Anonymised, illustrative composite. The target’s payroll was $1,800,000, so the buyer’s model carried no severance reserve at all. Then it closed the second plant, and s.64(1) turned out to have a second limb with no payroll test in it.
At a glance
A private buyer acquired the shares of a metal fabricator running two plants roughly four kilometres apart in the same Ontario municipality. Price $9,400,000, annual payroll $1,800,000 across 118 employees. The thesis was consolidation: move all production into Plant A within a year and close Plant B.
The buyer’s employment model was one line long. Payroll is under $2.5 million; therefore no severance; therefore reserve nil. The model was right about the payroll and wrong about the Act.
ESA s.64(1) has a chapeau and two limbs, and only one limb mentions payroll: “An employer who severs an employment relationship with an employee shall pay severance pay to the employee if the employee was employed by the employer for five years or more and, (a) the severance occurred because of a permanent discontinuance of all or part of the employer’s business at an establishment and the employee is one of 50 or more employees who have their employment relationship severed within a six-month period as a result; or (b) the employer has a payroll of $2.5 million or more.”
Limb (a) has no payroll threshold. It is satisfied by a permanent discontinuance plus a headcount, and the headcount is a matter of what the buyer does after closing rather than anything on the target’s balance sheet at closing. A small-payroll target becomes exposed the moment the buyer’s own restructuring crosses fifty.
Read the chapeau carefully, because this is where models go wrong in the other direction. The five-year requirement sits above both limbs and applies to each of them. So the fifty is a trigger counted across everybody severed as a result of the discontinuance, while the entitlement accrues only to those of them with five years or more. Treating the fifty as the population that gets paid overstates the bill; treating limb (a) as subject to the payroll test understates it to nil.
Plant B closed over nineteen weeks in three tranches of 22, 17 and 12 — 51 people. Of the 51, 33 had five years or more, with average completed service of 9.6 years and average regular wages of $1,050 for a regular work week.
Severance under s.65(1) is the weekly wage multiplied by completed years plus part-years over twelve: 9.6 × $1,050 = $10,080 each, or $332,640 across the 33. Nobody approached the s.65(5) ceiling of 26 weeks, which on these wages is $27,300. Termination pay in lieu under s.61, calculated employee by employee on the s.57 scale across all 51, added a further $214,000.
None of it was recoverable. The working-capital true-up could not reach it because the liability did not exist at closing, and no representation was breached: the target had accurately disclosed a payroll of $1,800,000 and 118 employees. The buyer built the exposure itself in month seven.
The staggering was deliberate and it was aimed at the wrong provision. Section 58(1) is the group-notice rule: “Despite section 57, the employer shall give notice of termination in the prescribed manner and for the prescribed period if the employer terminates the employment of 50 or more employees at the employer’s establishment in the same four-week period.” The prescribed period under O. Reg. 288/01 s.3(1) starts at “Notice shall be given at least eight weeks before termination if the number of employees whose employment is terminated is 50 or more but fewer than 200.”
Three tranches nineteen weeks apart never put 50 people inside a four-week window, so s.58 was avoided and individual notice under s.57 applied. But the severance trigger in s.64(1)(a) counts 50 or more severed “within a six-month period”, and nineteen weeks is comfortably inside six months. Two provisions, two different clocks, one restructuring schedule built against only one of them.
The remaining argument was that the two plants were a single establishment. Under the s.1(1) definition that is right — an establishment is “a location at which the employer carries on business but, if the employer carries on business at more than one location, separate locations constitute one establishment if, (a) the separate locations are located within the same municipality” — and both plants were in the same municipality. Section 64(4) answers it: “A location shall be deemed to be an establishment under subsection (1) if, (a) there is a permanent discontinuance of all or part of an employer’s business at the location; (b) the location is part of an establishment consisting of two or more locations; and (c) the employer severs the employment relationship of 50 or more employees within a six-month period as a result.” The closed plant is its own establishment, so the survival of Plant A decides nothing.
The buyer paid $546,640 it had not reserved, funded by drawing on an acquisition facility that had been sized for capital expenditure. The consolidation still made sense — it delivered $1,240,000 of annual saving — but the payback moved out by five months and the capital programme slipped a year.
The lasting change was to the model. Employment exposure is now run twice: once on the target as bought, and once on the target as the buyer intends to operate it, with the second run keyed to the integration plan’s own headcount schedule. See statutory severance pay and what an asset buyer inherits for the inherited side of the same question.
The severance component was entirely a function of the buyer’s own schedule. Had the discontinuance been genuinely capable of being done with fewer than fifty severances in any six-month window — a smaller Plant B, or a redeployment programme that moved people to Plant A rather than out — limb (a) would never have engaged, and at $1,800,000 of payroll limb (b) never would either. The $332,640 would have been nil.
That is a statement about how the provision is built, not a scheduling trick. The six-month window runs from each severance, so a closure that genuinely needs 51 people to go cannot be dressed as 34 by pushing seventeen of them past an arbitrary date — the later severances still occur “as a result” of the same permanent discontinuance and still fall to be counted against the ones before them. What the buyer could have done, and did not, is price it.
The reserve line read “nil — payroll below threshold”. Any severance analysis that reaches its conclusion by citing the $2.5 million figure alone has read one of the two limbs of s.64(1).
The second tell was in the integration deck rather than the diligence file: a closure schedule with tranche sizes chosen to avoid a headcount trigger. Somebody had done that arithmetic against s.58’s four weeks and nobody had done it against s.64(1)(a)’s six months. When a plan is engineered around one statutory threshold, the useful question is which other thresholds sit on the same facts with different units attached.
A 30-minute call is enough to tell you whether AI pays for itself here.
Canadian small-business transaction data is not published anywhere, so most valuations in this country quote an American benchmark. The Deavo–Treadstone Acquisition Index is a daily record of Canadian listings built to replace that: asking-price distributions by province and city are published now, and days on market, departure rates and asking-to-sale spreads follow as the series lengthens. Leave an email and we will tell you as each measure lands.
No pitch, no listings. One email when the first report lands.
