The Non-Compete Calendar: How Far Ahead a Fund Must Hire a Senior Quant in 2026

A senior portfolio manager who resigns from a large multi-strategy platform in October 2026 will, on the paper now typical in New York and Connecticut, trade for his next employer in the spring or summer of 2028; on Florida paper signed since July 2025 the date can sit in 2030. How far ahead a fund must hire is therefore no longer a judgment about urgency. It is a date on someone else's contract, read through the law of the state or country the seat sits in.

Forward hiring is the practice of opening a senior search against the date a candidate's restrictions expire rather than the date the seat opens. What makes it unavoidable in 2026 is that the restriction dates moved in at least nine jurisdictions within eighteen months, in both directions, so the calendar a hiring desk ran in 2024 is wrong in most of them.

How long is the sit-out now, and who sets it?

Three numbers rather than one. The notice period, during which the employee is paid and usually kept away from the desk; the post-employment non-compete, which may or may not be paid; and the vesting schedule of deferred compensation, which decides what a rival has to buy out. The first two add. The third sets the price of shortening them.

The upper bound moved again in August 2026. Bloomberg reported, and Hedgeweek carried on 14 August 2026, that Citadel was requiring non-competes of up to two years for its investing staff, including some analysts, with the length of the garden leave linked to total compensation and a twelve-month minimum for analysts; at other large multi-strategy funds, analyst non-competes generally run nine to twelve months. That followed the firm's move to 21-month non-competes for senior portfolio managers, reported in January 2025. Put a twelve-month notice in front of a twenty-four-month restriction and a senior candidate who resigns today is a hire for 2028 or 2029.

The legal ceiling now sits above the market practice. Florida's CHOICE Act, in force since 1 July 2025, lets a covered employer enforce a non-compete or a garden-leave notice period of up to four years against an employee earning more than twice the mean wage of the employer's county, and requires a court, on application, to enjoin both the employee and the firm hiring him; the employee then has to prove his way out by clear and convincing evidence. As of August 2026 no reported Florida decision construing the statute had been identified. Citadel lobbied for the bill in the spring of 2025. Four years is the ceiling for a Miami seat, and the first reported decision will tell the market how much of it is real.

London runs shorter, and the government has the numbers. The Department for Business and Trade's working paper of 26 November 2025 put the typical UK non-compete at around six months, found that 71 percent run longer than three months and some to 24, and counted around five million employees in Great Britain working under one. In a London hedge fund contract the restraint is mostly carried by notice and garden leave, with the non-compete behind it enforced only as far as a court finds reasonable, which is why a senior London seat can often be filled a year sooner than the same seat in Manhattan.

The third date is the one the capital side has been moving. Senior platform deferrals take three to four years to vest, and in the summer of 2026 the flagship funds termed out their investor liquidity to match, a convergence set out in the locked-capital piece. The vesting cliff prices the buyout, and the buyout is what a candidate's current employer is protecting when it writes the first two dates.

What changed in the law in 2025 and 2026, seat by seat?

The map as it stood on 1 October 2026, each entry with its source, so a reader can check the line that matters to his seat.

• United States, federal. The FTC's 2024 rule banning most non-competes never took effect. After a Texas court set it aside in August 2024 and the Commission dropped its appeal in September 2025, the FTC struck the rule out of its regulations on 12 February 2026. Enforcement is state law, contract by contract.
• New York. Senate bill S9759 would ban non-competes except for people earning above $500,000 and cap those at one year with full salary paid throughout. The Senate passed it by 40 votes to 21 on 3 June 2026, its second vote for the measure in two sessions, and delivered it to the Assembly, where the companion bill A10023 sat in the Labor Committee when the session ended. Nothing in a New York contract has changed. A paid one-year cap for every senior portfolio manager in Manhattan is an Assembly vote and the Governor's signature away, and the Governor vetoed a broader ban in December 2023.
• Connecticut. Raised Bill 5492, which would void non-competes below twice the minimum wage and confine the rest to the duties and geography of the employee's last two years, received a favourable committee report in March 2026, was tabled for the House calendar on 2 April and did not reach the floor before the session adjourned on 6 May 2026. The Greenwich and Stamford seats stay on common-law reasonableness, which in practice means the paper governs.
• Florida. The CHOICE Act, above: four years, a mandatory preliminary injunction against the employee and the hiring firm, a day-for-day credit for any paid garden leave already served, and no reported decision in its first year.
• Delaware. A large share of hedge fund employment and partnership agreements choose Delaware law. The Court of Chancery has stopped rescuing overbroad covenants: in HKA Global v. Beirise on 16 December 2025 and Boingo Wireless v. Georges on 19 February 2026 it declined to narrow non-solicitation covenants it found too wide, dismissing the Boingo claims at the pleading stage, and it did so on the authority of the Delaware Supreme Court's ruling in Sunder Energy v. Jackson that a court may refuse to blue-pencil. A covenant drafted wider than the business interest it protects can now fail entirely under Delaware law, whatever its stated length.
• United Kingdom. The November 2025 working paper consulted on a three-month statutory cap, a cap that varies with employer size, a salary threshold below which non-competes are unenforceable, and an outright ban. Responses closed on 18 February 2026 and the government has not announced a legislative timetable. Until it does, the six-month norm and the reasonableness test stand.
• Netherlands. The cabinet sent the bill modernising the non-compete clause to the Council of State on 26 June 2026: a twelve-month maximum, mandatory compensation whenever the employer invokes the clause, and a geographic scope that has to be specified in writing. For Amsterdam's proprietary trading firms, the two-year restraint will not be available once the bill passes.
• Hong Kong. On 13 February 2026 the Court of First Instance refused an interim injunction to enforce a twelve-month non-compete against a portfolio manager at a virtual-asset manager who had joined a rival, on the ground that the employer had not shown good prospects of establishing that a restraint of that breadth protected a legitimate interest and went no wider than necessary. Twelve months is not a safe assumption in Hong Kong for anyone, which shortens the calendar for a fund hiring there and raises the risk for one relying on its own paper.
• Singapore. The Ministry of Manpower told Parliament on 5 February 2024 that restraint clauses must balance the employer's interest against the employee's ability to earn a living, that the courts set the test, and that tripartite guidelines were due in the second half of 2024. Asked again, it told Parliament on 8 January 2025 that it was still discussing with the unions and the employers' federation how and when such clauses should be used, with details to follow in due course. No guideline had been published on the ministry's site as of 1 October 2026; reasonableness remains the only rule.
• United Arab Emirates. Onshore, Article 10 of the 2021 labour law caps a non-compete at two years and the courts award damages rather than keeping an employee out of a rival; the DIFC and ADGM apply their own regimes. A Dubai seat's lead time depends on which of the three the contract chose, and the difference can be a year.

What does the research say a non-compete buys the firm holding it?

A buyout, mostly. Liyan Shi's model in Econometrica in March 2023 treats the clause as what it is in practice: a device for enforcing a payment when the worker leaves, through which the incumbent extracts rent from the next employer. She finds the rent socially excessive and the optimal policy close to a ban; in her contract data about 64 percent of executives at listed firms had signed one. For a hiring desk the implication is plain. The bridge over the sit-out and the buyout of unvested deferrals are the purpose of the clause, and a budget that treats them as a sweetener has misread the document.

The clause also thins the field. Starr, Prescott and Bishara found in the Journal of Law and Economics in 2021 that about 18 percent of the US labour force is bound by a non-compete and 38 percent has signed at least one, with wages lower where the clauses are easier to enforce. Johnson, Lavetti and Lipsitz, in the Journal of Political Economy in September 2025, showed that stricter enforceability reduces earnings and job mobility, most for the workers likeliest to sign, and that the effects cross state lines: a New York platform hiring off Florida paper imports Florida's enforceability with the hire. Balasubramanian and co-authors, in the Journal of Human Resources, found that when Hawaii stopped enforcing the clause for technology workers in 2015 their job moves rose by 11 percent and starting pay by 4 percent, and that eight years into a job in an average-enforceability state a worker had held about 8 percent fewer jobs and earned 4.6 percent less. Lipsitz and Starr found the same direction in Management Science from Oregon's 2008 ban for hourly workers, which raised wages 2 to 3 percent and job-to-job mobility. The stricter the jurisdiction, the fewer people move, and the longer each move takes.

Matt Marx and Lee Fleming's review of the evidence for the NBER in 2011 adds the two findings a buyer most needs. A quarter of the technical professionals who signed a non-compete and then changed jobs also changed industries, a career detour that came with lower pay; and in nearly 70 percent of cases the worker was asked to sign after accepting the offer, having already turned down the alternatives, with nearly half reporting time pressure or being told the clause was non-negotiable. Read against a quant desk: the candidate who spent eighteen months at a non-competing firm may be a detour rather than a decline, and the hiring firm's own offer letter is where the next non-compete gets signed, which sets the lead time for the hire after this one.

How does a hiring desk build the calendar?

Start with three dates per candidate, asked at the first conversation and before any discussion of the book: the day notice ends, the day the post-employment restriction ends, and the day the next tranche of deferred compensation vests. Then apply the jurisdiction. On New York or Connecticut paper the first two dates govern as written. On Delaware-law paper the restriction is worth reading for breadth before it is priced, because a clause drawn wider than the firm's real interest may fail in full. On Florida paper the restriction may be longer than any court will uphold, and in 2026 nobody knows. In London the enforceable restraint is close to the notice period, and in Hong Kong a twelve-month clause failed to win interim enforcement against a portfolio manager in February 2026. None of this is legal advice: before a start date goes to an investment committee, the candidate's paper is read by counsel in the jurisdiction of the seat.

The second step is the buyout, which is now the largest line in a senior offer and which the vesting date prices to the month. The third is the start date the calendar produces, the only honest number to put in front of an investment committee. The platforms that grew capital at roughly 25 percent in the year to September 2026 while growing headcount 10 to 11 percent, the ratio reported in September 2026, are hiring into exactly this calendar, and much of the distance between those two numbers is paper.

A search opened against the restriction date is a different process from one opened against a vacancy. The shortlist is built from people who cannot start for a year or more, the first conversation asks for the three dates before it asks about the Sharpe ratio, and the offer carries a start date the candidate's lawyer has already checked. The candidate's view of the same window, including why his negotiating power peaks late in it, is set out from the other side in the sit-out piece; a buyer who has read it knows the late-window second offer is increasingly priced into the seller's contract as liquidated damages. Funds that run senior hiring this way are earlier rather than faster, and in a market where the paper runs two years, earlier is the whole advantage.

What moves the calendar into 2027?

Five dates, in rough order. New York's Assembly returns in January 2027 with A10023 in committee and a Senate that has passed the bill twice; if the Assembly passes it and the Governor signs it, every Manhattan senior contract acquires a paid one-year cap and the New York row of the calendar halves. The UK government owes a response to its November 2025 consultation, and every option it listed shortens the London row. The Dutch bill returns from the Council of State for a parliamentary vote, after which the Amsterdam row caps at twelve months with pay. Singapore's guidelines, promised for 2024 and still in tripartite discussion in January 2025, would turn a case-by-case test into a published norm. And Florida's first reported CHOICE Act decision will say whether a four-year restraint on a well-paid quant survives contact with a judge.

The direction is clear before any of them lands. The calendar is lengthening where the largest platforms book their senior employment, in Florida and on their own paper, and shortening in London, Amsterdam, Hong Kong and, if Albany follows its Senate, New York. A hiring desk that reads the map accordingly opens the London search a year after the Miami one for the same start date, prices the Delaware candidate's covenant for breadth before duration, and treats every New York contract signed in 2026 as one that may be capped at twelve months before it expires. The firms that seat the senior quants of 2028 will be the ones that began reading other firms' paper in 2026.