Before the Magnificent Seven, there was Nortel. In the year 2000, a single Canadian company accounted for more than a third of the entire Toronto Stock Exchange. Two years later, its stock had fallen 99.6 percent. Seven years after that, it was bankrupt. Pension funds, index investors and ordinary Canadians holding "diversified" portfolios rode the whole thing down, because the index itself was the concentration.
No American outlet can tell this story the way Canada can, because it happened to us. And its lesson for today's AI boom is sharper than any Wall Street analogy: we have seen what happens when one beloved stock becomes the market.
This post is for informational purposes only and is not financial advice. Figures are current as of October 2026. Investment rules, account types (RRSP, TFSA) and tax treatment discussed here apply to Canada.
Key takeaways
- At its 2000 peak, Nortel was more than a third of the entire TSX, with a market value around C$398 billion and 94,500 employees worldwide.
- The stock fell from C$124.50 in July 2000 to C$0.47 by August 2002, a 99.6 percent wipeout. A C$1,000 investment at the peak was worth about C$3.80 by the bankruptcy filing.
- Because Nortel was over 35 percent of the TSE 300 index, every index-tracking Canadian fund and pension plan held it on the way down. There was no passive escape.
- Nortel filed for creditor protection on January 14, 2009, after accounting scandals, restatements and years of losses. Its executives were later acquitted of fraud in Canada in 2013.
- The direct parallel to today: the Magnificent Seven are 34.9 percent of the S&P 500, almost exactly Nortel's share of the TSX at its peak. Canada has seen this movie.
What happened to Nortel
Nortel began as Northern Telecom, the equipment arm of Canada's telephone monopoly. When the internet arrived, it bet everything on optical networking gear, renamed itself Nortel Networks in 1998, bought Bay Networks, and became the picks-and-shovels supplier of the dot-com gold rush.
On May 1, 2000, BCE spun off its 35 percent Nortel stake, worth about C$88.5 billion. In July, Nortel shares hit C$124.50; by September its market value reached about C$398 billion, more than a third of the entire TSX. It employed 94,500 people, 25,900 in Canada. CEO John Roth cashed C$135 million in options and was named Canada's CEO of the year.
Then the cliff: a missed target on October 24, 2000 (down 20 percent in a day), slashed forecasts on February 15, 2001 (down 33 percent), and a US$19.4 billion quarterly loss, the largest in Canadian corporate history at the time. The startup carriers buying Nortel's gear were going bankrupt, and Nortel had lent many of them the money to buy it.
Then came accounting: two restatements, an SEC investigation, CEO Frank Dunn fired "for cause" in 2004. The SEC alleged reserve manipulation; Nortel settled without admitting or denying, and the executives were acquitted in Canada in January 2013. Hackers had also been inside Nortel's network for years, per a 2012 Wall Street Journal report.
On January 14, 2009, a day before a US$107 million interest payment was due, Nortel filed for creditor protection in Canada, the US and the UK. It was carved up and sold to rivals; the leftover US$7.3 billion sat in a "lockbox" while courts in two countries fought over the split until 2015.
The index trap: why passive investors could not escape
Nortel was not a speculators' stock. At over 35 percent of the TSE 300, it was the market: any index-tracking fund, including pension plans, automatically held enormous quantities. The "diversified" Canadian equity fund of 2000 was a Nortel fund with banks attached.
Today's American index investor is in the same position. The Magnificent Seven were 34.9 percent of the S&P 500 on October 8, 2026, essentially Nortel's share of the TSX. Cap-weighted indexing concentrates by construction: the bigger a stock gets, the more the index must hold, until it falls. Our index fund big-tech crash math runs the arithmetic for your portfolio.
The vendor-financing footnote that rhymes with 2026
Nortel's collapse had an accelerant familiar to AI-trade watchers: Nortel and Lucent lent their startup customers the money to buy their equipment. Nortel extended about US$1.4 billion, sometimes 130 percent of the equipment price. Lucent's customer credit and guarantee agreements reached US$8.1 billion, about 24 percent of revenue, including US$2 billion to WinStar Communications, bankrupt in 2001, leaving a US$700 million write-off. The supplier financed its own demand: sales booked as revenue, backed by its own loans. When customers failed, Lucent took US$2.2 billion in bad-debt provisions in 2001 and US$1.3 billion in 2002.
The 2026 version: Nvidia's US$105 billion guarantee of OpenAI's Ohio leases, its US$2 billion CoreWeave stake plus a US$6.3 billion buyback obligation through 2032, and the US$573 billion web of AI financings Ares Management mapped. Fancier instruments, same structure: the vendor underwriting its own customer's purchases, explained in full in our circular AI economy explainer.
Could a Nortel happen again in Canada?
Honestly, yes in mechanics, though the cast would be different. Any single stock can come to dominate a cap-weighted index; that is how the math works. In 2021, Shopify briefly became Canada's most valuable company at a market capitalization around C$268 billion, roughly 5 to 6 percent of the TSX Composite. It was nowhere near Nortel's third of the market, but it showed the pattern can recur: a beloved growth story, a pandemic-era multiple, and an index that must buy more of it the higher it goes.
Today's TSX is bank, energy and materials heavy, with no single name near a third of the index. The real concentration risk for Canadians is US equity exposure, where the Mag 7's 34.9 percent is Nortel-level concentration, plus a newer wrinkle: CPP Investments committed US$1.75 billion to EQT's AI infrastructure strategy in July 2026 and holds US$300 million in xAI debt. Your pension holds AI credit, not just AI stocks.
One more uncomfortable parallel: Nortel was a real company with real technology and real revenue (about US$30 billion at peak). It was not a fraud; the executives were acquitted. Bubbles do not require villains, just a great story, cheap conviction and a financing structure that hides how much demand is borrowed. All three are present in the AI buildout today.
Practical next steps
- Check where your real concentration is. It is probably not the TSX. Add up your US equity index exposure and assume roughly a third of it moves with seven AI-linked stocks. Decide if that matches your risk tolerance, especially if you are within a decade of retirement.
- Remember that home bias cuts both ways. The TSX looks diversified next to the S&P 500 but concentrates in financials and energy; see our Canadian dollar explainer for the macro side. No single market is a diversification strategy by itself.
- Treat dividends as income, not armour. Our best dividend stocks for 2026 covers the appeal. But a dividend did not save Nortel shareholders. Yield is not a hedge against a 90 percent drawdown.
- Diversify across the thing that actually failed. Nortel taught that the failure mode is single-name concentration inside an index, not the index itself. Broad global exposure, rebalanced on a schedule, is the boring answer that would have worked.
- If you want AI exposure, size it like a risk. A deliberate 5 percent position you can watch fall 50 percent is a strategy. A 35 percent accidental position you did not know you had is a surprise. Our AI stocks guide walks through sizing.
The bottom line
Nortel was Canada's AI bubble, a quarter century early: a real technology company, a world-changing story, more than a third of the national market, then a 99.6 percent collapse no index investor could dodge. The concentration math today is nearly identical at 34.9 percent, and the vendor-financing playbook is back in fancier clothes. Canada paid tuition for this lesson once; check your portfolio's concentration before the market does it for you.
Sources
- Wikipedia, Timeline of Nortel: market cap C$398 billion (Sept 2000), over one-third of TSX, share price C$124 to C$0.47, creditor protection January 14, 2009. https://en.wikipedia.org/wiki/Timeline_of_Nortel
- Reuters, Key dates in the history of Nortel: peak share price C$124.50 (July 2000), bankruptcy filing January 14, 2009. https://www.reuters.com/article/us-nortel-timeline-sb-idUKTRE50D3N120090114/
- Cantech Letter: Nortel at 36% of the TSX, market value about $350 billion. https://www.cantechletter.com/2012/05/canadas-ten-most-valuable-tech-stocks/amp/
- SEC EDGAR: Nortel creditor protection proceedings, Petition Date January 14, 2009. https://www.sec.gov/Archives/edgar/data/72911/000119312511056204/d10k.htm
- The Motley Fool / Stock Analysis: Magnificent Seven weight of 34.9% in the S&P 500 as of October 8, 2026. https://www.fool.com/research/magnificent-seven-sp-500/
- aistockwire: Lucent US$8.1 billion vendor financing (~24% of revenue), Nortel US$1.4 billion. https://aistockwire.com/blog/nvidia-openai-105-billion-backstop-lucent-vendor-financing-august-19-2026
- Fiscus Consulting: Nortel and Lucent vendor financing comparison. https://fiscusconsulting.co.ke/billion-dollar-deja-vu-what-nortel-and-lucents-collapse-can-teach-us-about-the-ai-boom/
- Unite.AI / Tech Times: Nvidia US$105 billion OpenAI guarantee, August 17, 2026. https://www.unite.ai/nvidia-guarantees-up-to-105b-for-8-gw-ohio-ai-campus-leased-by-openai/
- TradingView / Stocktwits: Ares Management US$573 billion across 26 financings, flagged by Michael Burry September 28, 2026. https://www.tradingview.com/news/stocktwits:b144aaa57094b:0-big-short-michael-burry-drops-can-t-wait-info-on-573b-ai-financing-web-what-happens-if-spending-stops/
- CPP Investments (CNW press release): US$1.75 billion EQT/EdgeConneX commitment, July 2026. https://www.newswire.ca/news-releases/cpp-investments-partners-with-eqt-to-support-global-digital-infrastructure-growth-826226428.html
- Wealth Professional / Palos Funds: Shopify peak market cap about C$268 billion, 2021. https://funds.palos.ca/issue-no-47-shopify-takes-the-pole-position-as-canadas-largest-company/
Quick answers
Frequently asked questions
01
What happened to Nortel?
Nortel Networks, Canada's telecom equipment giant, rode the dot-com boom to become more than a third of the TSX with a market value around C$398 billion in 2000. When the bubble burst, its stock fell from C$124.50 in July 2000 to C$0.47 by August 2002. Years of losses, accounting restatements and an SEC investigation followed, and on January 14, 2009, Nortel filed for creditor protection in Canada, the US and the UK.
02
How much of the TSX was Nortel at its peak?
At its peak in 2000, Nortel accounted for more than a third of the total value of the Toronto Stock Exchange, over 35 percent of the TSE 300 index. Its market capitalization reached about C$398 billion in September 2000. Any fund tracking the index, including pension funds, automatically held large quantities of Nortel stock.
03
Could a Nortel happen again in Canada?
The mechanics can recur anywhere a single stock dominates a cap-weighted index. In 2021, Shopify briefly became Canada's most valuable company at around C$268 billion, about 5 to 6 percent of the TSX Composite. Today's bigger concentration risk for Canadians is in US equity exposure, where the Magnificent Seven are 34.9 percent of the S&P 500, a larger share than Nortel ever held of the TSX.
04
Why did Nortel's stock fall so far?
The telecom bubble burst, drying up demand from the startup carriers Nortel sold to. Nortel had extended about US$1.4 billion in vendor financing to shaky customers, so its own sales evaporated when they failed. It posted a US$19.4 billion quarterly loss, the largest in Canadian corporate history at the time, then suffered accounting scandals, restatements and executive fraud charges, though the executives were acquitted in Canada in 2013.
05
What is vendor financing and how did it hurt Nortel?
Vendor financing is when a supplier lends its own customers the money to buy its products. Nortel extended about US$1.4 billion of it, sometimes financing 130 percent of the equipment price. The sales looked real on the income statement, but they were funded by Nortel's own loans. When customers collapsed, the receivables became bad debt. It is the same structure as today's circular AI deals, where Nvidia guarantees OpenAI's data center leases.
