is a right no corporation gets by investing in a genuinely third-party fund, and it is one of the most valuable and least discussed features of the sponsored model.

It keeps the name in the market. The corporation’s brand remains attached to the fund, which both drives deal flow to the program and keeps the corporation visibly present in its innovation ecosystem.

It keeps a professional home for its own flow. Whatever inbound opportunities do arrive at the corporation now have a dedicated, professional team to evaluate, nurture, govern and exit them – rather than dying in a business unit inbox.

It often keeps economics. The sponsor frequently retains a share of the carried interest or a stake in the general partner – its compensation for lending the brand, the anchor capital and the origination platform – subject always to the cap trap discussed above.

It keeps negotiated strategic access. That includes co-investment rights for the balance sheet, information and reporting rights (calibrated against the walls discussed below), and covenants about the fund’s strategic focus.

At this point, the arrangement starts to look a great deal like simply investing in a third-party strategic fund – and that resemblance is the point. The corporation is deliberately converging on the independent model while retaining a defined premium over it: the name, the admission control, the house for inbound flow and a closer seat than any ordinary LP gets. Whether that premium justifies the sponsorship burdens is the perennial board-level question, and it is exactly the question the reversing semiconductor company answered twice, differently, in one year. The structural menu for engineering that retained premium – running from light contractual rights all the way to an equity stake in the manager itself – is taken up in the documents discussion below.

Two footnotes on this spectrum. First, a corporation need not build or retain the team at all. There is now a cottage industry of “CVC-as-a-service” firms that will supply the professional venture team as an outsourced manager for a corporate-sponsored fund, trading depth of integration for speed and simplicity. And – more rarely but more interestingly – the outside manager is sometimes not a service shop but a first-tier independent venture firm in its own right. In one recent, well-publicized example, the corporate AI investing arm of one of the world’s largest technology companies partnered with a marquee venture firm to co-invest, check for check, in early-stage AI companies sourced through the venture firm’s own seed program, with the corporation adding capital, compute and technical access alongside the firm’s sourcing engine. Why would an elite independent firm lend its platform to a corporate program? Where the corporation’s strategic zone sits adjacent to the firm’s franchise, rather than on top of it, the answer is straightforward: lateral reach into new territory, added credibility and resources for its founders, and incremental economics – none of it at the expense of the core brand. The corporation, for its part, is buying instant access to a first-rate sourcing network, the very asset this primer keeps insisting is the hard part to build. Second, the operations spectrum is where a captive program quietly accumulates the assets that make an eventual spinout possible: a separable team, a separable brand, a track record attributable to identifiable individuals and a portfolio that can anchor a first independent fund. Corporations should decide deliberately whether they are building those assets – because they are building them either way.

What the documents actually look like

Having drawn the spectrums, we can describe the paperwork because the documents change character at each stage. Readers who want the general architecture should start with our primers on carried interest and on structuring the general partner and management company. Here, we focus on what is different in the corporate setting.

The captive stack. A wholly captive single-LP fund is documented with a short-form limited partnership agreement (there is no arm’s-length negotiation to memorialize and no offering to run, so there is no private placement memorandum or subscription book to speak of), a shell general partner entity, and a management company that is either a corporate subsidiary or simply a cost center. The economically real documents are elsewhere: the investment committee charter and investment policy, which function as the program’s true constitution by governing scope, check sizes, delegated authority, follow-on reserves, conflicts and recusal rules; the budget or expense-reimbursement arrangement that substitutes for a management fee; intercompany agreements covering services, secondment, systems and use of the corporate name; and compensation-plan documents, where any synthetic carry plan typically resides, drafted with the deferred compensation rules in mind. One structural weakness deserves flagging because founders price it: A captive fund is terminable at the sponsor’s pleasure. Committed capital, a numbered fund series and a public track record of following on are how captive programs manufacture the credibility that structure alone does not supply.

The moment outside money arrives. The first noncorporate dollar transforms the stack. Now there is a real, negotiated LPA: management-fee provisions (which, in sponsored funds, are often discounted, budget-based or subsidized by the sponsor), real carry and a real waterfall, an LP advisory committee, transfer and withdrawal mechanics, the fund term and any extensions – and, sometimes, key person provisions. Sometimes, because the corporate setting scrambles the usual logic, a key person clause exists to suspend or unwind a fund when the specific individuals whose networks and judgment the investors underwrote stop running it. Where the outside LPs have been persuaded, as corporate sponsors sometimes actively market and as is sometimes genuinely true, that the deal flow belongs to the corporation itself rather than to any individual, the clause’s premise falls away. The network identifier is the institution, and the institution is not going anywhere. A fair number of corporate-sponsored funds accordingly close with no key person provision at all. Where the program is sourcing-dependent, however, LPs will insist on one. When the key persons are or were corporate employees, the resulting clauses require careful drafting – and can still read strangely.

The corporate setting, moreover, adds a distinctive layer ordinary funds do not have. There is the allocation policy: When an opportunity fits both the fund and the parent’s balance sheet, whose deal is it? Outside LPs will insist this be answered in writing, in advance. There is also the conflicts protocol for the parent: What happens when the parent wants to acquire, partner with or compete with a portfolio company? The answer may include recusal mechanics, information-handling procedures, and in some cases, notice or matching frameworks. It always requires a fairness architecture because the sponsor sits on both sides. Information walls, meanwhile, run in both directions: among the strategic LPs, who may tolerate one another but not one another’s access to their data, and between the fund and the parent’s business units. Portfolio companies will share sensitive information with their investor that they would never hand to the parent’s product teams, so the fund must be able to promise, credibly and contractually, that the wall holds. There is also competitor exclusion: defined-list covenants restricting admission of the sponsor’s named competitors as LPs, which outside institutional investors will generally accept even when they would never accept an open-ended sponsor veto over fund activity. And then there is the sponsor’s own side letter, covering matters such as anchor economics, co-investment rights, capacity rights in successor funds and brand-license terms.

The spinout stack. When a program takes the final step to independence, a third family of documents appears, and these transactions are genuinely intricate. They typically cover the trademark or name license (if the corporate brand travels with the team, including quality controls and termination triggers); transition-services agreements; transfer or warehousing arrangements for existing portfolio positions moving into the new fund (including pricing, consents and tag rights); rules governing use of the track record, including who may claim it, how it must be presented, and the LP and regulatory sensitivities surrounding performance attribution; employment-transition, benefits and restrictive-covenant arrangements; and the former parent’s anchor commitment, together with its negotiated economics, including fee breaks, capacity rights and, in some cases, a retained stake in the new manager or a share of carry. Done well (and the well-publicized examples were), the spinout leaves both sides better off: The corporation converts a cost center into an anchor-LP position plus retained economics, and the team gains the market structure that finally solves the compensation problem for good.

The retained stake. One item on that list deserves unpacking because it is unfamiliar to most corporate readers, and it is the piece corporations most often under-engineer. When the team spins out, the corporation stops owning the management business. But it rarely wants to drop overnight to the status of just another limited partner. It supplied the brand, the training ground, the track record and the anchor check, and it still wants ongoing influence over how the franchise is run, visibility into what the franchise sees and, often, a share of what the franchise earns. The retained stake is our shorthand for whatever continuing position in or around the new manager delivers those things, and it comes in a handful of distinct legal forms. (The same menu applies, incidentally, when a corporation wants hooks in an established independent manager it is backing from the outside. Strictly speaking, that is the adjacent case excluded by our opening scope note, but the machinery is identical.)

The forms sit on a spectrum of their own, and the trade is always the same: Each step up buys more influence and more durable economics at the price of more entanglement, more liability and heavier regulation. The lightest form is purely contractual: The corporation owns nothing and instead holds a bundle of agreements, including a strategic advisory role, enhanced information and reporting rights, co-investment rights for the balance sheet, windows to see deals the manager sources, consent rights over a short list of fundamental changes, and a favored seat on the LP advisory committee. Contractual rights are cheap to build, easy to unwind and nearly invisible to regulators. However, they are only promises, enforceable like any contract and no further, and every special right the corporation extracts tends to prompt the fund’s other limited partners to demand matching treatment. A step up is a seat at the decision table: representation on the manager’s investment committee, as a voting member or – more often – a nonvoting observer, occasionally with defined veto rights over the categories the parent cares about most, such as deals above a particular size threshold or in the parent’s competitive zone. These seats need careful engineering, including recusal protocols when a deal conflicts with the parent, clarity about whether the corporation’s designee acts personally or as the corporation’s agent, and often, an agreed sunset at the end of the investment period. They also carry a trap worth naming: A corporate designee who actively steers investment decisions can start to look, in the law’s eyes, like an unregistered investment advisor to the fund, with the fiduciary exposure that accompanies the label. That risk is a large part of why so many of these seats are deliberately nonvoting.

The heavier forms involve actual ownership. A minority equity stake in the general partner or management company gives the corporation board representation, enumerated veto rights, and a pro rata share of the management fees and carry, protected by the usual minority-owner machinery: anti-dilution, transfer restrictions and buy-sell exits. It is the most durable hook because it is built into the manager’s own constitution rather than merely promised in a contract. It is also the most entangling because the corporation now shares responsibility to fund investors it did not choose, and its conflicts policies can collide with the fund’s freedom to invest. Even a change of control at the corporation becomes the fund’s problem. At the far extreme, the corporation’s own entity can be named a co-general partner of the fund itself. That structure is rare, and the reason can be stated in one phrase: joint and several general partner liability sitting on a corporate balance sheet, usually alongside an advisor registration obligation and the daily friction of two-headed governance.

In practice, negotiated outcomes blend elements from more than one of these forms. The design question is the one running through this whole primer: How much influence does the corporation genuinely need, and how much liability, entanglement and regulatory weight is it willing to carry to get it?

The regulatory map, in plain English

A corporate venture fund lives under the same regulatory regime as any private fund, plus a few corporate-specific overlays. What follows is not advice but rather a high-altitude map – and every item on it has changed at least one deal we have seen.

Investment Company Act. A fund that is not just a balance sheet budget must fit a private fund exemption – classically, the 100-beneficial-owner exemption or the qualified purchaser exemption – with the knowledgeable-employee concept doing quiet work when insiders participate. Broad employee participation would generally cause the fund to fail these classic exemptions and require pursuing the employees’ securities company route described above, with its SEC exemptive order.

Investment Advisers Act. A manager advising only a wholly captive vehicle for its corporate parent generally has no third-party clients and, as a practical matter, typically stays outside the registration regime and the “investment adviser” definition altogether – one of the underappreciated simplicities of the captive model. Admit outside investors, and the manager is likely managing a “private fund.” Most corporate-sponsored venture managers then rely on the venture capital fund advisor exemption. That exemption imposes real conditions on what the fund may hold and how it operates, including 80% in equity of private operating companies, minimal leverage, no redemption rights and holding itself out as a venture strategy. Alternatively, managers may rely on the sub-$150 million private fund advisor exemption, becoming exempt reporting advisors subject to public Form ADV filings and compliance obligations to a subset of the Investment Advisers Act provisions.

Securities offerings. Interests in any fund with outside (or employee) investors are securities, offered under the usual private placement exemptions, with the usual discipline around accreditation and general solicitation. A subtle trap for corporate sponsors deserves emphasis here: Corporate communications teams love announcing funds, and a press release drafted for a strategic audience can raise general solicitation concerns once outside capital is contemplated. Unless the offering is made pursuant to Rule 506(c), which requires verification of accredited investor status for each purchaser not satisfying minimum commitment amounts, the offering discipline of Rule 506(b), which prohibits general solicitation, must extend to the sponsor’s publicity machine – and the coordination must start early.

Antitrust. Two corporate-specific notes: Minority venture investments by a strategic investor may require premerger (HSR) filings above the size-of-transaction threshold because the “solely for investment purposes” exemption is generally understood to be unavailable to an investor taking board seats or investing with strategic intent – a routine trap for corporate programs writing large late-stage checks. Additionally, board seats need screening for interlocking-directorate issues, where portfolio companies compete with the parent or with each other.

Cross-border regimes. Foreign-parented programs investing in US startups must navigate CFIUS – especially for critical technology, data and infrastructure businesses, where staying genuinely passive (no board seat, special information or governance rights) is often the price of avoiding filings. US-parented programs investing abroad now also face the newer US outbound investment restrictions covering certain sensitive technologies in countries of concern. Both regimes cut directly against the strategic access that motivates corporate venturing, and they should shape structure and deal terms from the start. For a fuller tour of these regimes, see our recent primer on deploying private fund capital globally in complex regulatory times.

Financial reporting. A public company sponsoring a corporate venture program faces meaningful financial reporting complexity regardless of whether it invests directly off the balance sheet or through a fund vehicle. Although outside the scope of this post, financial reporting is a critical issue nonetheless. Under generally accepted accounting principles (GAAP), the corporate sponsor must determine whether each portfolio holding is carried at fair value through earnings, accounted for under the equity method or consolidated – a classification that turns on ownership percentage, board representation and the degree of influence the corporation exercises, all of which are in tension with the strategic access rights the program is designed to preserve. Because venture-stage holdings are predominantly in illiquid, non-publicly traded companies, fair value measurements require inputs that are inherently judgment-intensive and subject to audit scrutiny, and any mark-to-market movements flow through the income statement, introducing earnings volatility that can be difficult to explain to public market investors. This area is often a deciding factor between balance sheet and fund structures.

How founders – and their other investors – see corporate money

It may seem odd to end a structuring primer with a section about other people’s perceptions. It is not. Reputation is where the loop closes. Structure drives behavior, behavior drives reputation, reputation drives deal flow – and deal flow, as we have seen, drives everything else.

Founders and institutional co-investors bring a specific, learned set of concerns to corporate money:

Signaling risk. If a strategic investor with deep pockets and inside knowledge declines to follow on – or never acquires the company – the market wonders why. Startups burned when corporate programs pulled back in past downturns have taught a generation of founders to underwrite a corporate investor’s durability, which is precisely why committed funds outrank budget lines.

Speed. Approval chains are visible from the outside. Slow money loses competitive rounds.

Optionality damage. Anything that looks like an option on the company – rights of first refusal on a sale, broad exclusivity, blocking rights over exits – will be priced brutally against the investor or shown the door entirely. The surviving market norm is modest notice rights at most, with commercial arrangements negotiated separately, on their own merits, as commercial arrangements.

Information anxiety. Portfolio companies fear their data flowing to the parent’s product teams, which is why credible, contractual walls between fund and business units are now table stakes at well-run programs.

Governance. Board observer seats, rather than director seats, remain the common corporate accommodation – lighter on conflicts, antitrust and fiduciary complications for everyone.

Well-run corporate programs treat this list not as criticism but as a design specification. Structure a fund that is durable, fast, walled and free of option-like overreach, and the corporate affiliation flips from discount to premium. Founders affirmatively seek the strategic investor who can deliver the customers, distribution, technical depth and credibility that no purely financial investor can match.

Practical market observations

Pulling the threads together, here are a few observations from the field.

Most corporate venture activity is still not a fund, and the decision to form one is best understood as a commitment decision, not a paperwork decision. The vehicle’s deepest function is to make the program durable – against budget cycles, against leadership changes, and in the eyes of the founders and co-investors whose willingness to work with the program is the whole game.

The three spectrums travel together, and internal consistency is the design test. A program with a sourcing-dependent mandate, salary-only compensation and fully in-house operations is internally inconsistent, and that inconsistency will resolve itself the hard way – through the revolving door. A processing-oriented program with inbound flow can sit happily at the captive end of all three spectrums for decades. Neither position is inherently right or wrong; mismatched positions are.

Strategic focus predicts longevity, compensation predicts retention, and committed capital predicts market credibility. The research and the market experience agree on all three, and they are the three levers most within a sponsor’s control.

Movement along the spectrums, in both directions, is normal and should be planned for rather than improvised. The assets that make a spinout possible – separable team, separable brand, attributable track record – accumulate whether or not the corporation intends them to. Decide on purpose.

The far end of the spectrum converges, purposefully, with the independent venture model, layered on top with a negotiated corporate premium: the name, the admission control, the professional home for inbound flow and the closer seat. The recurring board-level question is whether that premium is worth the sponsorship burdens. There is no universal answer; there is only the honest comparison, refreshed every vintage.

Conclusion

Done well, corporate venture capital is one of the great compounding strategic assets available to a large corporation – a durable window on the future of its industry, decades of relationships in its innovation ecosystem and, run with discipline, genuine financial returns. Done half-heartedly, it is a revolving door at a logo graveyard. The venture market has a long memory for the difference.

The structuring questions, in the end, are few. They are the ones we opened with. Whose money is in the fund – and who may ever be admitted? Should the program operate from the balance sheet or through a committed vehicle? Does deal flow come to us, or must we go and earn it? How will we pay the people who earn it – and can our culture survive the cost? How far from the mother ship should the team sit? While related, these questions are not the same. The programs that endure are the ones whose answers fit each other, fit the deal-flow reality and get revisited – deliberately, at the top of the house – every vintage.