Leave Carry Forward and Year-End Leave Close

Most HR teams discover their leave carry forward rules the same way — in a heated email thread in the first week of the new leave year, when an employee opens the portal, sees a balance that is smaller than the one they remember, and asks who authorised the deletion. By then the damage is done. The balances have already rolled over, the payroll cut-off has passed, the finance team has closed the books on a provision that may or may not have been right, and HR is reconstructing a decision from memory because nobody wrote down what the carry-forward cap was supposed to be.

The leave year close is one of the few HR events that touches operations, payroll, finance and employee trust at the same time, and it is almost always under-planned. This guide covers how to run it properly: how to decide whether your leave year should follow the calendar or the financial year, which rules a leave policy has to define before the leave year end arrives, how to build an eight-to-ten week close runbook with named owners, how to give finance a leave liability number they can provision against, how to survive the December or March leave rush without stalling delivery, and how to clean up leave balance data so the opening numbers on day one of the new year are defensible.

Why the leave year end is an operational event, not an admin formality

The reason leave year close goes wrong is that it looks small. Somebody flips a setting, balances reset, done. In practice, four separate problems converge in the same six weeks, and each one costs real money or real goodwill.

Unbudgeted liability arrives all at once

Accrued leave that can be encashed or that must be paid out on exit is a liability. It sits quietly on the balance sheet, grows a little every month, and then becomes visible exactly when finance is trying to close the year. If your carry-forward cap is generous and your encashment window is open, a cohort of long-tenured employees can convert a large amount of accumulated leave into cash in a single payroll cycle.

The problem is rarely that the cost is unaffordable. The problem is that it was not forecast. A payroll run that is materially heavier than the previous eleven runs, landing in the same month as appraisal arrears and variable pay, creates a cash-flow surprise for a founder who was planning runway to the rupee. Finance does not object to paying leave encashment. Finance objects to finding out about it on the twenty-sixth of the month.

The December and March leave rush stalls delivery

If leave lapses at the year end, people will burn it before it lapses. That is rational behaviour and you should expect it. What you should not do is let it happen unmanaged.

The predictable failure looks like this. Through the year, the people who take the least leave are the ones carrying the most delivery load — the senior engineer holding a migration, the finance controller mid-audit, the one person who understands the billing logic. Those are precisely the people with the largest unused leave balance at year end. When the deadline approaches, they all apply in the same three weeks, managers approve because refusing feels unfair, and the last month of the leave year becomes the month with the lowest effective capacity in the calendar. Customer commitments made in a planning meeting six months earlier now collide with a staffing reality nobody modelled.

It gets worse when the leave year ends in March, because in India that month is already loaded: financial year close, statutory filings, investor reporting, appraisal cycles, and for services businesses, client year-end deadlines. Adding a leave rush to that month is a self-inflicted capacity problem.

"My leave vanished" disputes

Leave disputes are almost never about the number of days. They are about the feeling of having something taken away without notice. An employee who is told in October that ten days will lapse on 31 December, and is reminded again in November and December, is annoyed but not aggrieved. An employee who sees the same ten days disappear on 1 January with no prior communication believes the company deliberately withheld information to save money.

That belief is corrosive out of proportion to the amount involved. It spreads in team channels, it comes up in exit interviews, and it makes every subsequent policy communication less credible. The fix is cheap — visibility and reminders — and the cost of skipping it is not.

The audit trail problem when balances live in a spreadsheet

The deepest problem is evidentiary. When leave balances are maintained in a spreadsheet, an adjustment is just a changed cell. There is no record of who changed it, when, from what to what, or why. Six months later, when an employee disputes a balance or an auditor asks how the closing figure was derived, you cannot reconstruct the chain.

This is not hypothetical risk. It shows up in three places:

• Full and final settlements. An exiting employee claims a higher encashable balance than the one you computed. Without an accrual and consumption log, the negotiation is a memory contest.
• Statutory inspections and audits. An inspector or auditor asks for evidence that earned leave was granted, tracked and carried forward in line with the applicable state rules. "Here is our spreadsheet" is a weak answer if the spreadsheet has no version history.
• Due diligence. In a funding round or acquisition, employee-benefit liabilities are examined. An unreconciled leave register turns into a diligence finding, and diligence findings turn into escrow.

A system that logs every accrual, application, approval, adjustment and lapse with a timestamp and an actor is not a luxury at the leave year end. It is the difference between a close you can defend and one you can only describe.

Leave year vs financial year vs calendar year: choosing your cycle

Before you can close a leave year cleanly, you have to be clear about what your leave year is. Many Indian companies have never made this an explicit decision — they inherited January-December because the first HR person set up the tracker that way, or April-March because payroll was configured around the financial year and nobody separated the two.

The three cycles that matter:

• Calendar year (January-December): the leave year matches the year employees think in.
• Financial year (April-March): the leave year matches the books, payroll year and tax year.
• Anniversary-based (each employee's own joining date): each person's leave year starts on their date of joining.

Anniversary-based cycles are common in some multinational setups and they solve the rush problem elegantly, because no two employees hit their year end on the same day. They also create a permanent reconciliation headache: three hundred different close dates, three hundred different pro-rating windows, and a liability number that has to be computed continuously rather than at a point in time. For most Indian SMBs, the administrative cost outweighs the benefit. The real choice is between calendar and financial year.

There is no universally right answer, but there is a decent default. If your biggest operational risk is capacity, pick the cycle whose year end falls in your quietest month. If your biggest risk is financial control and provisioning discipline, align to the financial year and accept the March congestion, but then build an early encashment window and a hard blackout so the rush does not land in the last three weeks.

One practical compromise many companies land on: run the leave year April-March to match the books, but shift the practical deadlines forward. Close the encashment window in January, set the carry-forward cut in mid-February, and treat March as a blackout month for discretionary long leave. You keep financial alignment without putting a leave rush on top of your statutory peak.

How to change your leave year mid-stream

If you decide to switch — and switching is usually a one-time cost worth paying if the current cycle is actively hurting you — the transition needs a stub period, clear pro-rating and heavy communication.

1. Define the stub period explicitly. Moving from April-March to January-December means you have a nine-month stub from April to December, after which the new twelve-month cycle starts. Moving the other way gives you a three-month stub. Name it in writing: "transition leave period, 1 April to 31 December".

2. Pro-rate the annual entitlement for the stub, and round in the employee's favour. If the annual entitlement is normally accrued monthly, a nine-month stub simply accrues for nine months and no special maths is needed. If your entitlement is credited upfront at the start of the leave year, you need a pro-rated upfront credit for the stub. Whatever rounding rule you use, round up. The goodwill is worth more than the fractional days.

3. Carry forward the full balance into the stub without applying a cap. This is the single most important transition rule. Employees should not lose leave because the company changed its own calendar. Apply the normal carry-forward cap only at the end of the first full new-cycle year, and say so.

4. Decide what happens to the lapse rule during transition. The cleanest position: no lapse at the stub boundary. Lapse resumes at the end of the first complete cycle under the new calendar. Anything else invites the accusation that the change was a cost-saving exercise dressed up as an administrative improvement.

5. Re-base encashment eligibility. If encashment eligibility is tied to a minimum residual balance or a minimum tenure, state how the stub period counts. Generally, count it in full.

6. Communicate three times, in three formats. An announcement email with the exact dates and the pro-rating formula; a one-page FAQ; and a live session where people can ask the awkward questions. Then publish every employee's transition-period opening balance individually, in the system, where they can see it.

7. Freeze the policy for at least one full cycle afterwards. Changing the leave year and the carry-forward cap in the same breath is how you turn an administrative change into a trust problem. Do one, let it settle, then do the other.

The rules a leave policy actually has to define

Most leave policies are two pages long and answer about a third of the questions that arise at leave year end. The gaps are not obscure — they are the same ten or twelve decisions, unmade, every time. Here is the full set, with the options and what each option costs you.

The policy decision table

Which leave types should and should not carry forward

The general principle: carry forward the leave types that represent earned time, and reset the ones that represent an allowance for unpredictable events.

Whatever you decide, the policy must say it in the same words the system uses. A policy that says "annual leave" while the portal says "Earned Leave (EL)" and the payroll register says "PL" guarantees confusion at year end.

Statutory floor versus company policy

This is where a lot of well-intentioned leave policies quietly go wrong.

In India, leave entitlements are governed by a layered framework. Central labour legislation sets baseline requirements for certain categories of establishment, and each state's shops and establishments legislation — along with rules made under it — sets requirements for commercial establishments operating in that state. Between them, these rules address things like the minimum rate at which earned leave accrues, the circumstances in which accrued leave must be allowed to accumulate and carry forward, and how leave is treated on separation.

Three things follow from this, and they matter more than any specific number.

First, the statutory position is a floor, not a menu. A company policy can be more generous than the applicable statutory requirement. It cannot be less generous. If your policy says accrued earned leave lapses entirely at year end, and the applicable rules require a minimum level of accumulation, your policy is unenforceable to that extent — regardless of what the employee signed. Employment agreements do not override protective legislation.

Second, the requirements vary by state and by establishment type. A company with a Bengaluru head office, a Pune delivery centre and a Gurugram sales office is potentially operating under three different state frameworks, and a factory or manufacturing unit may sit under a different regime again from an IT or commercial office. Some states have issued exemptions or modified requirements for particular sectors, sometimes for limited periods, sometimes with conditions attached. "We follow Karnataka rules everywhere because that's where we're registered" is a common shortcut and a wrong one.

Third, this area changes. Labour law in India has been in a period of consolidation and revision, with implementation timelines and state-level adoption varying. What was correct when your policy was drafted may not be correct now.

The practical implications for your leave year close:

• Get your leave policy reviewed against the current rules for every state where you have employees, by someone qualified to do it, before you finalise carry-forward and lapse rules. Not after.
• Do not copy a leave policy from another company, a template site, or a previous employer. It encodes someone else's state footprint and someone else's compliance date.
• Where a statutory minimum applies, set your policy comfortably above it rather than exactly at it, so that a small rule change does not instantly put you out of compliance.
• Keep the statutory reasoning documented separately from the policy document. When someone asks in two years why the cap is what it is, you want the answer on file.
• If you operate across states with different requirements, decide deliberately between a single harmonised policy set at the most generous applicable level, or state-specific annexures. Harmonising is simpler to administer and costs more; annexures are cheaper and create the risk of an employee transfer silently breaking compliance.

This article deliberately does not quote day counts, accumulation limits, ratios or section references, because those vary and they change. Verify the current position for your states before you press the button that lapses anybody's leave.

The leave year close runbook

The close should start eight to ten weeks before the leave year ends, and every step should have a named owner. Below is a runbook you can lift and adapt. Weeks are counted backwards from the last day of the leave year.

The steps that need more explanation

Freeze and reconcile. The freeze is not optional. If applications for the closing leave year can still be submitted after balances are computed, every downstream number — encashment, carry-forward, liability — is provisional. Set a freeze date, publish it, and enforce it. Genuine exceptions after the freeze should go through a single approval path with a written reason, not through whoever answers first.

Chasing pending applications and un-regularised absences. These are two different problems. A pending application is a request sitting in a manager's queue: the leave may or may not have been taken, and the balance is wrong either way. An un-regularised absence is a day where attendance shows nothing and no leave was applied: someone was away and it was never recorded. Both distort the closing balance, and the second one is the more dangerous, because it usually means the employee took leave that was never deducted and will now be encashed or carried forward.

Run both reports weekly from W-5. Give managers a clear deadline and escalate to function heads after it.

Manager balance sign-off. This step is where most of the value sits, and most companies skip it. Send each manager a simple sheet: employee, opening balance, accrued, consumed, pending, closing balance, days at risk. Ask them to confirm or flag. You are not asking them to audit the arithmetic — you are asking them to spot the cases they know about that the system does not: the person who took a week off informally during a family emergency, the person whose approved leave was cancelled by phone, the contractor who converted to employee mid-year.

Sign-off also does something political: it makes the closing balances a shared decision rather than an HR imposition. When a dispute comes up later, "your manager confirmed this balance in writing three weeks before close" is a much stronger position than "the system says so".

The encashment window as a planning decision. Treat the encashment window as a capacity and cash management tool rather than a benefit administration task. Open it early enough that finance can plan the payout, close it early enough that you are not still computing amounts when payroll locks, and set the eligibility rules so that the encashment reduces the same excess balances that would otherwise drive the year-end rush.

The single design choice that matters most: make encashment available only on the portion of the balance that exceeds the carry-forward cap. Employees are then choosing between cash and lapse, not between cash and carrying the days forward. That converts encashment from a cost into a release valve, and it makes your liability model much simpler because you know exactly which slice of the balance is in play.

Blackout and capacity planning. Decide, function by function, whether the final weeks of the leave year need restricted leave. If they do, publish the blackout at W-8, not at W-2. A blackout announced late is indistinguishable from a trick to force lapse, and it will be read that way.

A fair blackout has four features: it is published well ahead of the deadline, it is paired with an extended encashment window or a carry-forward exception for the affected people, it applies to a defined role or function rather than to individuals, and it has an escape hatch for genuine emergencies.

Carry-forward processing. Before you run it live, run it in preview. Pick fifteen employees across the range — a new joiner, a long-tenured person with a large balance, someone on notice period, someone who returned from long leave, someone who transferred location — and compute their expected closing and opening balances by hand. Compare against what the system produces. If any of the fifteen differ, stop and find out why before processing the other few hundred.

Archive a complete pre-rollover snapshot: every employee, every leave type, opening balance, accruals, consumption, adjustments, closing balance. Store it somewhere immutable. This single file resolves the majority of disputes that arise over the following year.

Opening balance publication. On day one or two of the new leave year, every employee should be able to see their opening balance, broken down into current-year entitlement and carried-forward days, with the expiry date of the carried days if there is one. Push it, do not just make it available — a notification with the number in it, not a link to a dashboard nobody opens.

Then open a query window with a clear end date. Most queries arrive in the first ten days. A defined window means you handle them as a batch rather than as a trickle across the year.

Post-close audit. Two weeks after rollover, take a sample and verify three things: that no employee's carried-forward figure exceeds the cap, that every lapsed day was communicated in advance, and that every encashed day was actually earned and within the eligibility rules. Document the sample and the result. It takes half a day and it is the thing that makes the whole close defensible.

Leave liability and provisioning

Finance needs a number before year end, and HR is the only function that can produce it. The number matters for three reasons: the accounting provision for accumulated compensated absences, the cash forecast for the encashment payout, and the accuracy of settlement amounts for anyone exiting.

How to think about the cost

At its simplest, the liability for any employee is the encashable portion of their accumulated leave, valued at the applicable daily rate under your policy, grossed up for any statutory employer contributions that apply to that payment under the rules you operate under. Your finance team and auditor will tell you the correct basis for valuation and whether an actuarial approach is required for your size and accounting framework — that is their call, not HR's. HR's job is to supply a clean, complete, reconciled day count.

What HR should supply:

• Accumulated days by employee, split by leave type, with only the encashable types included
• Days above the carry-forward cap separated from days within it
• A flag for employees on notice period, whose balance will convert to cash imminently
• A flag for employees with disputed or unreconciled balances
• The prior-year comparative, so finance can see the movement

What drives the number up

The forces are predictable, which means they are manageable:

• A high or absent carry-forward cap. The dominant driver. Every day above the cap that survives into the next year compounds.
• Long average tenure combined with low leave utilisation. A stable senior team that does not take leave is a growing liability, and also a burnout risk. The two problems have the same solution.
• Salary growth. Accumulated days are valued at current rates, so an appraisal cycle revalues the entire accumulated stock upward. This is the effect most teams forget. A modest increment across the board increases the leave liability by roughly the same percentage, without a single extra day being accrued.
• Upfront crediting with weak recovery. If the full year's entitlement is credited on day one and mid-year exits do not recover unearned days, you are paying out leave that was never earned.
• Comp-off with long or no expiry. Comp-off generated during a crunch period, never used, never expired, quietly becomes an encashable balance in some policy designs.
• Approval bottlenecks. If leave is hard to get approved, people stop applying and the balance grows. Rejection rates and liability growth are correlated.

Modelling the impact of a carry-forward cap change

This is the most useful piece of analysis HR can put in front of a founder, and it is simple arithmetic. The numbers below are illustrative only — invented to show the shape of the calculation, not drawn from any real company or dataset. Substitute your own.

Assume a company of 100 employees. Assume an average encashable daily rate of Rs 3,000 (again, illustrative). Suppose the current balance distribution at year end looks like this:

At an illustrative Rs 3,000 per day, the gross accumulated stock is 1,640 x 3,000 = Rs 49.2 lakh before any statutory gross-up. Now model three carry-forward caps:

The band totals are approximations for illustration; in a real model you would compute per employee, not per band, because the distribution inside each band matters.

What the model tells you, and what to take to the founder:

1. Tightening the cap does not save money immediately; it moves the money. Days above the cap either become cash now or become nothing. If they become cash, this year's outflow rises. If they lapse, this year's outflow does not rise but you take the full morale and trust cost in one hit.
2. The savings are structural, not immediate. A tighter cap reduces the rate at which the liability compounds in future years. That is a genuine benefit but it shows up over three or four years, not in this quarter.
3. A cap change concentrated on your most tenured people is a retention decision, not an accounting one. Look at scenario C: the 12 employees in the top two bands hold a disproportionate share of the days. They are usually your senior technical and functional leads. Ask whether the balance-sheet improvement is worth the conversation you will have to have with them.
4. Phasing beats cliffs. Moving from no cap to a 30-day cap over two years, with a grandfathered pool for existing excess balances, achieves most of the structural benefit with a fraction of the disruption.

Run this model in October or November for a calendar leave year, or December for a financial leave year. Finance then has a number, a range, and a set of levers, well before the close.

Managing the year-end leave rush

The rush is a forecasting problem with a communication solution. Handle it in four moves.

1. Forecast who has to burn balance

At W-7, produce a single report with four columns: employee, current balance, carry-forward cap, days at risk. Sort descending by days at risk. That list is the whole problem, and it is usually shorter than people expect — typically a minority of the workforce holds most of the at-risk days.

Break it down by team, and hand each manager their own slice. A manager who is told in the first week of November that four of her nine people are collectively carrying twenty-six days that must be used before year end can plan around it. The same manager told in the third week of December cannot.

2. Stagger approvals deliberately

Set a simple sequencing rule and publish it:

• Leave requests for the final six weeks are approved in the order received, subject to minimum staffing.
• Employees with days at risk get priority over employees who are within the cap, for the same dates.
• Requests submitted after a stated date are subject to available capacity only.

The second rule is the important one and it needs explaining. Someone whose leave will lapse has more at stake than someone whose leave will carry forward. Making that explicit prevents the perception that approvals are arbitrary, and it nudges people with safe balances to move their plans.

3. Set minimum-staffing rules per function, not per team

A minimum-staffing rule states what coverage a function must maintain. It should be written at the function level, because that is where cover actually comes from. Examples of the form it takes:

• Support: at least a stated number of people on shift per time band, and at least one person who can handle escalations.
• Payroll: a named primary and a named backup available for the days around the payroll cut-off.
• Delivery teams with a client commitment: at least one person per critical system available on any given day.
• Finance during statutory filing weeks: a named owner present for each filing.

Write these once, keep them in the manager pack, and apply them the same way every year. Ad hoc refusals feel personal; a published staffing floor does not.

4. Make reminders that people actually act on

Generic reminders fail because they contain no number. "Please use your leave before year end" is ignorable. "You have 8 days that will lapse on 31 December" is not.

A cadence that works:

That last one feels uncomfortable to send. Send it anyway. An employee who receives a clear statement of what lapsed, having received three prior warnings, cannot credibly claim their leave vanished. An employee who receives nothing will.

Sample message outlines — adapt the wording to your house style:

W-6 individual notification: State the exact number of days at risk and the date they lapse. State the carry-forward cap and the employee's current balance so the arithmetic is visible. Give two clear actions with links: apply for leave, or submit an encashment request if eligible. Give the encashment window closing date. One short paragraph, no policy preamble.

W-2 final notification: Repeat the number, note that it has not changed, state the application freeze date and the lapse date on separate lines so neither is missed. Add one line telling the employee to speak to their manager if approval is the blocker rather than intent. Copy the manager.

W-0 lapse confirmation: State the number of days that lapsed, the date, the policy rule under which they lapsed, and the new opening balance split between current-year entitlement and carried-forward days. Add the query window end date. Do not apologise and do not editorialise; state the facts.

Communicating a leave policy change without a morale problem

If you are changing carry-forward rules, caps or lapse treatment for the coming year, the change itself is rarely the problem. The handling is.

Give notice measured in months, not weeks. A carry-forward cap that tightens should be announced at least one full leave year before it bites, or with a transition arrangement if you cannot wait that long. Announcing a tighter cap in November for a December year end is functionally a retrospective change, because people made leave decisions all year on the old rules.

Grandfather existing balances. The cleanest transition protects what people already have. Freeze existing excess balances into a named grandfathered pool that can be used or encashed under the old rules, apply the new cap only to days accrued from the effective date onward, and set a horizon for the grandfathered pool if you want it to eventually disappear.

Consider transition credit. If you are shortening the carry-forward window or introducing an expiry on carried days, a one-time transition credit — extra days, an extended usage window, or a one-off widened encashment window — buys a disproportionate amount of goodwill relative to its cost. It changes the story from "the company took something away" to "the company changed the rules and made it right".

Put these things in writing, every time:

• The exact effective date, and what happens to balances accrued before it
• Worked examples with real arithmetic, covering at least three cases: an employee well within the cap, an employee just above it, and an employee far above it
• What happens to grandfathered balances and when, if ever, they expire
• The encashment position under the new rules
• Who to contact with a question, and the deadline for raising one
• A statement that the policy has been reviewed against applicable requirements, without quoting figures you cannot stand behind

Give managers an FAQ before employees hear anything. Managers who learn about a policy change from their own team lose authority, and a manager who improvises an answer creates a precedent you then have to honour. The manager pack should cover, at minimum:

• Why the change is being made, in one honest paragraph. If it is about liability control, say so. Employees can tell when a cost decision is being described as a wellbeing initiative.
• What happens to each person's existing balance, with their team's numbers attached
• What a manager may and may not approve as an exception, and who decides the rest
• What to say when asked "can you make an exception for me" — the answer should be a process, not a judgement call
• What to say when asked "is this legal" — the answer is that the policy has been reviewed against applicable requirements for the relevant states, and a route to escalate if someone believes otherwise

Do not bundle. Announcing a leave policy change in the same communication as a compensation change, a return-to-office change or a restructure guarantees that all of them are read as cost-cutting. Separate them by weeks.

Data hygiene checklist before rollover

Run this in the two weeks before close. Each item is a real source of wrong balances.

Two items on that list deserve emphasis. LOP reconciliation against payroll is the one that costs actual money in both directions, and it is worth a dedicated review rather than a spot check: pull the LOP days from the leave system and the LOP deductions from the payroll register for the full leave year, match them employee by employee, and investigate every difference. Approved-over-email leave is the one that is hardest to find, because by definition it is not in the system — which is exactly why the manager sign-off step exists.

Common mistakes, what they cost, and how to fix them

Frequently asked questions

Can we simply lapse all unused leave at the end of the leave year?

Not safely, and not as a blanket rule. Earned leave in particular is subject to statutory protections that vary by state and by establishment type, and those protections typically address the extent to which accrued leave must be permitted to accumulate. A company policy cannot fall below the applicable floor, and an employee's signature on a contract does not waive a protective entitlement. Before you write any lapse rule, get the current position confirmed for every state where you employ people. Separately from the legal question, blanket lapse is usually bad practice: it encourages people to take leave for the wrong reasons at the wrong time and it converts a routine administrative step into an annual grievance cycle.

What is a sensible carry-forward cap?

There is no universal answer, and any number quoted without reference to your state footprint, your policy design and your statutory floor is guesswork. The useful way to arrive at one is to work from constraints: start with the applicable statutory minimum for your states, set your cap comfortably above it, then model the liability at two or three candidate levels using your actual balance distribution, and check what each level does to your most tenured people. Pair the cap with an encashment route for the excess, so the choice employees face is cash versus lapse rather than just lapse. And whatever you choose, make sure your system enforces it automatically — a cap that depends on someone remembering to apply it is not a cap.

Should carried-forward leave expire, or sit in the balance indefinitely?

An expiry window on carried-forward days is generally the better design, for two reasons. It stops indefinite accumulation, which is what drives liability growth. And it creates a second, smaller deadline part-way through the new leave year, which spreads consumption out instead of concentrating it all at the year end. The requirement is that the expiry must be visible: employees need to see, in the portal and in reminders, how many of their days are carried-forward days and when those specific days expire. An expiry rule that exists in the policy but is invisible in the balance display will produce the same "my leave vanished" complaint as a silent lapse.

Our leave year and financial year are different. Is that a problem?

It is workable but it costs you something. The main friction is provisioning: the liability figure you hand finance at financial year close reflects a leave year that is only part-way through, so the number moves again before the leave year actually closes. You also end up running two separate close processes a few months apart, each with its own communication.

The compensating benefit is real though, particularly for Indian companies: a December leave year end keeps the leave rush away from March, when finance, audit, statutory filings and appraisals are already competing for the same people. If you keep the split, make the leave year the one that is visible to employees, and keep the financial-year reconciliation as an internal finance exercise rather than something employees have to understand.

How do we handle an employee who is on long leave when the leave year closes?

Decide it in the policy, before it happens, and handle it as a named exception in the runbook. Three questions need answers. Does accrual continue during the absence, pause, or stop — and does the answer differ by the type of absence, given that some categories of leave carry their own protections? Does the carry-forward cap apply normally to someone who had no realistic opportunity to use their balance? And who signs off their closing balance, since their usual manager may not have visibility?

The workable default: apply the normal cap but allow a documented exception for people whose absence made consumption impossible, approved by the HR head rather than the line manager, and logged with a reason. Also make sure someone actually contacts the employee before close — a lapse notification landing in the inbox of someone on extended medical leave, with no prior conversation, is a bad outcome regardless of whether the rule was correctly applied.

What should we do about employees on notice period at the leave year end?

Flag them separately at W-9 and treat them as a distinct cohort. Their balance is about to convert to cash, so accuracy matters more for them than for anyone else. Confirm four things: whether accrual continues through the notice period under your policy, how the balance is pro-rated to the last working day, whether any negative balance or unearned upfront-credited leave will be recovered from the settlement, and whether their carry-forward is even relevant given their exit date.

The practical rule: a notice-period employee's balance should be reconciled and signed off by both the employee and their manager before the leave year close, not during the settlement. Settlements are adversarial by nature and it is much easier to agree a number while the relationship is still normal.

How do we stop everyone taking leave in the last three weeks?

Forecast early, communicate individually, and give people a reason to move. The forecast at W-7 tells you exactly who has to burn balance — it is usually a minority of the workforce. Managers get their slice of that list and plan around it in November rather than reacting in the last fortnight.

Then give people alternatives to the rush: an encashment window that closes before the crunch, a carry-forward expiry window that lets some consumption spill into the new year, and priority in approvals for people whose days are genuinely at risk. Pair that with published minimum-staffing rules per function so that refusals are grounded in a stated rule rather than a manager's mood. If you need a blackout, publish it at W-8 with a compensating arrangement for the people it affects. What does not work is refusing requests in December without having said anything in October.

What is the minimum we should keep as an audit trail?

At a minimum: every accrual event with its date and basis; every application with submission time, approver, approval time and status; every adjustment with the actor, the before and after values and a written reason; every encashment with its computation; every lapse with the number of days and the policy rule applied; and a full pre-rollover snapshot per employee. Keep the reminder communications too — the fact that an employee was warned three times is evidence, and it is the evidence that resolves most disputes in your favour.

If any of that currently lives in a spreadsheet with no version history, that is the single highest-value thing to fix before the next leave year close. Not because of any one dispute, but because the absence of a trail turns every future disagreement into an argument about memory.

Closing the year properly is mostly about doing things early

Nothing in this guide is difficult. The runbook is a list of ordinary tasks. The liability model is arithmetic. The reminder cadence is three emails with a number in them. What makes the leave year close go wrong is not complexity — it is timing. Almost every failure traces back to something that should have happened in week eight happening in week one, or not at all.

So set the dates now. Decide the carry-forward cap and get it reviewed against the rules for your states. Pull the balance report and find out who is actually at risk. Give finance a draft liability number while there is still time to do something about it. Brief managers before employees hear anything. Publish the calendar. Then run the close in the order it is written down, and archive the snapshot when you are done.

If you are doing all of that in a spreadsheet, the close will consume more of your November or February than it should, and the audit trail you need in a year's time will not exist. CozyHR handles the parts that are mechanical — accrual by any basis you configure, carry-forward caps and expiry windows applied automatically at rollover, oldest-balance-first consumption, automated at-risk reminders with each employee's own numbers in them, manager sign-off workflows, and a complete adjustment log showing who changed what and why — so your team can spend the eight weeks on the decisions that actually need judgement. Set up your leave year, caps and encashment window in CozyHR and run a preview of the rollover before you commit to it, so the opening balances on day one are ones you can defend.