Hedge Fund Non-Compete and Garden Leave: The Mechanics of Team Lift-Outs

Most institutional hedge fund businesses begin as a lift-out. A portfolio manager and two or three colleagues leave an established platform, and the question that determines whether the new fund launches on schedule is rarely an investment question. It is what the departing team is contractually permitted to take, and when it is permitted to start. Hedge fund non-compete and garden leave provisions sit at the centre of that question, alongside ownership of models, code and performance history. This article treats those provisions as a commercial planning problem rather than a legal argument, because the teams that launch cleanly are the ones that mapped the constraints before anybody resigned.

"The launches that go wrong are almost never the ones where a covenant was breached. They are the ones where nobody built a timetable around the covenants that everybody already knew about. We ask managers to establish their notice date, their restricted period and their track record position at the first meeting, because those three things set the entire launch plan."David Lloyd, Chief Executive Officer at CV5 Capital

Executive Summary

A lift-out is a transfer of human capital, and the value transfers only if the surrounding rights transfer with it. Investment judgment and process discipline are portable. The performance record, the code that generates signals, the research library and the investor relationships frequently are not, or are portable only in part. Restrictive covenants govern the timing of the move. Ownership and confidentiality provisions govern what arrives when it happens. Managers who separate those two questions plan far better than those who treat the whole subject as one undifferentiated legal risk.

• Garden leave and non-competes control when a team can trade; ownership and confidentiality terms control what it can bring.
• Track record attribution is a documentation exercise before it is a marketing exercise, and the former employer usually holds the pen.
• Models, code, research and data are typically assigned to the employer under standard intellectual property clauses, whatever any individual contributed.
• Enforceability of post-termination restraints varies widely between financial centres, and identical drafting can produce different outcomes.
• Investor and colleague non-solicitation provisions often run longer than the non-compete and bite harder on a first capital raise.
• The practical remedy is sequencing: build the launch timetable around the restricted period rather than against it.

What a Departing Team Can and Cannot Take

A lift-out moves people. It does not automatically move the things that made those people valuable. It helps to separate the departing team's assets into four categories, because each is governed by a different mechanism and each behaves differently under pressure.

The first category is skill and knowledge. Investment judgment, market understanding and process discipline travel with the individual and cannot realistically be restrained. The second is documented performance history, which is usually owned by the fund or the firm even though the departing team generated it. The third is tangible intellectual property: models, code, backtests, research notes and curated data. The fourth is relationships, meaning investors, clients, counterparties and colleagues.

Only the first category is unambiguously portable. The other three are governed by contract, and that contract was signed years before anybody contemplated a departure. This is why the most useful step a prospective manager can take is to retrieve and read their own employment agreement before speaking to a single investor. Teams that institutionalise a strategy on assumption rather than documentation tend to discover the constraint at the worst possible moment, a pattern we examine in our guidance on moving from a proprietary trading desk to a Cayman fund.

The distinction is commercial as much as legal. An allocator evaluating a new manager underwrites a process, a record and a team. If the record cannot be substantiated and the models must be rebuilt, the manager is raising capital on narrative rather than evidence. The launch remains possible. It is simply slower, smaller and more expensive.

Track Record Attribution: The Most Overestimated Asset

Performance history is the asset departing teams consistently overvalue in their own planning. In most institutional settings the record belongs to the fund or the management company, not to the individual who produced it. What the individual holds is a claim to attribution: the right to describe their role in producing a result that somebody else owns and reports.

Attribution is only useful if it can be substantiated. A manager quoting a prior return figure without support will be discounted heavily in diligence, and in some cases the claim is treated as a marketing risk rather than a credential. The substantiated version usually requires the former employer to confirm the role, the period, the mandate and the numbers. That confirmation is discretionary. It is far easier to obtain from an employer left on good terms than from one left in dispute.

Where confirmation is unavailable, the alternative is to treat prior performance as context and to build a verifiable record from day one. That means independent administration, an auditor and disciplined monthly investor reporting from the first month of trading. A short live record produced under institutional controls persuades a serious allocator more than a long unverifiable one. The mechanics are set out in our note on building and presenting an institutional track record.

Managers should also be precise about how prior performance appears in marketing materials and in the offering document. Descriptions must be accurate about who managed the assets, under what mandate, with what leverage and at what fee load. Overstatement here is one of the more avoidable ways a promising launch damages its own credibility.

Models, Code and Research Data

Employment agreements in asset management almost always assign to the employer any intellectual property created in the course of employment. This is standard drafting, and it reaches further than most professionals assume. It commonly captures code written outside working hours, models developed on personal hardware and research conducted on subjects related to the employer's business.

The practical consequence is that a quantitative team generally cannot bring its signal library, execution logic, backtest infrastructure or curated datasets. Nor can it bring the derived artefacts. Parameter sets, feature definitions and calibration outputs are usually treated as the same property as the code that produced them.

Rebuilding rather than transporting

The workable route is reconstruction from public knowledge and independent effort. The discipline that makes reconstruction defensible is documentary. Build in a clean environment on new infrastructure, keep a record of what was used and when, and retain nothing belonging to the previous employer. That includes personal notebooks, exported spreadsheets and archived correspondence.

Reconstruction carries a commercial benefit that managers rarely anticipate. A model rebuilt with several additional years of perspective is frequently better specified than the original. Because the rebuild usually overlaps with a period of paid leave, it is not dead time. It is often the most productive uninterrupted research window a manager will ever have.

Hedge Fund Non-Compete and Garden Leave Provisions Compared

Restrictive covenants are usually discussed as a single concept. They are better understood as a set of distinct instruments with different scopes, different durations and very different effects on a launch timetable.

Why garden leave is usually the binding constraint

Garden leave operates during employment rather than after it. The individual remains employed and paid, remains subject to duties of loyalty and confidentiality, and is simply removed from the desk and from information. Because the employment relationship subsists, the restraint tends to be more robust than a post-termination covenant. In practice it is the provision that sets the earliest possible launch date.

Non-competes attract more attention and are far more frequently contested. A non-compete restrains activity after employment has ended, and most financial centres test whether it protects a legitimate business interest and whether it reaches further than necessary to do so. Duration, geographic scope and the definition of a competing business are all live points.

Non-solicitation provisions are the quiet ones. They are generally easier to uphold than non-competes, they frequently run for longer, and they reach the two things a new manager needs most: investors and colleagues. A team non-poach clause can prevent a founder from hiring the very people the strategy depends on, which creates exactly the exposure described in our analysis of key person risk facing emerging managers.

How Enforceability Differs Across Financial Centres

Restrictive covenant enforceability is among the least uniform areas of commercial practice. The same words in the same contract can produce materially different outcomes depending on where the employment sits, which entity is the employer and where enforcement is pursued. The generalisations below are directional only. Outcomes turn on specific facts, and managers should obtain independent professional advice on their own arrangements.

Two structural points are worth drawing out. The first is that garden leave travels better than a non-compete. Where a jurisdiction is sceptical of post-termination restraints, it will often still respect a notice period during which the individual remains employed and paid. Managers who assume a sceptical jurisdiction means an immediate exit are frequently wrong.

The second is that the employing entity matters as much as the individual's location. Large managers commonly employ staff through group entities in particular jurisdictions, and the choice of employer entity and governing law can decide which regime applies. A team spread across several offices may find its members face different constraints, which complicates the sequencing of a group departure.

None of this makes covenants safe to disregard. Even where enforceability is doubtful, litigation risk is real, and a new manager carrying an unresolved dispute is difficult to allocate to. Institutional allocators consistently prefer a clean, uneventful departure to a defensible one.

Sequencing the Lift-Out

The difference between a smooth launch and a stalled one is almost always sequencing. Covenants are known quantities well in advance. The failure mode is not breaching them. It is failing to build the timetable around them.

The correct order of operations runs backwards from the earliest date the team can lawfully begin managing capital for the new business. Fund formation, regulatory registration, service provider onboarding and account opening can generally proceed in parallel with a restricted period, because forming a vehicle is not the same as competing. Marketing and capital raising cannot proceed on the same basis, because approaching covered investors is precisely what a non-solicitation provision addresses.

Before notice is given, a departing manager should establish the following:

• The exact end date of the notice period and whether the employer may elect to impose garden leave.
• Whether any non-compete runs from termination or is reduced by garden leave already served.
• The precise contractual definition of a competing business, which is often narrower than the team assumes.
• Which investors and clients are covered by the non-solicitation, and whether it restrains approach only or also acceptance of unsolicited interest.
• Whether former colleagues can be recruited, from what date, and by whom.
• How the employer has actually behaved on previous departures, which is frequently more informative than the drafting.

Formation work should begin early. A Cayman structure, its regulatory registration, its board and its service provider arrangements take time that overlaps neatly with a restricted period. Using that window is how a team moves from resignation to first trade without an idle quarter. Our complete guide to launching and operating a Cayman hedge fund sets out those parallel workstreams in order.

The departure itself should be conducted so the former employer has no reason to escalate. Return devices and data. Copy nothing. Do not approach covered investors during the restricted period, including indirectly through intermediaries. The cost of a dispute is not principally financial. It is delay, and delay is what kills launches.

Documentation the New Manager Needs at Launch

A team that has just navigated a set of restrictive covenants usually arrives at its own launch with a sharpened view of what employment documents should say. That instinct is correct and should be applied immediately, because the new management company will eventually face the same question from the other side of the table.

Founding documentation should address ownership and departure before there is anything to argue about. Equity in the management company, vesting and the treatment of a departing founder are the obvious items. Less obvious, and far more frequently omitted, are the provisions that decide whether the business survives a second lift-out three years later.

At a minimum, the launch documentation should cover:

• Assignment to the management company of all intellectual property created by founders and employees.
• Clear ownership of the performance record by the fund and the manager rather than by any individual.
• Proportionate notice periods and an express garden leave right for investment personnel.
• Non-solicitation of investors and employees, drawn narrowly enough to be credible.
• Confidentiality covering models, code, investor identities and commercial terms.
• Founder equity vesting with good leaver and bad leaver treatment aligned to the fund's key person provisions.

These terms interact directly with the fund's own documents. A key person clause in the offering memorandum that names a founder means little if that founder's employment terms permit an abrupt exit. Seed investors test the alignment closely, as we discuss in our review of what emerging managers give up in a seed deal. Structuring the management company itself is the related exercise, covered in our overview of fund manager formation.

What Allocators and Diligence Teams Actually Test

Talent mobility is rarely presented to allocators as a standalone diligence topic. It nonetheless runs through an operational review, and it generally surfaces in three forms.

• Provenance of the record: who managed the assets, under what mandate, and whether the figures can be independently confirmed.
• Legal overhang: whether any former employer has raised or reserved a claim, and whether the manager is free to trade the strategy being marketed.
• Key person concentration: whether the strategy depends on individuals who are contractually free to leave and take colleagues with them.

The first two resolve quickly where the manager holds documentation. The third is structural, and it is where the choice of launch route changes the answer. Where a manager operates inside an established regulated platform, the fund, its board, its service providers and its regulatory permissions sit outside the management company. The departure of an individual is disruptive but not existential, because the vehicle and its governance persist.

That separation is one reason allocators have grown comfortable with platform structures for emerging managers. It converts key person risk from a single point of failure into a governance question with a documented answer. The CV5 Capital hedge fund platform is built on that principle, with the segregated portfolio, the board and the operating infrastructure independent of any one manager.

Managers should expect these questions early, sometimes in a first meeting. Answering them precisely, with dates and documents rather than reassurance, is a credibility signal well beyond its apparent significance. It tells an allocator that this manager plans ahead, documents decisions and closes loops.

Key Takeaways

• Skill and process travel with a departing team, while the performance record, models, code and investor relationships generally do not.
• Track record attribution should be requested from the former employer before notice is given, when goodwill is at its highest.
• Intellectual property assignment clauses reach further than most professionals assume, so models and code usually have to be rebuilt rather than transported.
• Garden leave is often the binding constraint on a launch date because it operates during employment and is respected more consistently than a post-termination non-compete.
• Non-solicitation provisions covering investors and colleagues frequently run longer than the non-compete and shape the first capital raise more directly.
• Fund formation, registration and service provider onboarding can proceed during a restricted period, converting notice into preparation time rather than idle time.

Frequently Asked Questions

Can a hedge fund manager take their track record to a new firm?

Generally the record itself belongs to the fund or the former employer, not to the individual. What a departing manager can usually claim is attribution: a description of their role, mandate and period of responsibility. Allocators will expect that attribution to be confirmed independently, which normally requires the former employer's cooperation, so it is best requested before notice is given.

What is garden leave and how long does it usually last?

Garden leave is a period during which an employee who has resigned remains employed and paid but is removed from the desk, from clients and from confidential information. In asset management it commonly runs from one to twelve months, depending on seniority and the contractual notice period. Because the employment relationship continues, it tends to be the most reliably respected restraint on an early launch.

Are hedge fund non-competes enforceable?

It depends heavily on the jurisdiction, the drafting and the facts. Most financial centres will consider whether the restraint protects a legitimate business interest and whether its duration, geography and scope go further than necessary. Some jurisdictions require the individual to be compensated during the restricted period, and others are markedly less receptive to post-termination restraints altogether. Independent professional advice on the specific contract is essential.

Can a new fund be formed during a non-compete or garden leave period?

Forming a vehicle is not the same activity as competing, and in many cases structuring, registration and service provider onboarding can proceed while a restricted period runs. Marketing and soliciting covered investors is a different matter and is usually the activity the covenants target most directly. The practical approach is to complete formation during the restricted period and to begin the raise once the restraint expires.

Can a departing manager rebuild the same trading models at a new firm?

Rebuilding from public knowledge and independent effort is generally the accepted route, whereas transporting code, datasets, parameter sets or backtests is not. The discipline that protects the new manager is documentary: a clean build environment, new infrastructure, and a contemporaneous record of sources and dates. Retaining nothing from the former employer, including personal archives, is the baseline expectation.

Can a founder hire former colleagues after leaving?

Team non-solicitation clauses commonly restrain recruiting former colleagues for a defined period, and they are often easier to uphold than a non-compete. The restriction can also outlast the non-compete, which affects hiring plans well into the first year of trading. Founders should confirm the covered population and the expiry date before making any commitment to a prospective hire.