Beyond Compounding: the guide as text

Opening Same returns. Opposite outcomes.

Two retirees get the exact same 30 years of market returns. Only the order differs.

• Broke in year 26 Actual order (25 withdrawals, then empty)
• $1.65M Reversed order (after 30 withdrawals, in today's dollars)
• 9.7% Average return, both (per year, before inflation)

Bad years at the start, while you're withdrawing, do far more damage than the same bad years at the end. This is sequence-of-returns risk.

Compounding assumes a smooth ride. Everything in this deck is about the bumps: which ones you can afford, and how to build around them.

Source: Damodaran (S&P 500 with dividends, 10-yr US Treasuries), Shiller CPI. $1M, 60% stocks / 40% bonds, 4% of the starting value withdrawn each year and raised with inflation, rebalanced yearly.

Map Ten acts, one idea

There is no riskless asset, only different risks. Pick the ones your timeline and finances can survive, keep costs and taxes low, and stay the course.

1. 1Compounding in the real worldInflation, volatility drag, and the three leaks: fees, taxes, bad timing.
2. 2What risk really isCrashes, inflation, sequence risk. When bonds help, and when they don't.
3. 3The worst casesGermany 1925–65, Japan, markets that vanished, and the one defence.
4. 4Time changes everythingHow the odds of losing money shift between 1 year and 30.
5. 5DiversificationOne stock can go to zero. Canada is 3% of the world.
6. 6FactorsFama and French in plain words, droughts included.
7. 7Total return vs risk-adjustedWhat the Sharpe ratio says, and when it doesn't matter.
8. 8Designing a portfolioTolerance, capacity and need. You are a bond (or a stock).
9. 9The Canadian toolkitTFSA, RRSP, FHSA, rent vs buy, one-ticket ETFs.
10. 10Putting it togetherA step-by-step process, four examples, and a framework explorer to try.

• Returns are real (after Canadian inflation), in CAD, before fees and taxes, unless a label says otherwise.
• Underlined words like drawdown show a plain-language definition.
• Charts show values when you hover over them, or tab to a chart and use the arrow keys. Charts built only from public government data also have a Table button.
• Slides marked Deep dive can be skipped; turn them off at the bottom in Present mode.
• In Present mode use ← → or swipe. Scroll view reads like an article.

Educational illustration, not personal financial advice. Past returns don't guarantee future results.

1.1 Inflation: the invisible tax

A dollar loses buying power every year. What matters is what your money can buy, not how many dollars you have.

• $375 $100 of 1980 groceries costs today (prices up 2.9% a year since 1980)
• $2,746 $100 of 1914 groceries costs today (more than a century of compounding, in reverse)

Inflation is compounding working against you. The 1970s averaged 7.5% a year; the 1930s had falling prices (−1.8%); the 2010s were calm at 1.7%.

So this deck measures everything as a real return: the return after inflation. A 5% nominal return with 3% inflation is about a 2% real return. That's how much more your money actually buys.

Source: Statistics Canada tables 18-10-0004 and 18-10-0005 (Consumer Price Index), 1914–2025. Decade figures are compound averages.

1.2 Volatility drag: why the average isn't what you get

Ups and downs cost you money even when the average stays the same. A 50% loss needs a 100% gain to break even.

+50%, then −50% = −25%

$100 rises 50% to $150, then falls 50% to $75. The average return says 0%. You actually lost a quarter.

What you earn is the compound annual growth rate. The bigger the swings (volatility), the further it falls below the average year.

In real data the gap is roughly swing² ÷ 2. From 1977 to 2025, Canadian stocks averaged 11.9% a year but compounded at 10.7%. A diversified all-in-one stock mix lost less to drag: 12.8% vs 11.9%. Bonds barely lost any (7.0% vs 6.8%).

Source: Illustration: 30 years of returns alternating above and below an 8% average. Real-data gaps: Canadian-dollar returns, Jan 1977–Dec 2025 (Ken French; Statistics Canada, with bonds estimated from Government of Canada yields).

1.3 The three leaks: fees, taxes and bad timing

Costs compound exactly like returns do, only against you. Try your own numbers.

A typical mutual fund (~2% a year, the MER) leaves you about 30% less than a 0.2% all-in-one ETF after 30 years. Buying high and selling low (the behaviour gap) costs about as much as a 1.5% fee.

• $642k You keep (≈ $354k in today's dollars)
• $0 Lost to fees (0% of the no-cost result)
• $0 Lost to bad timing (toggle on to include)
• $0 Lost to tax (none in this account)

You put in $190k. With no costs at all it would grow to $642k ($354k in today's dollars).

A TFSA never taxes growth or withdrawals.

Illustration only: a steady 6.83% a year before costs (our assumption), which real markets never deliver. Costs are subtracted as a yearly drag. Not personal advice.

Interactive in the guide: the controls on this slide only work with JavaScript.

Source: Expected return before fees: the deck's own assumption, today's Government of Canada bond yields (Statistics Canada) plus a stock premium you can change; inflation: the Bank of Canada's 2% target. The bar chart uses the default premium. Tax drag on a stock portfolio from 2026 federal + Ontario rates. Behaviour gap: Morningstar Mind the Gap 2026 (US funds, 10 years).

1.4 What you actually keep

Your result is the return, minus inflation, fees, taxes and bad timing. Small leaks decide whether compounding works for you.

• $38.8k Low-leak: $10k after 30 years (in today's dollars)
• $12.2k Leaky: $10k after 30 years (in today's dollars)

Same market, same 30 years. One investor ends with about 3.2 times the purchasing power of the other, and none of the difference came from picking better investments.

The rest of this deck is about the first term in the equation: the return, and the risk that comes with it.

Source: Approximation that subtracts yearly rates: expected return before fees = the deck's own assumption (Government of Canada bond yields from Statistics Canada plus the stock premium set on slide 1.3); inflation = Bank of Canada 2% target; fees as stated; tax drag for a stock portfolio in a non-registered account at ~$100k income (2026 federal + Ontario); behaviour gap 1.2 pts (Morningstar Mind the Gap 2026).

2.1 Risk is a menu, not a number

"Risky" usually means "it can drop a lot." That's one risk of six, and the safest-looking assets carry the others.

Crash risk

Losing a lot, fast, possibly right when you need the money.

US stocks lost 79% after inflation from 1929 to 1932.

Inflation risk

Your "safe" money quietly buys less every year.

Canadian T-bills lost 45% of their buying power from 1935 and took 51 years to recover.

Sequence risk

For anyone withdrawing, bad years early hurt far more than bad years late.

Same 30 years of returns: broke in year 26, or $1.65M left.

Concentration risk

Too much in one stock, one sector, one country or one house.

Nortel was over a third of the TSX in 2000, then went bankrupt.

Catastrophe risk

War, hyperinflation, a currency reset. It hits the assets people think are safe hardest.

German government bonds were wiped out in 1923 and again in 1948.

Shortfall risk

Taking too little risk to reach your goal at all.

In the US since 1928, cash lost buying power over 35% of 30-year periods.

Source: Examples drawn from later slides: Ken French and Shiller (US 1929), Statistics Canada (T-bills), German currency laws and Dimson-Marsh-Staunton (Germany), Damodaran and Shiller (US cash, 1928–2025).

2.2 Risk gets paid, on average

Over 50 years, each step up in risk earned a lot more. That extra return is the payment for sitting through the drops.

• $25.2k Cash
• $47.6k Bonds
• $473k All-in-one stocks

The gap between stocks and safer assets is the equity risk premium. These are total return numbers: price changes plus dividends and interest.

Over 150 years of US data, $1 in stocks grew to about $44,000 after inflation (7.1% a year).

The chart is on a log scale: equal steps are equal percentage changes. The next slides show the price of that extra return.

Source: Real (after Canadian inflation), CAD, before fees, Jan 1977–Dec 2025, year-end points. Ken French data library (US, international, emerging and Canadian stocks), Statistics Canada (T-bills, CPI; bonds are an estimate built from Government of Canada yields), Federal Reserve (USD/CAD). US 1871–2026: Shiller.

2.3 Crash risk

Stocks can lose most of their value and take years, sometimes decades, to come back.

A drawdown is the fall from a previous high. The recovery time is how long until you're back to that high. Dividends shorten it: after 1929 the S&P 500's price took 25 years to come back, but with dividends reinvested it took 15.4.

1929: The answer to "how long did 1929 take?" depends on what you measure: price only, 25 years (1954); with dividends, 15 yrs (1945); with dividends and after inflation, 16 yrs (1945), and permanently above only from 1949.

Source: Table: month-end data, nominal. S&P 500 price: Shiller (monthly averages). Total returns with dividends: Ken French data library (US market; US high-tech industry; Japan and Canada market portfolios in local currency). Chart: Ken French US market deflated by Shiller CPI, worst month of each year.

2.4 Inflation risk: the slow bleed of "safe" money

Cash and bonds rarely crash. Instead, inflation can erase half their buying power over decades.

• 0.7%/yr Cash since 1935, after inflation (worst −45%, 51 years to recover)
• −0.59%/yr Cash after tax at ~$100k income (worst −53%; never recovered)
• −63% Long bonds from 1946 (43 years to recover)

Your balance never goes down, so it feels safe. But a T-bill or GIC that pays less than inflation loses purchasing power every year, and the interest is fully taxed.

Bonds can drop too when rates rise: a broad mix of Government of Canada bonds (our estimate from Statistics Canada yields) fell −18% from mid-2020 and was still below that peak at the end of 2025.

Source: Real (after Canadian CPI), month-end, to Dec 2025. Statistics Canada 10-10-0122 (3-month T-bill yields from 1935; Government of Canada bond yields: the 20-year bond and the broad bond series are estimates built from yields). After-tax figures use 2026 federal + Ontario marginal rates applied to history.

2.5 Every asset has an ugly chapter

There is no riskless asset, only different risks. Stocks fail fast; bonds and cash fail slowly or all at once.

Stocks: fast

A crash at the wrong time.

• US 1929–32: −79% after inflation
• Germany: −91% (1945–48) after inflation
• Japan: −97% (1944–47) after inflation

Bonds: inflation or a reset

Promises to repay paper money lose when the money does.

• Germany 1923: paper-mark bonds wiped out (1 new Rentenmark replaced 1 trillion old marks)
• Germany 1948: government (Reich) bonds never converted to the new money
• Canada long bonds from 1946: −63%, 43 years to recover

Cash: slow

Inflation and tax, a little every year.

• Canadian T-bills from 1935: −45%, 51 years to recover
• German savings in 1948: 100 Reichsmarks became 6.5 Deutsche Marks
• Cash over 1900–2000, after inflation: Japan −2.0% a year; Germany −0.6% a year even leaving out 1922–23

The question isn't "which asset is safe?" It's "which risk can I survive?"

Source: Real, in each investor's home currency. US: Ken French + Shiller CPI. Germany and Japan: Dimson, Marsh & Staunton, Triumph of the Optimists (2002), and German currency laws (quoted). Canada: Statistics Canada (bond returns estimated from yields).

2.6 Bonds' job: crash insurance, and it mostly works

When stocks crashed, high-quality bonds usually held up or rose. A 60/40 mix roughly halved the damage.

• 1.5× vs 0.46× US 1929–32, after inflation (Treasuries vs stocks (prices fell 23%))
• 2.4× vs 0.44× Japan 1990–2012, after inflation (government bonds (est.) vs stocks)

In 2007–09 stocks fell 40% while Canadian government bonds rose 17%. A 60/40 investor lost 21%, and could sell bonds to buy cheap stocks: that's rebalancing.

Recessions push interest rates down, which pushes bond prices up. That is why bonds earn less than stocks over time and still earn their place: diversification pays you when you need it most.

Source: CAD, nominal, Jan 1977–Dec 2025: returns over each peak-to-trough fall of the all-in-one stock mix (Ken French); bonds = Government of Canada estimate built from Statistics Canada yields. US 1929–32: real, Damodaran + Shiller CPI. Japan 1990–2012: real, Ken French (stocks), bond estimate from OECD 10-year government bond yields, World Bank CPI.

2.7 When the insurance fails

When inflation is the problem, stocks and bonds can fall together. 2022 was the textbook case.

Correlation by decade: 2000s −0.09, 2010s −0.17, 2020s so far 0.40. Years when both lost after inflation: 1981, 2022.

Rule of thumb: Bonds protect against recessions, not against inflation. Inflation protection comes from stocks over long periods, and from short-term or inflation-linked bonds over short ones.

Source: Jan–Sep 2022 (the stock mix's peak-to-trough), CAD, nominal: stocks = all-in-one global mix (Ken French data library), bonds = Government of Canada estimate from Statistics Canada yields, mixes rebalanced monthly; T-bills and CPI: Statistics Canada. Correlation: monthly CAD returns of the stock mix vs Canadian bonds over rolling 36-month windows, 1977–2025, shown at year-end.

2.8 Sequence risk: the retiree time machine

Once you're withdrawing, the order of returns matters as much as the average. Pick a retirement year and see.

• Broke in year 26 60/40 result (money ran out in 1991)
• $0 Lowest point (real value at its worst)
• $40k Withdrawn each year (in today's dollars)

1929 retiree at 4%: all stocks ran dry after 21 years of withdrawals, while 60/40 still had about $524k after 30. In 1966 every mix ran out, because a decade of inflation hit stocks and bonds together. The safe withdrawal rate research came from cases like these.

Interactive in the guide: the controls on this slide only work with JavaScript.

Source: US data: Damodaran (S&P 500 with dividends, 10-year Treasuries), Shiller CPI, 1928–2025. $1M start, withdrawals taken at the start of each year and raised with inflation, rebalanced yearly, real dollars. Failure rates: all 69 rolling 30-year US windows. Morningstar safe withdrawal rate 3.9% (Dec 2025, US data).

3.1 Germany: the stock investor

Through hyperinflation, depression, war and a currency reset, German stocks lost almost everything at the worst moment. Over the whole century they still beat inflation; German bonds and cash did not.

• −91% German stocks after WWII (1945–48) (after inflation)
• −89.6% Worst single year (1948) (the currency-reform year)
• +155.9% The very next year (1949) (after inflation)

If you needed your money in 1948, you were ruined. If you could wait, German stocks came back and still beat inflation over the century, by 3.6% a year. German bonds and cash lost buying power.

That difference is risk capacity: not how brave you are, but whether your life can wait out the worst.

Source: Dimson, Marsh & Staunton, Triumph of the Optimists (Princeton University Press, 2002), chapter 4: real returns in local currency, quoted. Canada shown from the same table for comparison. German bond and bill figures exclude the hyperinflation years 1922–23.

3.2 Germany: what happened to the "safe" investor

German savers in bonds and bank accounts were wiped out twice: by hyperinflation in 1923 and by the currency reform of 1948.

• +3.6%/yr German stocks, 1900–2000 (after inflation)
• −2.2%/yr German bonds (after inflation, even leaving out 1922–23)
• −0.6%/yr German cash (bills) (after inflation, even leaving out 1922–23)

1923: hyperinflation

• Prices exploded through 1922 and 1923.
• In November 1923 the currency was reset: 1 Rentenmark = 1 trillion paper marks.
• Anyone owed paper marks (bondholders, savers, lenders) was wiped out.
• Shares still owned real businesses: the nominal stock return in 1923 was about 305 billion percent.

1948: the currency reform

• Wages, prices and rents converted 1 to 1.
• Debts and bonds converted 10 to 1.
• Savings: 100 Reichsmarks became 6.5 Deutsche Marks, a 93.5% loss of buying power overnight.
• Government (Reich) bonds were never converted, and were cancelled for good in 1957.

A bond is a promise to repay a fixed amount of money. When the money itself is destroyed, the promise is worth nothing. A share of a factory still owns the factory.

Source: German currency laws: Rentenmark decree (Oct 1923); Currency Conversion Law (Umstellungsgesetz, 1948); General War Consequences Act (Allgemeines Kriegsfolgengesetz, 1957). 1923 stock return and century returns: Dimson, Marsh & Staunton, Triumph of the Optimists (2002), ch. 4.

3.3 Japan: two disasters

War wiped out Japanese stocks. Then the 1989 bubble left them below their peak for decades, while Japanese government bonds quietly compounded.

• −97% Stocks after WWII (1944–47) (after inflation)
• 34 years Nikkei 225 price index back above its 1989 peak (22 Feb 2024 (price only))
• 28 yrs With dividends, back to the 1989 peak (Sep 2017, in yen before inflation)
• 2.4× vs 0.44× 1990–2012, after inflation (government bonds (est.) vs stocks)

Bonds carried Japanese investors through the long slump. Then the roles reversed: from 2013 to 2025 stocks grew 4.4× after inflation, and over 1990–2025 as a whole stocks (1.9×) and bonds (1.9×) ended roughly level. Holding both was the calm path through.

Source: Stocks: Ken French data library, Japan market portfolio, total return in yen, year-end points (not the Nikkei 225 price index); peak, bottom and recovery dates from month-end data. Bonds and cash 1990–2025: estimates built from OECD 10-year government bond yields and the overnight call rate (via FRED); inflation: World Bank CPI. War years: Dimson, Marsh & Staunton (2002). Nikkei price record: news reports, Feb 2024.

3.4 The worst cases, and markets that vanished

The deepest losses in history come from wars, hyperinflations and revolutions. Some markets disappeared entirely.

Bonds were no shelter in those countries: over 1900–2000, German bonds lost 2.2% a year after inflation (even leaving out 1922–23) and Japanese bonds lost 1.6% a year.

Why US history looks so good: We only study the countries that survived. From 1921 to 1996 the US had the best real stock return of 39 markets (4.3% a year); the typical market earned 0.8%. Russia and China went to zero and drop out of most datasets. That's survivorship bias.

Source: Germany and Japan: Dimson, Marsh & Staunton, Triumph of the Optimists (2002), ch. 4, quoted (real, local currency). US 1929–32: Ken French + Shiller CPI. Markets that vanished: Dimson-Marsh-Staunton; survivorship: Jorion & Goetzmann (Journal of Finance, 1999).

3.5 The only defence: own the world

In each disaster the damage came from having everything in one country or one kind of asset. Spreading out is the only defence that doesn't depend on guessing which one fails.

Caveat: Hindsight warning: the US was the best major market of this period, which flatters the foreign side. The point isn't picking the winner; it's never having everything in the one that loses. Today a single all-in-one ETF holds 8,000 to 14,000 companies around the world for about 0.2% a year.

Source: Real multiples in each investor's home currency. US: Damodaran + Shiller CPI. Japan: Ken French stocks, bond and cash estimates from OECD yields, World Bank CPI. Canada: Ken French + Statistics Canada (bonds estimated from yields). Japanese investor comparison: Ken French Japan and US market returns, Federal Reserve yen/dollar rates, World Bank CPI.

4.1 Odds of losing purchasing power

Over long horizons, the "safe" assets lost purchasing power more often than stocks did, at least in the markets and decades we can measure.

Stocks, 10-year Treasuries and T-bills.

• 0.0% Stocks, 30 years (of 69 periods ended below the start, after inflation)
• 33% Bonds, 30 years (of 69 periods ended below the start, after inflation)
• 35% Bills (cash), 30 years (of 69 periods ended below the start, after inflation)

Over one year, every asset lost purchasing power roughly a third of the time. Stretch the horizon and stocks pull away, while US bonds and bills stayed near one in three.

In the US since 1928, stocks beat bonds in 85% of 10-year, 99% of 20-year, 100% of 30-year windows.

Long-history check: These samples are flattering. Outside the US our data starts in 1975, a period with no world wars or hyperinflations, and the US was the best market of the century. A study of 39 developed markets from 1841 to 2019 found a 12% chance that stocks lost purchasing power over 30 years, against 1.2% using US data alone (Anarkulova, Cederburg & O'Doherty, 2022).

Caveat: Windows overlap, so they aren't independent. Markets that vanished (Russia 1917, China 1949) are missing, which makes every asset look safer than it was.

Source: US: Damodaran annual returns (S&P 500, 10-yr Treasuries, T-bills) with Shiller CPI, 1928–2025. 22 developed markets: Ken French country returns in local currency with World Bank CPI inflation (US from 1927, the other 21 from 1975 or later, to 2025). Canada: Statistics Canada (T-bills, bond estimate) and Ken French (stocks). Overlapping windows, real returns. Long-history check: Anarkulova, Cederburg & O'Doherty, Journal of Financial Economics 143 (2022).

4.2 The range of outcomes narrows with time

Annual results swing wildly. Twenty-year results, in Canada since 1977, never went negative for a diversified stock mix.

• −32.7% Worst 1-year (per year, after inflation)
• −2.6% Worst 10-year (per year)
• 2.6% Worst 20-year (per year)
• 6% 10-year periods losing to inflation (share of all 10-year windows)

The shaded funnel is every rolling period since 1977: the outer band runs from the worst to the best, the inner band holds the middle half. Time doesn't make stocks safe, but it shrinks the range of what a year can do to you.

Caveat: 1977–2025 holds only about 2½ independent 20-year periods, in an era of falling interest rates that flattered bonds. That's why slide 4.1 adds 22 markets, US history back to 1928 and a 39-market study.

Source: Our CAD series, Jan 1977 – Dec 2025, rolling monthly windows, real annualized returns. All-in-one = XEQT-like global stocks (Ken French data); bonds = Government of Canada bond estimate built from Statistics Canada yields; cash = 3-month T-bills.

4.3 Money you need soon

Match the asset to when you need the money. A down payment due in three years has no time to recover from a crash.

Put $50,000 aside for a down payment needed in 3 years. What was left in the worst 3-year stretch?

• 10% All-in-one stocks (of 3-year periods ended below $50k · median $70.4k)
• 6% 60% stocks / 40% bonds (of 3-year periods ended below $50k · median $65.7k)
• 0% T-bills (cash) (of 3-year periods ended below $50k · median $56.5k)

Stocks usually won, and the median result was higher. But about one time in 10 the buyer came up short, at worst by $19.3k.

For money with a date on it, the common tools are a HISA, a GIC laddered to the date, or short-term bonds. Long-term money is a different bucket with a different job.

Source: Our CAD series, every 3-year window from Jan 1977 to Dec 2025. Nominal dollars (what the account would show).

5.1 One stock can go to zero

Nortel was about a third of Canada's stock market in 2000 and ended worthless. Most single stocks lag cash; owning the index guarantees you own the rare big winners.

• ≈35% Nortel at its 2000 peak (of the TSE 300 (sources say 35–37%); worthless after bankruptcy)
• 57% Non-US stocks that lagged T-bills (1990–2020, over each stock's life (US: 55%))
• 2.4% Firms that created all net wealth (of 64,000 stocks worldwide, 1990–2020)

Bessembinder's team followed 64,000 stocks around the world. Most of them did worse over their lifetimes than one-month treasury bills. A tiny minority produced all of the stock market's net gains.

That's why picking a few stocks is a different bet from owning the market: you'll probably miss the handful of big winners, and one collapse can sink a concentrated portfolio. An index fund owns all of them by design.

Even the index can be concentrated: the top 10 TSX stocks are 39% of it, and 16 stocks make up half.

Caveat: Nortel is one famous example; the evidence is the 64,000-stock base rate, not any single story.

Source: Bessembinder, Chen, Choi & Wei, Financial Analysts Journal 79(3), 2023 (64,000 stocks, 1990–2020). Nortel: CBC, Globe and Mail, Toronto Star. TSX holdings: iShares XIC, Oct 26 2026.

5.2 Canada is 3% of the world

The TSX is a fine market, but it's small and lopsided: banks, oil and mining, with almost no health care.

• 68% Financials + energy + materials (of the TSX)
• 16 Stocks that make up half the TSX (of 217 holdings (S&P 500: 22))
• 22.6% Big Six banks (of the TSX)
• 0.3% Health care in the TSX (vs 9.3% of the S&P 500)

Holding much more of your own country than its share of the world is called home bias. The US has its own concentration: information technology is about 40% of the S&P 500 and one company, NVIDIA, is 7.9%. A market-cap weighted world fund spreads both bets.

Source: MSCI ACWI IMI factsheet, Sep 30 2026 (South Korea, China, France and India derived from MSCI market caps); iShares XIC and State Street SPY holdings, Oct 26 2026.

5.3 Countries take turns

No country stays on top. Japan went from 2nd-best in the 1980s to last in the 1990s; the US went from 19th in the 2000s to 1st in the 2010s.

• 18 Different #1 countries (across 49 calendar years)
• 1 Years Canada was #1 (US: 3)
• 10th Canada's median rank (of about 20 countries)

Source: Ken French international country returns (MSCI 1975–2006, Bloomberg 2007+), calendar years, in US dollars so countries are comparable.

5.4 Canada vs the world, a fair fight

Going global won most decades. But Canada's worst decade was milder than the world's, and a blend had the smoothest ride.

• 33% Canada beat world weights (of 40 10-year windows ending each December)
• 7.1% vs 8.6% Real return per year (Canada only vs world weights, 1977–2025)
• 1.9% vs −3.2% Worst 10 years, per year (Canada vs world (all-in-one −1.7%))

The case against a Canada-only portfolio: it's 3% of the world and lopsided, and it lost most 10-year races.

The case for keeping some Canada: Canadian dividends get the dividend tax credit, there's no foreign withholding tax, fees are low, and your spending is in Canadian dollars.

Middle ground: Common guidance lands well above 3% but well below the 50%+ many Canadians hold: about 30% (Vanguard), about a third (PWL Capital). All-in-one ETFs hold 25–30% Canada.

Source: Our CAD series (Ken French stock returns, Federal Reserve USD/CAD), Jan 1977 – Dec 2025, real, before fees; fixed weights rebalanced monthly, shown as year-end points. Guidance on the Canadian share: Vanguard (2023) ≈30%, PWL Capital ≈⅓.

5.5 The loonie is a shock absorber

The Canadian dollar tends to fall when stocks fall, so foreign stocks lose less in CAD during a crash. It cuts both ways in a commodity boom.

• -0.43 USD-vs-CAD move vs Canadian stocks (correlation, 1977–2025: negative = cushion)
• +0.6% Currency effect, long run (per year on US stocks, 1971–2025)
• −4.5% to +3.3% Currency effect, any 10 years (per year (worst decade ended 2012, best 2001))

This is why many Canadians don't hedge foreign stocks: the currency usually helps exactly when it's needed. It's still a currency risk: from 2002 to 2007 a rising loonie erased six years of US gains.

Source: Federal Reserve USD/CAD exchange rate; Ken French US market returns; our CAD series. 1971–2025. The 10-year currency effect is measured over windows ending each December.

6.1 What are factors?

Beyond "stocks beat cash", a few traits of companies have been linked to higher long-run returns. Fama and French call them factors.

A factor premium is the gap between two groups of stocks. The t-statistic says how unlikely the gap is to be luck: around 2 is suggestive, 3 or more is strong.

US value has averaged 4.2% a year since 1926 (t = 3.4). Emerging-market value: 7.8% (t = 4.9). Developed-market size has been about zero since 1990.

Caveat: These are paper long-short returns, before costs and taxes. You can't buy them directly; funds tilt toward them and capture only part.

Source: Ken French Data Library: US 1926 (size, value, momentum) or 1963 (profitability, investment) to Sep 2026; developed 1990+, emerging 1989–92+. Long-short premiums, arithmetic average per year.

6.2 The catch: long droughts

Every factor has gone a decade or more without paying. US value fell 58% from 2006 to 2020 and still hasn't fully recovered.

• 4.2% Average premium (per year since 1926 · t = 3.4)
• 86% Positive over 10 years (of windows (20 years: 92%))
• −58% Worst drought (2006-12 → 2020-09 · recovered: not yet)

• −58% US value fell (Dec 2006 → Sep 2020)
• −33% Still below its 2006 peak (as of Sep 2026)
• 3.2% / 7.8% Value outside the US (per year: developed (1990+) / emerging (1989+))

On paper, US small value beat the market in 96% of 20-year windows since 1926.

Real funds don't get the paper number. They pay fees and trading costs, owe taxes, and hold a slightly different set of stocks, so they have historically captured only part of the premium, with a bumpier ride than the plain market.

Caveat: Paper premiums are long-short portfolios, before any costs. Judge a factor fund against its own benchmark over 10+ years, not against these figures.

Source: Ken French Data Library, summary figures only: US long-short premiums compounded within each decade (to Sep 2026), long-run averages, hit rates and worst droughts; developed and emerging value averages; US small value vs total market from Ken French's 2×3 portfolios, 1926 onward.

6.3 How to think about tilts

Factor tilts are an optional extra on top of a diversified core, never the foundation. The price is patience through years of lagging.

For a tilt

• Value, profitability and investment premiums have t-stats near or above 3 in the US.
• They also showed up outside the US, where they were tested later (emerging-market value 7.8%/yr).
• Low-cost funds now exist in Canada.

Against, or at least caution

• Premiums vanish for 10–20 years: US value is still below its 2006 peak.
• Real funds capture only part of the paper premium, after costs.
• Careful planners shrink historical premiums before counting on them; Dimson-Marsh-Staunton find 22% of country-decades had negative premiums.
• Tracking-error regret: lagging the plain market for years makes people quit at the worst time.

US-listed factor funds from Avantis and Dimensional have short records (mostly since 2019–2021), and Canadian-listed versions launched in 2026. Five to seven years is too short to tell a premium from luck, so check each fund's own fact sheet against its benchmark over time.

A way to decide: If you tilt, size it so you could live with a decade of lagging the plain index. A tilt you abandon halfway costs more than no tilt at all.

Source: CIBC fund pages (Avantis CIBC ETFs: management fees, launched Feb–Mar 2026); Ken French Data Library (summary figures); Dimson-Marsh-Staunton.

7.1 What the Sharpe ratio says

From 50% to 100% stocks, return per unit of risk barely changed. Ending wealth nearly tripled.

• 6.4% Real return (per year after inflation)
• −12.9% Worst calendar year (in 2008)
• −26% Worst real drawdown (5.3 years longest underwater)
• 0.54 Sharpe ratio (extra return per unit of swing)
• $209k $10,000 became (after inflation, 1977–2025)

The Sharpe ratio measures extra return over cash per unit of swing. It peaked at 70% stocks, but it's nearly flat from 50% up. Surprise: all bonds was not the safest mix here. 40% stocks had the smallest real drawdown, because inflation hit bonds hard in 1977–81.

Interactive in the guide: the controls on this slide only work with JavaScript.

Source: Our CAD series, Jan 1977 – Dec 2025: all-in-one (XEQT-like) stocks (Ken French data) + Canadian bonds (Government of Canada estimate from Statistics Canada yields), rebalanced monthly; real, before fees. The growth chart shows year-end points. Crisis windows run from the month after the stock peak to the trough, nominal.

7.2 The leverage test, and the big debate

A high-Sharpe mix only beats 100% stocks if you borrow cheaply to scale it up. Most people won't, so total return is the number that matters to them.

Leverage means investing borrowed money. Only at the T-bill rate, which almost no individual gets, does the levered 70/30 edge out 100% stocks (8.5% vs 8.2%). Levered bond-heavy mixes also had deep drawdowns when inflation hit.

Total return camp

Cederburg, Anarkulova & O'Doherty: a lifetime all-stock portfolio (about 1/3 domestic, 2/3 international (2025 version; 50/50 in 2023)) gave +39% more retirement wealth than target-date funds, with a 7.0% chance of running out vs 19.7%.

Risk-adjusted camp

Cliff Asness (AQR, "Why Not 100% Equities", 2024): pick the most efficient mix, then use leverage to reach your target return. Diversifying across assets is the free lunch; concentrating in stocks wastes it.

Where that leaves a saver

If you won't borrow to invest and have decades, the highest expected total return you can sit through matters most. If you're spending from the portfolio soon, the path matters, and bonds earn their place (sequence risk, slide 2.8).

Source: Our CAD series, 1977–2025: each mix levered to the volatility of 100% stocks, borrowing at the T-bill rate plus a spread, real. Cederburg, Anarkulova & O'Doherty (SSRN 4590406, 2025 version); Asness, AQR, Feb 12 2024.

8.1 Three different questions

How much risk to take depends on three separate things. Mixing them up is the most common design mistake.

How far your portfolio can fall before you'd sell. People usually overestimate this until they live through a real crash.

Tested by: how you behaved in 2008, 2020 or 2022.

Whether a fall would actually change your life: your horizon, job stability, emergency fund, pensions, debts.

Tested by: what happens if it drops 40% the year you need it.

The return your savings rate and timeline need to reach the goal. It sets a floor, not a ceiling.

Tested by: the math of goal, savings and years.

Take the lower of tolerance and capacity. Then check it against need.

If the safe mix can't reach the goal, the fix is usually to save more, extend the timeline or adjust the goal, not to take risk you can't hold or afford.

The terms: risk tolerance, risk capacity, risk need.

Source: Standard financial-planning framework (risk tolerance, risk capacity and risk need), as explained by Ben Felix (PWL Capital) and others.

8.2 Capacity: you are a bond (or a stock)

When you're young, your biggest asset isn't your portfolio. It's your future paycheques.

• $1.36M Future earnings at 25 (3% real discount) (96% of total wealth)
• $50.7k Median financial assets, under 35 (Survey of Financial Security 2023)
• 55% Share of total wealth at 55 (human capital is shrinking)

Human capital behaves like an asset. A stable job (public sector, tenure, steady demand) pays like a bond, so someone holding it can carry more stock risk in the portfolio.

A volatile or market-linked job (commissions, start-ups, the oil patch) behaves like a stock. If it's tied to the TSX, holding mostly Canadian stocks doubles the bet.

At 35, with a bond-like job, holding 60% stocks across total wealth would mean about 671% of the portfolio in stocks. That's impossible without borrowing, but it shows why many young savers with steady jobs choose high-stock portfolios.

Caveat: Values are per person and pre-tax. The discount rate is a judgement: at 2% (bond-like job) the 25-year-old's figure is $1.61M, at 5% (stock-like) $1.02M.

Source: Statistics Canada 11-10-0239 (2024 median employment income by age), Survey of Financial Security 2023 (median financial assets, holders only, approximate). Present value of earnings to age 65, pre-tax, no real wage growth.

8.3 Capacity: the floor under you

CPP and OAS are inflation-indexed income for life, worth about $334k of bonds you already own at 65.

• $161k OAS only ($762.50/month at 65–74)
• $334k Average CPP + OAS ($19,450/yr, indexed)
• $464k Maximum CPP + OAS (CPP max $1507.65/month at 65)
• 20.9 yrs Life expectancy at 65 (15% reach 95)

A retiree with the average CPP and OAS already holds the equivalent of a large, inflation-protected bond. That raises capacity for risk in the rest of the portfolio.

Longevity risk cuts the other way: money may need to last 30 years, which is a long-horizon problem even at 65.

Other capacity checks: an emergency fund of 3–6 months; no high-interest debt; insurance; and how soon each dollar is needed.

Starting CPP at 70 instead of 65 adds 0.7% per month of delay. At this discount rate the average case is worth about $177k vs $173k.

Caveat: Per person; a couple roughly doubles these. Taxes, the OAS clawback (from $93,454 of income) and residency rules can reduce them.

Source: CPP/OAS 2026 amounts (Government of Canada; verified), valued as indexed life annuities with Statistics Canada life tables 2022/2024, discounted at the real-return bond yield of 1.88% (Bank of Canada, Oct 26, 2026). Pre-tax; OAS clawback ignored.

8.4 Tolerance: what each mix felt like

Slide through the mixes. The question isn't the average return. It's whether you'd have held on through the worst.

• 6.4% Real return per year (1977–2025)
• −12.9% Worst calendar year (2008)
• −26% Worst real drawdown (64 months to recover at worst)
• 12% Losing years (of calendar years (nominal))

$100,000 in this mix became $78.8k in 2008–09 and $85.4k in 2022.

Ask: could I have held at the bottom, and waited six years to get back? The behaviour gap says many people don't. Investors earned 1.2 points a year less than their funds (Morningstar, 2026), but only 0.7 points in all-in-one allocation funds, where the rebalancing is automatic.

Notice the bottom of the curve: 100% bonds had a deeper real fall (1977–81 inflation) than 30–40% stocks.

Interactive in the guide: the controls on this slide only work with JavaScript.

Source: All-in-one stocks (XEQT-like weights, Ken French data) + Canadian bonds, monthly rebalanced, real CAD, Jan 1977 – Dec 2025, before fees; summary figures. Canadian bonds are an estimate built from Government of Canada yields (Statistics Canada) for the whole period. Crisis windows: month after the stock peak to the trough, nominal.

8.5 Need: what to expect looking forward

Expected returns are lower than the last 50 years. Here is the range of outcomes per $10,000, after inflation.

With stocks 6.83% (bonds 3.83% + a 3.0-point premium), inflation 2%: at 5 years, 40% stocks has the lowest chance of a real loss (16%), vs 20% for all stocks and 30% for all bonds. Over 30 years the median gap widens in favour of stocks. Move the premium slider to see how much the answer depends on that one assumption.

Caveat: A smooth model understates the tails. A study of 39 developed markets from 1841 to 2019 found stocks lost purchasing power over 30 years 12% of the time (Anarkulova, Cederburg & O'Doherty, 2022; slide 4.1), not the 2% this model shows for all stocks. Use these bands for orientation, not precision.

Interactive in the guide: the controls on this slide only work with JavaScript.

Source: The deck's own assumptions: bonds = current Government of Canada bond-mix yield (Statistics Canada, Sep 2026), stocks = bonds + an adjustable premium (default 3 points), inflation = Bank of Canada 2% target. Volatility and correlation from real CAD returns 1977–2025. Lognormal model, independent years. GIC path: current 1–3 year Government of Canada yield, assumed riskless.

9.1 Accounts, 2026

The same investment can end up 30% bigger or smaller depending on the account it sits in.

TFSA and RRSP are mirror images. They give the same result when your marginal tax rate is the same going in and coming out, as long as the RRSP refund is invested too. The FHSA gets both breaks for a first home.

Caveat: Simplified: the RRSP withdrawal is taxed at one rate, and the non-registered account at the income-now rate throughout. Values use the default 3-point stock premium; the ranking of accounts doesn't depend on it. It ignores OAS clawback, pension income splitting and changes in tax law.

Source: 2026 limits: CRA / Government of Canada. Outcomes: $7,000/yr after tax for 30 years in 100% stocks at the deck's default assumption (6.83% a year before inflation = Government of Canada bond-mix yield + a 3-point premium), with an assumed dividend split (0.75% Canadian eligible dividends, 1.42% foreign dividends, the rest price growth; 10% of gains realized yearly). Combined federal + Ontario 2026 marginal rates applied throughout. Today's dollars at the Bank of Canada's 2% target.

9.2 Where to hold what

Interest is taxed hardest, so bonds usually go in registered accounts first.

• 0.17%/yr Foreign withholding tax, TFSA via Canadian ETF (≈ $21.6k over 30 years on $7k/yr)
• 0% US US-listed ETF held in an RRSP (treaty exemption for retirement accounts)

Asset location is the second-order decision after the mix itself. At ~$100k income, tax takes about 31% of a bond's return in a taxable account but about 14% of a stock's.

The trade-off: putting stocks in the TFSA also makes the tax-free account riskier. Foreign withholding tax is a small, steady cost, worth knowing but rarely worth complexity for small accounts.

Source: Combined federal + Ontario 2026 marginal rates. The deck's default assumptions: stocks 6.83% and bonds 3.83% a year before inflation. Withholding tax: assumed dividend yields of 1.2% on US stocks × 15% and 3.0% on international stocks × an assumed 10%.

9.3 Rent vs buy

Compare rent with owning's unrecoverable costs, not with a mortgage payment.

• $3,608 Owning: unrecoverable cost per month (4.77% of the price per year)
• $2,937 Renting: monthly rent (rent ÷ break-even = 0.81)

Renting is cheaper by about $671 a month, if the difference is invested.

Long run: Canadian real house prices rose 2.4% a year since 1970, are −29% below their 2022 peak after inflation, and the 1989 bust took 14.7 years to recover. Flat prices over 5 years cost −24% of the cash put in once buying and selling costs are paid.

With this deck's own assumptions, the cost of capital is about 3.8% (expected stock return 6.8% minus 3% expected house-price growth), above the rule's 3%; at that rate renting comes out cheaper in 6 of 6 cities.

A home is a levered, concentrated, single-asset bet, and also a place to live. Both rent and the unrecoverable costs of owning buy shelter; neither builds wealth on its own.

Caveat: The answer flips with the cost of capital: at 2% owning wins in all six cities, at 4% renting wins in all six. Maintenance, closing costs, condo fees and lifestyle are not in the rule. Benchmark homes vs apartment rents is an imperfect match.

Interactive in the guide: the controls on this slide only work with JavaScript.

Source: 5% rule idea: Ben Felix (PWL Capital, 13 May 2019): property tax ~1% + maintenance ~1% + cost of capital ~3%. Prices: Sep 2026 benchmarks (Toronto, Ottawa primary; Vancouver, Calgary secondary; Montreal is a single-family median; Halifax unconfirmed). Rents: Rentals.ca 2-bed asking rents, Sep 2026. City property-tax rates 2026. House prices: BIS residential property prices. Leverage example: Statistics Canada 5-year conventional mortgage rate (5.18%, 2026-09).

9.4 One-ticket portfolios

A single all-in-one ETF holds thousands of stocks and bonds worldwide, rebalances itself and costs about 0.2% a year.

• ≈0.2% All-in-one ETF (per year (MER))
• 0.5–1% Robo-advisor, all-in (fee plus ETF costs)
• 1.93% Typical Series A balanced fund (includes ~1% advisor trailer)
• 99% Canadian equity funds lagging the index (over 10 years; US equity funds 94%, international 100% (SPIVA Canada, mid-2026))

These are all-in-one ETFs: each holds a fixed stock/bond split and keeps it there. The tickers are examples of the category from three large issuers, not recommendations. Paying MER of 2% instead of 0.2% costs about 30% of ending wealth over 30 years (Act 1).

Advice can be worth paying for: planning, tax help, and someone who stops you selling in a crash. The fair comparison is advice vs no advice, not fee vs no fee.

Source: Issuer fact sheets, Sep 30 2026 (iShares, Vanguard Canada, BMO). OSC 2024 Series A mutual fund MERs. MoneySense (Feb 2026) robo-advisor costs. S&P SPIVA Canada Mid-Year 2026. Shown as examples of a category, not recommendations.

10.1 A common process, step by step

The order planners commonly follow. Most of the value comes from the first steps and the last one.

1. Emergency fund3–6 months of expenses in a HISA or cashable GIC.
2. Pay down high-interest debtA 20% card is a guaranteed 20% return when paid off.
3. List goals by dateEach goal gets its own horizon: house, kids, retirement.
4. Short money stays safeAnything needed within ~3–5 years: HISA, GICs, short bonds.
5. Check capacityHorizon, job stability, pensions, what a 40% drop would do.
6. Check toleranceHow you behaved, or would behave, in 2008 or 2022.
7. Pick the mixThe lower of the two, checked against need.
8. Choose accountsEmployer match first, then TFSA / RRSP / FHSA by your tax rate.
9. Buy broad and cheapOne all-in-one fund, or a few index funds, global.
10. Automate and rebalanceContribute on payday. Rebalance yearly or when off by 5 points.
11. Write it downA one-page investment policy statement, read before any change.

Terms: emergency fund, HISA, GIC, rebalancing, investment policy statement.

Source: Synthesis of this deck. Account order and amounts: 2026 rules (Government of Canada).

10.2 Four example people

The same framework gives different answers for different lives. These are illustrations, not recommendations.

Maya, 26

Her job and pension act like a large bond. Both ceilings are high; her tolerance is the limit. Long-term money only: the down payment she's also saving for is a separate bucket.

Daniel & Priya, 33

Two goals, two horizons, two answers. The house money can't risk a 2008; the retirement money is limited by their tolerance.

Kevin, 45

Comfortable with swings, but his income already rides on oil and the TSX. Capacity binds; holding less Canadian energy avoids a double bet.

Linda, 62

Her long-term money follows the framework; 1–2 years of spending sits in cash so a crash never forces a sale. CPP/OAS is her bond floor. Sequence risk is the main threat.

Educational illustration, not personal financial advice. Past returns don't guarantee future results.

Source: Fictional examples run through the framework explorer's rules: capacity = horizon plus adjustments, tolerance = answers to T1–T3, range = the lower ceiling minus 20 points.

10.3 Try it: the framework explorer

Pick answers for one example goal to see how a common textbook rule turns capacity and tolerance into a range of mixes, and what living with each mix has meant. A teaching tool, not a recommendation.

Capacity: can your finances take it?

Tolerance: can you sit through it?

Educational illustration, not personal financial advice. Past returns don't guarantee future results.

With these answers, tolerance (how much decline the investor would sit through) sets the limit, at 60%.

Caveat: A rule of thumb, not advice for you. Nine questions leave out most of a real situation: taxes, other savings and debts, family, health, and goals that compete for the same money. A licensed advisor looks at all of it.

• 6.4% Real return per year
• −12.9% Worst year (2008)
• −26% Worst real drawdown (64 months underwater at worst)
• 12% Losing years

• $12.7k Bad case (5th pct)
• $25.6k Median
• $51.8k Good case (95th pct)
• 1% Chance of a real loss

Forward figures use stocks 6.83% (bonds 3.83% + a 3.0-point premium), inflation 2%. The tool shows mixes of asset classes, not products.

Need check: Required real return: 2.5% per year. At the median, 60% stocks gets there (3.8%/yr real). Half of outcomes are worse, so the median alone leaves no margin.

Interactive in the guide: the controls on this slide only work with JavaScript.

Source: Rules: this deck's framework (capacity = horizon plus adjustments, tolerance = answers to T1–T3, range = the lower ceiling minus 20 points). History: all-in-one stocks (Ken French data) + Canadian bonds (Government of Canada estimate from Statistics Canada yields), real CAD, Jan 1977 – Dec 2025, before fees; summary figures. Forward: the deck's own assumptions (Government of Canada yields, Sep 2026, plus an adjustable stock premium; Bank of Canada 2% target) with 1977–2025 volatility, lognormal; per $10,000, today's dollars.

10.4 Staying the course

The plan only works if you're still following it at the bottom of the next crash.

• 9.9% Funds returned (per year, 10 years)
• 8.7% Their investors earned (buying high, selling low)
• −0.7% Gap in all-in-one funds (vs −1.2% overall)

• Selling after a fall locks in the loss. Rebalancing does the opposite: it sells some of what held up (bonds) and buys what fell (stocks).
• Re-read your plan, not the news.
• Remember the history: every drawdown in this deck was temporary for a diversified, global investor, though some took over a decade.
• Rehearse in advance. Decide now what you'll do at −20% and −40%.

• Goals, with dates and amounts.
• Target mix, and the range it came from (tolerance, capacity, need).
• Accounts and funds used.
• Contribution amount and schedule.
• Rebalancing rule: yearly, or when a part drifts 5 points.
• What you will do in a 30% fall.
• When you'll review it: once a year, or at a life change.

Terms: behaviour gap, rebalancing, investment policy statement.

Source: Morningstar Mind the Gap 2026 (US funds, 10 years to 2025). Rebalancing and IPS: standard practice.

10.5 Ten takeaways

Everything in this deck, in ten lines.

1. Real returns matter, not nominal. Inflation took 73% of a dollar's value since 1980.
2. Every asset has a failure mode. Stocks crash, bonds and cash lose slowly to inflation.
3. Your horizon decides which risk matters. Short money stays safe; long money can ride out crashes.
4. Bonds are crash insurance, not inflation insurance. They helped in 2000 and 2008, not in 2022.
5. Diversify globally. Canada is 3% of the world; countries take turns leading.
6. Factors are optional. Real premiums, with decade-long droughts.
7. Total return wins for patient savers who won't borrow. The path matters more once you're withdrawing.
8. Take the lower of tolerance and capacity. Then check it against what the goal needs.
9. Fees and taxes compound too. 2% a year takes about 30% of 30-year wealth; use the right accounts.
10. Write a plan and keep it. The behaviour gap is the most avoidable cost of all.

Educational illustration, not personal financial advice. Past returns don't guarantee future results.

Source: This deck. Full data sources, methods and licences are in the appendix.

A Glossary

Every term used in the deck, in plain language. Type to filter.

Returns and growth

Compound growth: Earning returns on past returns. The one thing the audience already knows

Nominal return: The raw % change in dollars, ignoring inflation

Inflation / CPI: The rise in prices. The Consumer Price Index measures it

Real return: Return after inflation: how much more *stuff* your money buys

Purchasing power: What your money can actually buy

Arithmetic average: Add up the yearly returns and divide. Overstates what you earn when returns swing

CAGR (compound annual growth rate): The steady yearly rate that gets you from start to finish. What you actually earned

Volatility: How much returns swing from year to year (here: the typical yearly swing, the standard deviation)

Volatility drag: Swings lower compound growth: −50% needs +100% to recover

Total return: Price change **plus** dividends or interest. Price-only charts understate it

Equity risk premium: The extra return stocks have paid over safe assets, as payment for bearing risk

Expected return: A best guess of the average future return. Not a promise. Here: today's Government of Canada bond yields plus a stock premium you can adjust

Risk

Drawdown: The fall from a previous high to a low

Recovery time / underwater: How long until you're back to the previous high

Crash risk: Losing a lot quickly, possibly when you need the money

Inflation risk: Your "safe" money silently buys less every year

Shortfall risk: Not reaching your goal because you took too *little* risk

Sequence-of-returns risk: For someone withdrawing money, bad years early do far more damage than bad years late

Safe withdrawal rate: The % of a portfolio you can take out yearly (inflation-adjusted) without running out; often quoted around 4%, recent estimates ~3–4%

Concentration risk: Too much in one stock, sector, country or house

Correlation: Whether two things tend to move together (+1), oppositely (−1) or unrelated (0)

Survivorship bias: Looking only at winners (like the US) makes the past look safer than it was

Longevity risk: Living longer than your money lasts

Currency risk: Foreign assets change value in CAD when exchange rates move

Currency hedging: Using contracts to remove currency moves (costs money; for Canadians it removed a useful cushion)

Assets and products

Stock / equity: A share of ownership in a company

Bond: A loan to a government or company that pays interest and returns the principal

T-bill: A very short-term government loan (e.g., 3 months); the classic "cash"

GIC: Guaranteed Investment Certificate: a bank deposit locked for a set term at a set rate (CDIC-insured up to limits)

HISA: High-interest savings account (or HISA ETF like CASH)

Index fund / ETF: A fund that simply holds a whole market (all its stocks) at low cost; an ETF trades like a stock

All-in-one (asset allocation) ETF: One fund holding a fixed global stock/bond mix that rebalances itself (e.g., 100/0, 80/20, 60/40)

MER: Management expense ratio: the annual fee a fund charges, as % of your money

Active fund: A fund whose manager tries to beat the market by picking stocks

Market-cap weighting: Each company held in proportion to its total market value

Home bias: Holding much more of your own country than its share of the world

Pfandbrief: A German covered (mortgage-backed) bond; part of the Germany case study

Factors

Factor: A characteristic of stocks that has been linked to different long-run returns

Market factor: Stocks minus T-bills: the basic equity premium

Size (SMB): Small companies minus big companies

Value (HML): Cheap stocks (high book-to-market) minus expensive ones

Profitability (RMW): Robust-profit firms minus weak-profit firms

Investment (CMA): Firms that invest conservatively minus those growing assets aggressively

Momentum: Recent winners minus recent losers (not in the five-factor model; shown as an aside)

t-statistic: How confident we can be a premium isn't luck; around 2+ is "probably real", 3+ is strong

Tracking-error regret: The pain of lagging the plain market for years when you tilt away from it

Portfolio design

Asset allocation / mix: How your money is split between stocks, bonds and cash

Rebalancing: Selling what has grown and buying what has shrunk, to get back to your target mix

Diversification: Spreading money so no single failure sinks you

Sharpe ratio: Extra return over cash per unit of volatility: "return per unit of risk"

Leverage: Investing borrowed money; magnifies gains and losses

Risk tolerance: How much decline you can *emotionally* sit through without selling

Risk capacity: How much decline your *finances* can absorb without derailing your goals

Risk need: How much return your goal requires

Human capital: The present value of your future earnings; your biggest asset when young

Time horizon: When you'll need the money

Emergency fund: 3–6 months of expenses in safe, accessible cash

Investment policy statement (IPS): A one-page written plan: goals, mix, accounts, rebalancing rule, what you'll do in a crash

Behaviour gap: The difference between a fund's return and what its investors actually earned (bad timing)

Canada-specific

TFSA: Tax-Free Savings Account: no deduction going in, no tax coming out

RRSP: Registered Retirement Savings Plan: tax deduction going in, taxed coming out

FHSA: First Home Savings Account: deduction in, tax-free out for a first home

RESP / CESG: Education savings plan; the government adds 20% on up to $2,500/yr per child

Non-registered account: A regular taxable investment account

Marginal tax rate: The tax rate on your *next* dollar of income

Capital gains inclusion rate: The share of a capital gain that's taxable (50%)

Dividend tax credit: Lower tax on dividends from Canadian companies

Foreign withholding tax: Tax other countries keep from dividends paid to you (US: 15%)

Asset location: Which investments go in which account to cut taxes

CPP / OAS / GIS: Canada Pension Plan, Old Age Security, Guaranteed Income Supplement: government retirement income, indexed to inflation

5% rule (rent vs buy): Owning costs ≈ 5% of the home's value per year in unrecoverable costs (property tax + maintenance + cost of capital). If rent is below that, renting is cheaper

Unrecoverable cost: Money spent that doesn't build wealth: rent, interest, property tax, maintenance, the return your down payment could have earned

Stress test: Mortgage qualification at the higher of contract rate + 2% or 5.25%

Interactive in the guide: the controls on this slide only work with JavaScript.

Source: Definitions written for this deck; the slide number shows where each term first appears.

B Methodology and data notes

How each number was built, how it was checked, and where it is an estimate.

• Canadian stocks (CAD, total return): Ken French / MSCI Canada in local currency, Jan 1977 – Dec 2025.
• US, international developed and emerging stocks: Ken French market returns in USD, converted to CAD with month-end USD/CAD from the Federal Reserve (via FRED). International uses the EAFE-weighted index before Jul 1990 and Developed ex-US after.
• All-in-one stocks: fixed weights like XEQT (US 46 / Canada 25 / international 24 / emerging 5), rebalanced monthly. World weights: MSCI ACWI IMI, Sep 2026 (US 63 / Canada 3 / international 22 / emerging 12).
• Canadian bonds: an estimate built from Government of Canada yields (Statistics Canada 10-10-0122) for the whole period, a ladder of 2, 4, 7 and 20-year par bonds weighted like the universe bond index.
• Cash: 3-month T-bill yields from Statistics Canada, from Jul 1935. Inflation: Canadian CPI from 1914.
• Mixes: all-in-one stocks plus Canadian bonds, rebalanced monthly. Real = after Canadian CPI.
• Japan after 1990: stocks from Ken French (total return in yen); bonds and cash are estimates built from the OECD 10-year government bond yield and overnight call rate; inflation from the World Bank.
• Odds by horizon: 22 developed markets from Ken French country returns with World Bank inflation (US from 1927, the others from 1975 or later); US stocks, bonds and bills from Damodaran with Shiller CPI, 1928–2025; Canada from Statistics Canada and Ken French.

• Canadian bonds are a government-only estimate: they miss the extra yield of provincial and corporate bonds in normal years and did better than a universe index in credit crises such as 2008.
• Canadian stock total returns exist here only from 1977. Earlier Canadian figures are quoted from Dimson, Marsh & Staunton (1900–2000: stocks 6.4%, bonds 1.8%, bills 1.7% a year after inflation).
• The Japanese bond and cash series are estimates built from yields, not index returns.
• Germany and Japan's war years are told with short quotations from Dimson, Marsh & Staunton (2002) and the German currency laws; no data series from those years are included.
• The 22-market sample is mostly 1975–2025, with no world wars or hyperinflations, so it understates long-run risk. The 39-market study by Anarkulova, Cederburg & O'Doherty (2022) is quoted as the long-history check.
• 1977–2025 in Canada was an era of falling interest rates, kind to bonds and lucky overall. Forward-looking numbers use the deck's own assumptions instead (below).
• Forward outcome bands use a lognormal model with independent years; it understates extreme outcomes.

• Cash: the latest 3-month Treasury bill yield, 2.39% (Statistics Canada, 2026-09).
• Bonds: the current yield of the deck's Government of Canada bond mix, 3.83% (Statistics Canada, 2026-09).
• Inflation: the Bank of Canada's 2% target.
• Stocks: bonds plus an equity premium. The default is 3 points, below the 5.1 points stocks beat bonds by in Canada (1977–2025) and 5.5 in the US (1928–2025), so the deck doesn't assume the past repeats.
• The premium slider (slides 8.5 and 10.3) changes every forward-looking number at once. Volatility and correlation come from the deck's 1977–2025 real CAD returns.
• The tax examples use an illustrative dividend split (an assumption, not data).

• Ken French, Damodaran and Shiller data appear only as yearly points in charts and as summary figures (averages, worst cases, hit rates). Their full series are not included, and those charts have no table view or download.
• Charts built only from Statistics Canada, Bank of Canada, BIS, OECD, World Bank, Federal Reserve or government data keep a Table button.
• Factor premiums are shown as decade averages and long-run summary statistics, not as time series.
• Drawdown charts on yearly points use the worst value of each year, so they still show how deep each fall went.

• Default: real (after Canadian inflation), in CAD, before fees and taxes. Nominal, price-only and estimate labels are shown where they apply.
• Per year = compound annual growth rate. Typical yearly swing = standard deviation of calendar-year returns.
• Drawdowns are real, month-end, peak to trough; recovered = back to the prior real peak. Famous crashes use month-end Shiller and Ken French data.
• Odds by horizon use overlapping windows, so the windows are not independent.

• During development the rebuilt stock series were checked internally against ETF records over the years they overlap. Those price records are not included or redistributed here.
• Statistics Canada marks missing values as −99.999; nine early T-bill months (1934–35) were caught and excluded, so cash history starts Jul 1935.
• Fourteen key facts (account limits, CPP/OAS, SPIVA, Morningstar, Cederburg et al., housing and others) were re-checked against primary sources; 13 confirmed, 1 corrected (Avantis figures are management fees, not MERs).
• Account rules are 2026 values; the Bank of Canada policy rate (2.25% on Oct 26, 2026) had a decision due Oct 28, 2026.

Source: Built with this deck's own data and analysis scripts from the sources on the next slide. Market data to December 2025 unless stated.

C Sources and licences

Where every number comes from, how to cite it, and the terms it is shared under.

Educational illustration, not personal financial advice. Past returns don't guarantee future results.

Source: Licence notes summarise each provider's published terms as read on Oct 27, 2026. They are not legal advice; check each provider's current terms before reuse.