Queue item four, and the fourth article this session. The gate stopped the obvious version of this piece: an excise duty article on wine scored 5 against the brewery, distillery and cannabis excise articles already published. That was the gate working. The silo has enough excise. What it did not have is what happens after the excise, which is where the larger numbers live.
Key Takeaway
Ontario wine sold in an on-site winery retail store is taxed at zero. The same wine sold in that winery's off-site store is taxed at twelve percent. Non-Ontario wine is 19.1 percent on-site and twelve percent off-site. So Ontario wine should go on-site and blended wine should go off-site, and the tax system is content-blind in one channel and content-sensitive in the other.
The Verdict, Stated First
Five claims, in descending order of confidence.
One. Ontario now applies three different wine tax rates depending on content and channel. Zero percent on Ontario wine and wine coolers from on-site winery retail stores, 19.1 percent on non-Ontario wine from on-site stores, and twelve percent on owner's wine, whether Ontario or not, from off-site stores. This is Ontario's own published position and it is not in dispute.
Two. The zero rate is conditional on the channel, not just the content. An Ontario winery's own Ontario wine moves from zero to twelve percent by being sold in that winery's off-site store rather than at the vineyard. Nothing about the wine changes.
Three. The best channel therefore inverts between the two product types. Ontario wine is better on-site by twelve points. Non-Ontario wine is better off-site by 7.1 points. A winery holding both is optimised in opposite directions simultaneously.
Four. Off-site, the tax is content-blind, so cost alone decides the blend. At twelve percent on both, the tax differential between Ontario fruit and imported bulk disappears entirely, and on our own arithmetic the cheaper input wins by the full input saving with nothing offsetting it.
Five. A transitional rule let manufacturers keep tax over-collected between 1 and 24 April 2026. The rates took effect on 1 April and the legislation passed on 24 April. The difference is treated as part of the purchase price, so it is not remitted. That makes it revenue, and we suspect a good number of wineries have it sitting in a payable account. Least confident of the five, because it is an accounting inference from a tax rule rather than a stated accounting treatment.
Our Grades For These Claims
We grade our own sourcing before anyone else has to.
The rate table is as well sourced as anything we have published. It appears in the Annex to Ontario's 2026 Budget on the province's own budget site, and again, in operative form, in Ontario's notice to alcohol manufacturers on the 1 April 2026 tax changes. We also found it reproduced consistently by two professional firms and by trade press. Four independent renderings agree.
The transitional rule is from Ontario's own manufacturer notice, which states that amendments to the Liquor Tax Act, 1996 treat the difference in taxes collected between 1 April and 24 April 2026 as part of the purchase price, so manufacturers do not have to remit it.
Everything about a winery's costs is ours and invented. The $18 bottle, the $6.50 Ontario input cost and the $3.80 blended input cost are assumptions chosen to be plausible. We did not obtain any winery's cost structure.
The channel and market statistics are older and weaker. The 513-store winery retail network, its split between on-site and off-site, and the ownership concentration come from a 2021 foreign agricultural service report. The average price points for blended and VQA wine come from an industry statistics aggregator citing an association, and are older still. We use them to describe shape, not to compute anything, and no dollar figure in this article depends on them.
We did not read the Liquor Tax Act, 1996 or the amending legislation. That is the gap under claims one, two and five, and it matters more than usual because this is a rate table that changed four months before we wrote.
A Note On Method
What we obtained: the Annex to Ontario's 2026 Budget on budget.ontario.ca, for the proposed single rates on beer, wine and spirits effective 1 April 2026; Ontario's notice to Ontario alcohol manufacturers on the 1 April 2026 tax changes, for the operative rates and the transitional rule; two professional firm summaries of the budget measures; trade press reporting on the same; a 2021 United States Department of Agriculture foreign agricultural service report on the Ontario wine market, for the structure of the winery retail store network; and an industry statistics compilation for indicative price points.
What we did NOT obtain:
- The Liquor Tax Act, 1996, or the 2026 amending legislation. Every rate in this article comes from a government publication describing the law rather than from the law.
- The LCBO's current wholesale markup schedule. The budget states the rate changes align with a new LCBO wholesale markup structure, and we did not obtain that structure. This is a significant hole, because markup is very likely larger than the tax we are analysing.
- Any winery's actual cost of production, by channel or by product.
- Current data on the winery retail network. Our figures are from 2021 and the network has been the subject of policy change since, including support for off-site stores relocating to grocery.
- British Columbia, Quebec or any other province. This article is about Ontario.
One caution about scope. We analyse the provincial liquor tax only. Federal excise duty, HST, LCBO markup and cost of service, and container deposit charges all sit on the same bottle and none of them is modelled here. Nothing in this article is a calculation of what a bottle of wine actually costs to sell. It is a calculation of one layer, isolated deliberately because that layer changed.
Three Rates On One Bottle
Ontario's 2026 Budget proposed combining the basic, volumetric and environmental alcohol taxes into single rates, effective 1 April 2026, to align with a new LCBO wholesale markup pricing structure [1]. For wine the resulting rates are the whole of this article.
From Ontario's own notice to alcohol manufacturers, the position from 1 April 2026 is:
- Zero percent on Ontario wine and wine coolers purchased from on-site winery retail stores
- 19.1 percent of the retail price on non-Ontario wine and wine coolers purchased from on-site winery retail stores
- Twelve percent of the retail price on the owner's wine and wine coolers, Ontario and non-Ontario alike, purchased from off-site winery retail stores [2]
Read the third bullet twice. The headline everyone reported was the zero rate for Ontario wine, and the zero rate is real. But it attaches to Ontario wine in an on-site store. The off-site rate makes no distinction between Ontario and non-Ontario wine at all.
So the province did two different things in the same table. In the on-site channel it created a twelve percent difference between Ontario wine at zero and off-site treatment, and a 19.1 point difference between Ontario and non-Ontario wine. In the off-site channel it flattened the distinction entirely.
Whether that was the intention we cannot say and will not guess. What we can do is work out what it does.
The Same Wine, Two Doors
Start with the simplest case, because it is the one most likely to be costing money right now.
Take an Ontario winery with an on-site store at the vineyard and an off-site store in a plaza forty kilometres away. Take one bottle of its own Ontario wine, from its own fruit, at an $18 retail price.
Sold at the vineyard: zero provincial liquor tax.
Sold in the plaza: twelve percent of $18, which is $2.16.
Same wine. Same vintage. Same producer. Same price on the shelf. The tax difference is $2.16 a bottle and it is decided entirely by which door the customer walked through.
Ours. On a winery moving 40,000 bottles of its own Ontario wine through the off-site store in a year, that channel choice carries $86,400 of provincial liquor tax that the same volume sold at the vineyard would not attract.
We should be careful about what that number means. It is not a saving available for the taking. Off-site stores exist because customers do not drive to vineyards, and volume that moves in a plaza would not all move at the cellar door. The correct reading is not that the winery should close its off-site store. It is that the off-site channel now carries a twelve percent tax burden that the on-site channel does not, and that burden should be in the channel contribution analysis. If it is not, the off-site store looks more profitable than it is.
The Channel Preference Inverts
Now add the second product, and the structure gets more interesting than a single differential.
Ontario wineries do not only sell wine made from Ontario fruit. The winery retail store network also carries what the trade calls International-Canadian Blends, made by blending imported bulk wine with Ontario wine. For tax purposes what matters is that such a product is not Ontario wine.
Put both products through both channels at our $18 price:
- Ontario wine, on-site: zero percent. Tax $0.00
- Ontario wine, off-site: twelve percent. Tax $2.16
- Non-Ontario wine, on-site: 19.1 percent. Tax $3.438
- Non-Ontario wine, off-site: twelve percent. Tax $2.16
Now read down the two products.
Ontario wine is better on-site. Moving it off-site costs $2.16 a bottle.
Non-Ontario wine is better off-site. Moving it off-site saves $1.278 a bottle, because 19.1 percent becomes twelve percent.
That is an inversion, and we think it is the most useful thing in this article. A winery running both products through both channels does not have a channel strategy. It has two, and they point in opposite directions. The optimal routing is Ontario fruit to the vineyard store and blended product to the plaza, and any allocation of stock that does not reflect that is paying tax it need not pay.
None of this requires the winery to change what it makes, what it charges, or where it sells. It requires it to send the right product through the right door.
What The Inversion Is Worth
Ours, on an invented winery, and every input is an assumption.
Assume 40,000 bottles a year of Ontario wine and 40,000 bottles of blended product, all at $18, split evenly between the two stores because nobody thought about it.
Under that split, 20,000 Ontario bottles go off-site at $2.16 of tax each, which is $43,200. And 20,000 blended bottles go on-site at $3.438 each, which is $68,760. Total provincial liquor tax on the misrouted halves: $111,960.
Now route them correctly. All 40,000 Ontario bottles on-site at zero. All 40,000 blended bottles off-site at $2.16, which is $86,400.
The comparison is not quite that clean, because the correctly routed case still sells the same total volume through two stores of finite size, and we are ignoring the fact that a plaza store cannot absorb 40,000 bottles overnight. So treat the arithmetic as bounding rather than as a plan.
What it bounds is worth having. On these assumptions the tax cost of the routing decision alone, holding volume, price and product constant, is in the order of $25,000 to $45,000 a year on an 80,000 bottle winery. That is real money in a sector where most Ontario wineries have historically reported total sales of two million dollars or less.
And it is money that a costing system organised by product, which is the normal way, will never surface, because the differential is not a property of the product. It is a property of the pairing of product and store.
Off-Site Is Content Blind
The inversion has a second consequence that runs deeper than routing, and it concerns what the winery decides to make.
On-site, the tax system is strongly content-sensitive. Ontario wine pays zero and non-Ontario wine pays 19.1 percent. On an $18 bottle that is a $3.438 advantage to Ontario fruit, handed over by the tax system regardless of anything else.
Off-site, the tax system is completely content-blind. Both pay twelve percent. The advantage to Ontario fruit is exactly zero.
So the two channels do not merely favour different products. They embody different policies. One channel is an instrument of support for Ontario grape growing. The other is a flat consumption tax that does not care where the grapes came from.
Now put a cost on the two inputs, which is where it bites. Imported bulk wine is cheaper than Ontario fruit. That is why blends exist. Assume an Ontario bottle costs $6.50 of input and a blended bottle costs $3.80, so the blend saves $2.70 a bottle.
Off-site, where the tax is identical, that $2.70 is the entire story. The blend wins by $2.70 a bottle and nothing in the tax system offsets it.
Contribution per bottle off-site, ours: Ontario wine $18.00 less $2.16 tax less $6.50 input is $9.34. Blended wine $18.00 less $2.16 less $3.80 is $12.04.
The blend wins by $2.70, exactly the input saving, because the tax cancels.
Where The Blend Decision Flips
On-site, the same comparison runs the other way, and there is a crossover price at which it turns.
Contribution per bottle on-site, ours: Ontario wine $18.00 less $0.00 tax less $6.50 input is $11.50. Blended wine $18.00 less $3.438 tax less $3.80 input is $10.762.
Ontario fruit wins by $0.738 a bottle, despite costing $2.70 more to put in the bottle, because the tax system hands it $3.438.
That margin is thin, and the reason it is thin is that the tax advantage scales with price while the input saving does not. Work out where it flips.
The blend is better on-site whenever the input saving exceeds the tax cost, that is whenever $2.70 is greater than 0.191 multiplied by the retail price. Solving, the blend wins below a retail price of $14.14, and Ontario fruit wins above it.
So on-site there is a crossover price, and off-site there is not. Off-site the blend wins at every price, because the tax term is identical on both sides and cancels out of the comparison entirely.
We want to be careful with this. It is arithmetic on our assumed input costs, and the $14.14 crossover moves directly with them. If the input saving is $2.00 the crossover falls to $10.47. If it is $4.00 it rises to $20.94. The existence of a crossover on-site and its absence off-site does not depend on our numbers. The location of it depends on them entirely.
And Where The Market Actually Sits
There is a tempting observation here and we are going to make it and then immediately undercut it, because it is the kind of thing that reads as a finding and is not one.
Older industry figures put the average purchase price of an Ontario blended wine at around $8.22 and of a VQA wine at around $15.33 [7]. Our computed on-site crossover, on our own assumed input costs, is $14.14. The two observed averages sit on opposite sides of it.
That is a striking alignment and it is not evidence of anything.
The reasons it is not evidence are worth spelling out. The price figures predate the rate structure by years. The 19.1 percent on-site rate for non-Ontario wine took effect on 1 April 2026, and the price averages come from a compilation citing older association data. Prices formed under a different tax regime cannot have been sorted by this one. The crossover is computed from input costs we invented, so its location is our assumption, not a measurement. And the causation could easily run the other way, with blends being cheap because they are made from cheaper wine and sold to a more price-sensitive customer, entirely independently of tax.
We include it because a reader who knows the price points will notice the coincidence anyway, and because saying nothing about it would be a small dishonesty. The honest statement is that our arithmetic produces a crossover in the same region as the observed price gap, that this is interesting, and that we have no basis whatever for claiming one caused the other.
The Twenty-Four Days
Now a detail from Ontario's notice to manufacturers that is easy to read past and that we think has an accounting consequence.
The new rates took effect on 1 April 2026. The legislation implementing them passed on 24 April 2026.
For twenty-four days, therefore, the rates were operative but not yet enacted, and prices in the market may have been set on the old, higher rates. Ontario's notice states that recent amendments to the Liquor Tax Act, 1996 include rules treating the difference in taxes collected between 1 April and 24 April 2026 as part of the purchase price, with the result that alcohol manufacturers do not have to remit the difference resulting from the higher rate that may have been included in the price of beer, wine or spirits sold to consumers on or after 1 April 2026 [2].
Read what that does. Money was collected from consumers, in the price, on the footing that it was tax. It is not remitted. It is deemed to be part of the purchase price.
Ours, and stated as an inference rather than as advice. An amount collected in a price and not payable to anyone is not a liability. It is consideration. If a winery recorded those twenty-four days on its old rate table, it will have a credit balance in a tax payable account that has no corresponding obligation, and the correct treatment is almost certainly to recognise it rather than to carry it.
The amounts are not enormous. Twenty-four days is roughly 6.6 percent of a year. But it is the kind of balance that sits in an account for years because nobody has a reason to look at it, and it will not clear itself. Anyone closing a winery's 2026 year should look for it specifically.
We did not obtain the amending legislation and this reading rests on the notice's own description. A winery with a material amount at stake should get the provision read properly.
And The Deferred Returns
The same notice records a second transitional measure. To support a smooth transition, alcohol manufacturers were given additional time to file and pay their April, May and June 2026 returns [2]. Secondary summaries of the budget measures describe the filing and reporting requirements as deferred to July 2026, with no penalties where filed by 20 August 2026 [4].
Deferrals are welcome and they are also a trap of a specific and predictable kind.
Three months of liquor tax deferred is three months of a liability accruing without a payment against it. A winery that has been paying monthly for years, and whose cash forecast is built on that rhythm, now has a quarter with no outflow followed by a period with a larger one. If nobody adjusted the forecast, the business looks more liquid than it is for a quarter and then does not.
Ours. For a winery whose off-site store moves 40,000 bottles a year at $18 and twelve percent, monthly liquor tax runs about $7,200. Three months deferred is roughly $21,600 sitting in the bank that belongs to the province. On a business of that size that is not a rounding error, and it will be spent if it is not identified.
The second trap is the accounting one. A deferral changes when tax is paid. It does not change when it is incurred. A winery recording liquor tax on a cash basis, which is more common in small operations than it should be, will understate expense for the deferred period and overstate it afterwards, in a year where the rate structure also changed. Two changes in the same year, one to the rate and one to the timing, is exactly the configuration in which a variance analysis becomes uninterpretable unless somebody separates them deliberately.
Five Hundred And Thirteen Stores
The channel differential only matters if wineries have both channels, so it is worth knowing what the network looks like. Our figures here are from 2021 and should be treated as describing shape rather than current fact [5].
At that date the winery retail store network in Ontario comprised 513 stores, split between 221 located at wineries and 292 located away from them. Winery retail stores may sell only VQA wine, 100 percent Ontario wine, or International-Canadian Blend product made from the winery's own production. Overall the network accounted for less than fifteen percent of wine sales in Ontario.
The detail that matters most is the ownership. The report states that the off-site network has been consolidated over time and that the majority of those licences are now held by two companies, Arterra Wines Canada and Andrew Peller Limited, operating retail chains.
Sit that beside the rate table and something follows that we have not seen said.
The off-site channel, where the tax is content-blind, is concentrated in the hands of two large operators. The on-site channel, where the tax rewards Ontario fruit at 19.1 points, is where the small independent winery actually sells.
So the zero rate lands disproportionately on the small on-site producer, which is presumably the policy intention and is worth stating as a feature rather than an accident. And the content-blind twelve percent lands disproportionately on a channel controlled by two firms with the scale to source imported bulk efficiently.
We are describing structure, not alleging anything. But a small winery deciding whether the off-site channel is worth pursuing should understand that it is entering a channel where the tax system gives it none of the advantage its Ontario fruit earns at the vineyard.
The Vintage Problem Underneath All Of This
Everything so far has been about the tax on a finished bottle. Underneath sits a costing problem that predates all of it and that no rate change touches.
A vineyard is not productive when planted. Establishment costs are incurred over several pre-productive years before a single saleable grape is harvested. Then a harvest happens once a year, in a few weeks, and produces a vintage that may be sold over one year or five.
That creates three distinct timing mismatches in the same business.
Establishment against production. Costs incurred over years before any revenue, attaching to an asset with a long productive life.
Harvest against sale. A year's grape cost lands in a compressed window and is carried in inventory until bottles sell, which for a reserve wine may be years later.
Vintage against vintage. A poor harvest produces less wine from broadly similar fixed costs, so cost per bottle rises in exactly the year when there are fewer bottles to absorb it.
That third one is the same shape as the yield problem in meat processing, which we wrote about earlier in this session, and it has the same trap. A costing system that spreads annual fixed cost across actual production will report a high cost per bottle in a short year and invite the conclusion that the wine is less profitable. The wine is not less profitable. There is less of it.
We are deliberately not stating how vineyard establishment or vintage cost should be capitalised or amortised. That depends on the reporting framework and we did not obtain the relevant guidance. What we will say is that a winery whose management reporting does not separate the cost of a short harvest from the cost of a bad wine is going to draw the wrong conclusion from a cold spring.
Bulk And Bottled Are Different Assets
One more structural point, and it connects this article to the distillery piece published earlier this session.
Wine exists as bulk before it exists as bottles, sometimes for years. Those are not the same asset in any respect that matters.
Bulk wine is fungible, tradeable, and can still become several different finished products. Bottled wine is committed. It has a label, a vintage statement, a package, a price point and, after 1 April 2026, a tax rate that depends on where its content came from and where it will be sold.
The moment of bottling is therefore the moment at which a great deal of optionality is extinguished. Before bottling, the winery can still decide whether a parcel becomes an Ontario wine or part of a blend, which as we have shown is a decision worth $3.438 a bottle in tax at an $18 price in the on-site channel. After bottling, that decision is made.
Ours. This means the blend decision is a tax decision taken at a specific operational moment, and it is taken by a winemaker on quality grounds. We are not suggesting tax should drive winemaking. We are pointing out that the tax consequence is large, is computable in advance, and in most small wineries will not be in front of the person making the call.
It also means inventory carries a rate exposure. Bulk wine held at the date of a rate change has not yet been assigned to a channel or a content classification. Bottled stock has. A winery holding a large finished inventory when a rate structure changes has less room to respond than one holding bulk, which is a genuine and rarely stated advantage of bottling late.
The Definition The Whole Thing Turns On
An article built on a 19.1 point difference between Ontario wine and non-Ontario wine ought to be able to say where the line falls. We cannot, and the gap is worth naming rather than writing around.
The rate table distinguishes Ontario wine from non-Ontario wine. It does not, in the material we obtained, define either. The winery retail store rules describe what may be sold in those stores as VQA wine, 100 percent Ontario wine, or International-Canadian Blend product made from the winery's own production [5], which tells us there are at least three recognised categories and does not tell us how the two tax rates map onto them.
The obvious question follows immediately. An International-Canadian Blend is, by construction, part Ontario and part imported. Is it non-Ontario wine for the rate in every case? Is there a content threshold? Does VQA certification matter to the tax, or only to what may be sold in the store?
We do not know, and we are not going to reason our way to an answer from a rate table. The distinction is almost certainly defined in the Liquor Tax Act, 1996 or in regulations under it, and we did not obtain either.
Why it matters commercially is easy to state even without the answer. If the boundary is a percentage of Ontario content, then it is a cliff, and every cliff in a tax system creates a zone just below it where a small change in a blend produces a large change in tax. On an $18 bottle in the on-site channel that cliff is worth $3.438. A winemaker adjusting a blend by a few percentage points for balance could, on that structure, move the whole bottling across it without anyone in the building noticing.
We have seen this pattern before in this silo. The brewery excise article found a 3.5 times marginal cliff at 15,000 hectolitres, and the mortgage investment corporation article found a leverage test that fails for the whole year if it fails on one day. Threshold tests reward knowing exactly where the threshold is, and this is one we could not locate.
For a winery with material blended production, establishing the definition is the single highest-value piece of research available, and it is a question for someone who can read the Act rather than a rate table.
The Layer We Did Not Model, And Why It Probably Dominates
Honesty requires a section that undercuts the rest of the article, because there is a larger number sitting next to the one we analysed and we did not obtain it.
The budget states that the wine, beer and spirits rate changes were made to align with the implementation of a new LCBO wholesale markup pricing structure [1], and trade reporting notes the province describing a reduction in wholesale markups of about $200 million to help with the transition, with the LCBO becoming the exclusive wholesaler to licensees such as bars and restaurants from the same date [3].
Markup is not tax and it is very likely larger. Promotional material from a competing channel claims that around 65 percent of a shelf price at the liquor board has historically gone to taxes and markup combined [6]. We report that figure with considerable reluctance, because its source is a business whose commercial interest is in the number being large, and we could not verify it against the board's own published schedule, which we did not obtain.
What we can say without that figure is structural and still useful. The channel differential we have analysed, worth twelve points or 7.1 points depending on the product, sits inside the winery retail store network. It says nothing at all about the comparison between selling through a winery retail store and selling through the liquor board, because that comparison turns on markup and cost of service rather than on liquor tax.
So this article analyses the choice between a winery's own two stores. It does not analyse the choice between a winery's stores and the board. That is a larger question and a genuinely important one, and answering it needs the markup schedule we do not have.
If You Run A Winery
Five things, in the order we would look at them.
Route Ontario wine to the on-site store and blended product to the off-site store. On an $18 bottle that is worth $2.16 and $1.278 respectively, and it requires no change to what you make or charge.
Put the twelve percent into your off-site channel contribution. If the off-site store's profitability is calculated without the provincial liquor tax that the on-site store does not pay, it is overstated. That is not a small adjustment on a channel whose whole existence is a volume argument.
Find the twenty-four day balance. If you charged the old rates between 1 and 24 April 2026, the difference is not remittable and may be sitting as a payable that will never be paid.
Rebuild the cash forecast around the deferred returns. Three months of deferral is three months of tax in your bank account that is not yours. On our invented numbers that is roughly $21,600 on a modest off-site volume.
Compute your own crossover before the next blending decision. Ours came out at $14.14 on assumed input costs. Yours will be different, and the number that matters is the one built from your fruit cost and your bulk cost, not ours.
If You Advise One
Four checks we would run on any Ontario winery engagement this year.
Whether liquor tax is being computed by channel at all. A system that applies one rate to all winery retail store sales is now wrong in three different directions, and the error is invisible because the return will still foot.
The tax payable account, specifically for the April 2026 period. A credit balance with no corresponding obligation is the fingerprint of the transitional rule, and it is revenue.
Whether the 2026 variance analysis separates rate change from timing change. The rate structure changed on 1 April and the filing deadlines moved for the following quarter. Any year-over-year comparison that does not isolate the two will attribute one to the other.
Whether short-harvest cost per bottle is being read as a margin problem. It is a denominator problem. Fewer bottles absorbing similar fixed cost is not the same as less profitable wine, and the two look identical in a standard cost report.
And one thing to resist. Do not tell a client the zero rate means Ontario wine is untaxed. It means Ontario wine sold in an on-site winery retail store carries no provincial liquor tax. The same wine in the same winery's off-site store carries twelve percent, and federal excise, HST and markup are all still there in both channels.
What To Do
If you take one thing from this article, take the inversion. Ontario wine belongs on-site and blended wine belongs off-site, and a winery running both products through both channels is being pulled in opposite directions by a rate table it probably read as a single headline.
If you take two, take the content-blindness. Off-site, the tax makes no distinction between Ontario fruit and imported bulk, so cost alone decides the blend and cost favours the import. The support the province gave Ontario grape growing lives entirely in the on-site channel.
If you are advising an Ontario winery this quarter, the highest-value single question is whether the liquor tax calculation distinguishes on-site from off-site. If it does not, the figures are wrong in a way that no reconciliation will catch, because the return will balance either way.
The Limits Of This Analysis
Long and specific, because a limits section that is short is decoration.
This is Ontario and it is one tax layer. Federal excise duty, HST, LCBO markup, cost of service charges and container deposits all apply to the same bottle and none of them is in our arithmetic. Nothing here is a calculation of what it costs to sell wine.
We did not obtain the LCBO wholesale markup schedule, and markup is very likely larger than the tax we analysed. That is the biggest hole in the article and we gave it its own section rather than burying it here.
We did not read the Liquor Tax Act, 1996 or the amending legislation. Every rate comes from Ontario's budget annex and its manufacturer notice. Those are government publications describing the law, and four independent renderings agree, but they are not the law.
The 65 percent taxes-and-markup figure comes from a source with a commercial interest in it being high. We flagged it in the body and we repeat it here. It should not be relied on.
Every winery cost figure is invented. The $18 bottle, $6.50 Ontario input and $3.80 blended input are assumptions. The $14.14 crossover moves directly with them: a $2.00 input saving puts it at $10.47 and a $4.00 saving puts it at $20.94.
The volume assumptions are illustrative and the routing arithmetic is bounding, not a plan. A plaza store cannot absorb a vineyard store's volume and we said so where we used the number.
The network figures are from 2021. Store counts, the on-site and off-site split and the ownership concentration have all been the subject of policy change since, including provincial support for off-site stores relocating into grocery.
The price averages are older still and we computed nothing from them. We raised the coincidence between them and our crossover explicitly in order to say it is not evidence.
The twenty-four day treatment is our accounting inference from a tax rule. Ontario's notice says the difference is not remittable and is treated as part of the purchase price. That it should therefore be recognised as revenue is our reading and not a stated accounting treatment.
Cider is barely addressed. The queue entry covered wineries and cideries. The rate table we analysed speaks to wine and wine coolers, and cider sits in the beer rate structure in Ontario, which we did not model. That is a gap and it is ours.
Rates change. These took effect four months before we wrote and the legislation passed three weeks after they took effect. Check them.
Frequently Asked Questions
What are Ontario's wine tax rates from 1 April 2026?
Is Ontario wine really untaxed now?
Why does the best channel differ between products?
Does the tax system still favour Ontario grapes?
At what price does an Ontario blend beat Ontario fruit on-site?
What is the twenty-four day issue?
What happened to filing deadlines?
References
- Government of Ontario, Annex to the 2026 Ontario Budget, on budget.ontario.ca. Source of the proposal to combine the basic, volumetric and environmental alcohol taxes into single rates effective 1 April 2026 to align with implementation of the new LCBO wholesale markup pricing structure, and of the wine rates: a single rate of zero percent on Ontario wines and wine coolers and 19.1 percent on non-Ontario wines and wine coolers in on-site winery retail stores, and a single rate of twelve percent on owner wines and wine coolers sold in off-site winery retail stores. Note: the province's own budget document, and the strongest source here. It is a budget annex describing proposals, not the enacted legislation, which we did NOT obtain. Ontario
- Government of Ontario, notice to Ontario alcohol manufacturers on the 1 April 2026 tax changes. Source of the operative rate statements, being twelve percent of the retail price on the owner's wine and wine coolers, Ontario and non-Ontario, purchased from off-site winery retail stores; 19.1 percent on non-Ontario wine and wine coolers purchased from on-site winery retail stores; and zero percent on Ontario wine and wine coolers purchased from on-site winery retail stores. Also the source of the transitional rule, that amendments to the Liquor Tax Act, 1996 treat the difference in taxes collected between 1 April 2026 and 24 April 2026, the date the legislation passed, as part of the purchase price so that manufacturers do not have to remit it, and of the statement that manufacturers were given additional time to file and pay April, May and June 2026 returns. Note: the operative administrative notice and the load-bearing source for the twenty-four day point. Ontario
- Local news reporting of the 2026 Ontario Budget alcohol measures, syndicated across several Ontario outlets. Used for the reported reduction in wholesale markups of about $200 million to assist the transition, the LCBO becoming exclusive wholesaler to licensees from 1 April, the $100 million spent the previous year to reduce tax rates, markups and fees, the $35 million a year for five years to the wine sector for grape production, and $16.7 million over five years toward capital costs for off-site winery retail stores relocating to grocery. Note: secondary reporting of a budget, useful for context and programme amounts and not for anything load-bearing.
- Professional firm summaries of the tax measures in the 2026 Ontario Budget, used to corroborate the wine and spirits rate tables independently of the government sources, and for the description of filing and reporting requirements being deferred to July 2026 with no penalties if filed by 20 August 2026. Note: secondary, but valuable here precisely because it is independent: four separate renderings of the rate table agree.
- United States Department of Agriculture, Foreign Agricultural Service, report on the Ontario wine market, dated 2021. Source of the winery retail store network structure, being 513 stores split between 221 on-site and 292 off-site; the restriction that such stores may sell only VQA, 100 percent Ontario wine or International-Canadian Blend product from the winery's own production; the statement that the off-site network has consolidated with the majority of licences held by two operators; and that the network accounts for less than fifteen percent of Ontario wine sales. Note: a foreign government's market report, five years old at the date we wrote, describing a network that has since been the subject of policy change. Used for shape only. No dollar figure in this article depends on it.
- Promotional material from a craft winemaking business, cited only for its claim that around 65 percent of a liquor board shelf price has historically gone to taxes and markup, and for its reference to a new cost-plus wholesale pricing model. Note: the weakest source in this article by a wide margin. It is marketing copy from a business whose commercial interest lies in that percentage being large. We report it because we could not obtain the board's own markup schedule and thought the gap worth naming rather than leaving silent. It should not be relied on.
- Industry statistics compilation citing Ontario wine association data, used only for indicative average purchase prices of blended and VQA wine and for the observation that most Ontario wineries have reported total sales of two million dollars or less. Note: an aggregator citing an association, several years old, and predating the rate structure analysed here. We computed nothing from it and raised the one coincidence it produces expressly in order to say it is not evidence.