2027 Canadian Economic Outlook
Two clocks, one portfolio
Canada’s economy runs on two clocks. Families should set their watches by the slower one.
Dear audience, I know you want to hear that rates are about to fall. The honest answer depends on which clock you read.
Below, we look more closely at the two clocks and the themes related to them.
On the cyclical clock, which ticks with every inflation print and jobs report, the case for patience at the Bank of Canada is reasonable.
On the structural clock, which moves in decades, nearly everything points the other way: towards dearer capital, costlier construction and a higher floor under long-term interest rates.
Canada’s prime minister Mark Carney caught the mood in January when he called this moment a rupture, not a transition. In Ottawa that sounds like a lament. In Washington it has been policy for a decade towards Canada. Either way, the bill for a rupture arrives in the bond market. What follows is our reading of both clocks, one point and one chart at a time, and what we think a family balance sheet should do about them.
The cyclical clock: patience, with a stopwatch
Start with the clock everyone watches. Headline inflation rose to 3.1% in September, from 3.0% in August. Nearly all of that is petrol: gasoline prices are up 27.8% on a year ago, as conflict in the Middle East keeps crude dear. Strip out gasoline and inflation is 2.2%; grocery inflation has slowed to 1.7%.
The labour market, meanwhile, has wobbled. Employment fell by 68,000 in September. The Bank of Canada has held its policy rate at 2.25% since October 2025 and meets again on October 28. In September it warned that upside risks to inflation had grown, and markets now put the odds of a hike at roughly one in three.
Our reading: the Bank should hold rates steady and accept a temporary, energy-driven jump in inflation, because higher interest rates cannot bring down the price of oil. We expect it to do just that. Raising rates into falling employment to fight the price of crude would be treating the thermometer. But note what the chart implies. With headline inflation above the policy rate, real short rates are already negative, so this is hardly a tight stance. Anyone pencilling in deep cuts for 2027 is betting that the oil shock fades and that nothing takes its place.
The structural clock: a higher floor
The slower clock is harder to read and matters more. Three forces are pushing up the price of capital at once.
The first is the race to reindustrialise. China’s share of world manufacturing now blown past America’s, and Washington has spent a decade, under presidents of both parties, trying to claw production back with tariffs, subsidies and industrial policy. Factories need capital, and capital has a price.
In 2005 America and the European Union together made nearly half the world’s manufactured value added. Today China alone makes 27%, roughly ten points more than either of them.
The second is artificial intelligence. The four biggest American hyperscalers plan to spend about $733bn this year, nearly double 2025. Set against America’s economy that is roughly 2.3% of GDP, close to twice the intensity of the telecoms boom at its 2000 peak. And a boom that began on operating cash is now borrowing at the margin.
The third is the grid. Data centres, electric cars and reshored factories all run on electricity, and the West spent 15 years building little of it. Add rearmament, plus governments running deficits as far as the eye can see, and the private and public sectors are bidding for the same pool of savings.
None of this means rates must rise every year. It means the floor beneath long-term yields is higher than the 2010s taught investors to expect. Here we part company with the comfortable consensus that the long end has peaked. Mortgage spreads over government bonds have normalised, which tells you the spread has stopped widening. It says nothing about where the bond yield beneath it goes next.
Canada’s real problem: a decade of underinvestment
Canada’s most serious economic problem is not immigration, nor even tariffs. It is that its businesses stopped investing a decade ago (thank you... previous government ... sorry my personal feelings got the better of me here). By the third quarter of 2025 the average member of the workforce had 9% less capital to work with than in late 2015; for machinery and equipment the shortfall was 20%. For every dollar of new capital an American worker receives, a Canadian now gets about 55 cents.
The suspects are many: America’s 2017 tax reforms, which erased Canada’s tax edge; years of regulatory hostility to resource projects; a financial system that favours housing; and a surge of labour that made machines seem less pressing. Blaming red tape alone, as is fashionable, flatters the diagnosis.
Ottawa has at least noticed. In little more than a year it has created a Major Projects Office, legislated a fast track for projects in the national interest, signed up to NATO’s target of 5% of GDP for defence by 2035 and courted foreign capital with unusual zeal. It is the most ambitious industrial agenda in a generation. But announcements are not investment, and the binding constraint may be people. Estimates presented to the industry this autumn suggest the pipeline would need close to a million more construction workers by the end of the decade. Either the projects slip, or wages and materials costs climb. Neither makes a house cheaper, and the second makes the Bank of Canada’s job harder.
Canada and America: diversifying, slowly and partly by accident
The trade war has entered a second, nastier phase. Talks broke down in August. America imposed 50% duties on a list of Canadian goods from August 22nd, and Ottawa retaliated dollar for dollar, on $27.6bn of American imports, from September 8th. America’s share of Canadian goods exports has fallen to a record low of 68.3% over the 12 months to August, from about 74% a year earlier.
Before anyone declares diversification a triumph, look inside the numbers. More than half the rise in exports to the rest of the world is a single channel: unwrought gold shipped to Britain. Carmakers have found almost no substitute customers. And the exposure is far from evenly spread.
Ontario still sells the most south of the border, 36% of the national total, even though its own reliance has dropped by ten points. New Brunswick and Alberta remain the most dependent. Manufacturing in central Canada bears the brunt of the new duties; the Prairies, selling mostly energy, have so far been spared the worst.
The hopeful case is that America needs Canada more than its president admits. Canada is the biggest export market for 25 states, and a country rebuilding its industrial base needs Canadian energy, metals and water. We find that persuasive in the long run and of little comfort in the short. Protectionism can outlast an exporter’s balance sheet. Families should treat the relationship as a range of outcomes, not a reunion waiting to happen.
Housing: stabilising, not recovering
Housing explains itself once you stop staring at interest rates. Canada’s population grew by just 189,425 in the year to July 2026, or 0.5%, the slowest pace in percentage terms since 1915/16. Three years earlier it was growing by nearly 3% a year. The property business sells to new households, and when the supply of new households dries up, so does turnover.
The slowdown is not evenly spread. Ontario grew by 0.3%, Quebec by 0.2% and British Columbia by just 0.1%, because they took the largest share of the migrants who are no longer coming; Alberta grew by 1.5%. That is why Toronto and Vancouver have been the weakest markets, and why a flat national index can hide real regional losses for families concentrated in them.
Two policy dials will matter in 2027. The first is immigration: if Ottawa concludes it has squeezed too hard, growth will recover faster than current plans imply. The second is the ban on foreign buyers, which expires on January 1st 2027. The government has not said whether it will renew, revise or drop it; a revised regime that admits foreign money into new construction, as Australia does, looks likeliest. With three-quarters of Canadians backing the ban, do not expect a flood.
The case for a crash, meanwhile, is weak. Construction costs, pushed up by labour shortages and dear metals, put a floor under new-build prices. Our central case is flat national prices through 2027, with Toronto and Vancouver lagging the rest.
What it means for a family balance sheet
Forecasts are cheap; portfolios are not. Rather than bet on a single number, we plan around three paths and give each a weight.
| Scenario | Long-term rates | Canadian dollar | GTA housing | What tends to hold up |
|---|---|---|---|---|
| Grind | Elevated and sticky | Soft | Flat to lower | Diversification, US-dollar assets, shorter-duration bonds |
| Build-out | Higher, driven by growth | Firmer | Flat to modest gains | Real assets, Canadian cyclicals |
| Rupture deepens | Volatile, inflationary | Mixed | Weaker | Inflation hedges, liquidity |
Four principles follow, whichever path arrives. Own duration deliberately, as insurance rather than as a bet that the 2010s are coming back. Treat inflation sensitivity as a structural risk, not a passing one. For families whose obligations straddle Canada, America and Asia, hold currency where the money will actually be spent. And stress-test property: a family with most of its wealth in GTA real estate has made a large, concentrated bet on the slowest-growing corner of the country’s population. Plan mortgage renewals at today’s rates, not hoped-for ones.
The prediction
Economists who hedge every call are useless to the people who pay them. So here are ours, to be marked against the record at the end of 2027.
- The Bank of Canada holds at 2.25% next week and for the rest of 2026. Its next move will be up, not down, and we expect the policy rate to end 2027 between 2.50% and 2.75%. Anyone positioned for a return to emergency-low rates will be disappointed.
- Long-term yields end 2027 higher than they are today. Not dramatically, but enough to bury the idea that the long end has peaked. The floor has moved, and it will not move back.
- National house prices finish 2027 within 3% of where they started. Toronto and Vancouver underperform the national average; Alberta and Manitoba outperform it.
- The foreign-buyer ban is neither simply renewed nor scrapped. It is replaced by a narrower regime that admits foreign money into new construction.
- No comprehensive Canada–US trade deal is in force by the end of 2027. Some of today’s 50% duties will still be on the books when the CUSMA review grinds on.
- Population growth stays below 1% in 2026/27, and business investment per worker does not regain its 2015 level before 2028.
If we are right, the coming two years will reward discipline over daring: diversified, inflation-aware portfolios, deliberate duration and honest stress-tests of concentrated property. If we are wrong, it will most likely be because Ottawa’s building programme succeeds faster than its record suggests. That is a risk we would happily take. Set your watch by the slower clock.
Sources
- Bank of Canada policy rate and decision history (WealthNorth, from Bank of Canada series V39079, October 2026)
- The $733 billion year: measuring the AI infrastructure boom (The Meridian Report, September 2026)
- Canada’s Investment Crisis, Commentary 699 (C.D. Howe Institute, December 2025)
- Canada’s exports and the United States (Countability, from Statistics Canada, October 2026)
- Canada’s population in 2026 (Open Data Canada, from Statistics Canada table 17-10-0009-01)
Also drawn on, from news reporting: September 2026 CPI, September 2026 employment and the October 28 decision, Carney’s Davos speech, the foreign-buyer ban expiry, Statistics Canada demographic estimates, 2023 and Canada’s export partners by state.
两个时钟，一个投资组合
加拿大经济由两个时钟驱动。各家族应以走得较慢的那个为准。
亲爱的听众，我知道你们想听到利率即将下降。诚实的答案取决于你看的是哪个时钟。
下文将更深入地探讨这两个时钟及其相关主题。
在周期性时钟上，它随每一份通胀数据和就业报告而跳动，加拿大央行保持耐心的理由是合理的。
在结构性时钟上，它以数十年为单位运转，几乎一切都指向相反方向：资本更贵、建筑成本更高、长期利率的底部更高。
加拿大总理马克·卡尼（Mark Carney）在1月准确捕捉到了这种情绪，他称当下是一场“断裂，而非过渡”。在渥太华，这听起来像是哀叹；在华盛顿，这十年来一直是对加拿大的政策。无论如何，断裂的账单最终都会送到债券市场。以下是我们对两个时钟的解读，逐点展开、每点配一张图，并说明我们认为一个家族的资产负债表应如何应对。
周期性时钟：保持耐心，但手握秒表
先看人人都在关注的那个时钟。9月整体通胀率从8月的3.0%升至3.1%。几乎全部来自汽油：中东冲突令原油价格居高不下，汽油价格同比上涨27.8%。剔除汽油后，通胀率为2.2%；食品杂货通胀已放缓至1.7%。
与此同时，劳动力市场出现波动。9月就业人数减少6.8万。加拿大央行自2025年10月以来一直将政策利率维持在2.25%，并将于10月28日再次议息。央行在9月警告通胀上行风险已经加大，市场目前认为加息的概率约为三分之一。
我们的解读：央行应维持利率不变，接受由能源推动的暂时性通胀上升，因为加息无法压低油价。我们预计它正会这样做。在就业下滑之际为对抗原油价格而加息，无异于把体温计当成病因来治。但请留意图表的含义：整体通胀率已高于政策利率，实际短期利率已经为负，因此当前立场谈不上紧缩。任何押注2027年大幅降息的人，都是在赌油价冲击会消退，而且不会有新的冲击取而代之。
结构性时钟：更高的底部
较慢的时钟更难解读，也更为重要。有三股力量正同时推高资本的价格。
第一股是再工业化竞赛。中国在全球制造业中的份额如今已远超美国，而华盛顿十年来在两党总统任内，一直试图通过关税、补贴和产业政策夺回生产。工厂需要资本，而资本是有价格的。
2005年，美国和欧盟合计生产了全球近一半的制造业增加值。如今仅中国一家就占27%，比两者中任何一方高出约十个百分点。
第二股是人工智能。美国四大超大规模云服务商今年计划投入约7330亿美元，几乎是2025年的两倍。相对于美国经济，这约占GDP的2.3%，接近2000年电信业繁荣巅峰时期强度的两倍。而一场最初靠经营现金流推动的繁荣，如今在边际上正越来越依赖举债。
第三股是电网。数据中心、电动汽车和回流的工厂都依赖电力，而西方在过去15年里几乎没怎么建设电网。再加上重整军备，以及各国政府一眼望不到头的财政赤字，私人部门与公共部门正在争夺同一池储蓄。
这并不意味着利率必须年年上升，而是意味着长期收益率的底部高于2010年代让投资者习以为常的水平。在这一点上，我们与“长端利率已经见顶”这一令人安心的共识分道扬镳。按揭利率相对国债的利差已经正常化，这只说明利差不再扩大，却完全没有说明其下方的债券收益率接下来会走向何方。
加拿大真正的问题：十年投资不足
加拿大最严重的经济问题不是移民，甚至也不是关税，而是其企业在十年前就停止了投资（感谢……前任政府……抱歉，这里个人情绪占了上风）。截至2025年第三季度，平均每名劳动者可使用的资本比2015年底少9%；机器与设备方面的缺口达20%。美国劳动者每获得1美元的新资本，加拿大劳动者如今只能获得约55美分。
嫌疑因素众多：美国2017年的税制改革抹去了加拿大的税收优势；多年来监管环境对资源项目充满敌意；金融体系偏向住房；劳动力激增又让机器显得不那么迫切。时下流行把一切归咎于繁文缛节，这种诊断未免过于简单。
渥太华至少已经注意到了这一点。在一年多一点的时间里，它设立了重大项目办公室，为符合国家利益的项目立法开辟快速通道，承诺在2035年前达到北约国防开支占GDP 5%的目标，并以罕见的热情招揽外国资本。这是一代人以来最雄心勃勃的产业议程。但宣布不等于投资，而真正的硬约束可能是人。今年秋天向业界展示的估算显示，到本十年末，这批项目将需要近百万新增建筑工人。要么项目延期，要么工资和材料成本攀升。两者都不会让房子更便宜，而后者还会让加拿大央行的工作更加棘手。
加拿大与美国：多元化，进展缓慢且部分出于偶然
贸易战已进入第二阶段，而且更加激烈。谈判在8月破裂。美国自8月22日起对一系列加拿大商品征收50%关税，渥太华则以对等方式反制，自9月8日起对价值276亿元的美国进口商品征税。截至8月的12个月里，美国在加拿大商品出口中的占比降至68.3%的历史低点，一年前约为74%。
在宣布多元化大获成功之前，先看看数字内部。对世界其他地区出口的增长中，超过一半来自单一渠道：运往英国的未锻造黄金。汽车制造商几乎没有找到替代客户。而且风险敞口的分布远不均衡。
安大略省对美销售额仍居首位，占全国总额的36%，尽管其自身依赖度已下降十个百分点。新不伦瑞克省和阿尔伯塔省依然最为依赖美国。新关税的冲击主要落在加拿大中部的制造业；以能源出口为主的草原省份迄今幸免于最坏影响。
乐观的看法是，美国对加拿大的需要超过其总统所承认的程度。加拿大是25个州的最大出口市场，而一个正在重建工业基础的国家需要加拿大的能源、金属和水资源。我们认为这一论点从长期看有说服力，但在短期内帮助有限。保护主义持续的时间，可能比出口商的资产负债表撑得更久。各家族应把两国关系视为一系列可能的结果，而不是一场注定到来的重归于好。
住房：趋于稳定，而非复苏
一旦不再紧盯利率，住房市场的逻辑便不言自明。截至2026年7月的一年里，加拿大人口仅增长189,425人，即0.5%，按百分比计是自1915/16年度以来最慢的增速。三年前，人口年增速还接近3%。房地产行业的客户是新家庭，当新家庭的供给枯竭，成交量也随之萎缩。
放缓并不均衡。安大略省人口增长0.3%，魁北克省0.2%，不列颠哥伦比亚省仅0.1%，因为这些省份曾接收了如今不再到来的移民中的最大份额；阿尔伯塔省则增长1.5%。这就是多伦多和温哥华成为最疲弱市场的原因，也说明为何全国指数持平，却可能掩盖集中持有这些地区房产的家族所遭受的真实区域性损失。
2027年有两个政策旋钮值得关注。第一是移民：如果渥太华认定收紧过度，人口增长将比当前计划所暗示的更快恢复。第二是外国买家禁令，将于2027年1月1日到期。政府尚未表态是续期、修订还是取消；仿照澳大利亚的做法、允许外国资金投向新建住房的修订方案看来可能性最大。由于四分之三的加拿大人支持该禁令，不要指望外资会蜂拥而入。
与此同时，崩盘的理由并不充分。劳动力短缺和金属价格高企推高了建筑成本，为新房价格提供了底部支撑。我们的基准情景是全国房价在2027年前保持平稳，多伦多和温哥华落后于其他地区。
这对家族资产负债表意味着什么
预测很廉价，投资组合却不是。我们不押注单一数字，而是围绕三条路径做规划，并为每条路径赋予权重。
| 情景 | 长期利率 | 加元 | 大多伦多地区住房 | 通常较为抗跌的资产 |
|---|---|---|---|---|
| 胶着 | 高企且具黏性 | 偏软 | 持平至走低 | 分散配置、美元资产、较短久期债券 |
| 建设落地 | 走高，由增长驱动 | 走强 | 持平至小幅上涨 | 实物资产、加拿大周期性行业 |
| 断裂加深 | 波动、通胀性 | 涨跌互现 | 走弱 | 通胀对冲工具、流动性 |
无论哪条路径到来，都可以得出四条原则。审慎地持有久期，把它当作保险，而不是押注2010年代会卷土重来。把通胀敏感度视为结构性风险，而非一时之患。对于义务横跨加拿大、美国和亚洲的家族，应在资金实际支出的地方持有相应货币。还要对房产进行压力测试：一个大部分财富集中于大多伦多地区房地产的家族，实际上是对全国人口增长最慢的一隅下了一笔庞大而集中的赌注。按今天的利率规划按揭续约，而不是按期望中的利率。
我们的预测
每个判断都留有余地的经济学家，对付钱给他们的人毫无用处。以下是我们的判断，留待2027年底对照记录检验。
- 加拿大央行下周及2026年余下时间将利率维持在2.25%。其下一步行动将是加息而非降息，我们预计政策利率到2027年底将处于2.50%至2.75%之间。任何押注利率重回紧急低位的人都会失望。
- 2027年底长期收益率将高于当前水平。涨幅不会很大，但足以终结“长端已见顶”的说法。底部已经抬升，而且不会回落。
- 2027年底全国房价与年初相比变动不超过3%。多伦多和温哥华跑输全国平均水平；阿尔伯塔省和曼尼托巴省跑赢。
- 外国买家禁令既不会简单续期，也不会被废除。它将被一个范围更窄、允许外国资金投向新建住房的制度所取代。
- 到2027年底，加美之间不会有全面贸易协议生效。在美墨加协定（CUSMA）审议仍在推进之际，今天的部分50%关税仍将保留。
- 2026/27年度人口增长将保持在1%以下，而每名劳动者的企业投资在2028年之前不会恢复到2015年的水平。
如果我们是对的，未来两年将奖励纪律而非冒进：分散化、关注通胀的投资组合，审慎的久期管理，以及对集中持有房产的诚实压力测试。如果我们错了，最可能的原因是渥太华的建设计划比其过往记录所显示的推进得更快。这是我们乐于承担的风险。请以走得较慢的那个时钟为准。
资料来源
- 加拿大央行政策利率及决议历史（WealthNorth，基于加拿大央行V39079序列，2026年10月）
- The $733 billion year：AI基础设施投资热潮（The Meridian Report，2026年9月）
- 加拿大投资危机，第699号评论（C.D. Howe研究所，2025年12月）
- 加拿大出口与美国（Countability，基于加拿大统计局数据，2026年10月）
- 2026年加拿大人口（Open Data Canada，基于加拿大统计局17-10-0009-01表）
另参考新闻报道：2026年9月CPI、2026年9月就业数据与10月28日议息、卡尼达沃斯演讲、外国买家禁令到期、加拿大统计局2023年人口估计及加拿大出口伙伴。