FAQ
What is True North Alpha
True North Alpha is a systematic Canadian equity strategy focused on identifying large-cap TSX-listed companies trading at a significant premium to their historical valuation levels—when that premium is supported by strong fundamentals. The model is rules-based, fully transparent, and rebalanced annually, with no discretionary input, macro forecasting, or narrative overlays.
Why do high-multiple stocks continue to outperform?
Valuation expansion is often misunderstood. When supported by fundamentals, elevated multiples can persist for extended periods. Empirical research supports the premise that:
Stocks undergoing multiples expansion, driven by higher earnings expectations, tend to outperform—both in absolute terms and relative to peers.
This aligns with documented effects in academic literature on earnings momentum, price momentum, and valuation persistence.
It is also premised on the concept of "valuation re-rating". This happens when the market starts to assign a higher value to a company—not just because earnings went up, but because investors are willing to pay more for each dollar of those earnings. In other words, the company’s valuation multiples (like P/E ratio or price-to-sales) increase.
This usually happens when the market realizes a company has stronger, more durable growth than previously thought—maybe it’s expanding faster, more profitable, or less risky than expected.
Is the model optimized using hindsight?
No.
The model architecture—composite scoring, factor weightings, selection rules—remains consistent across all years. Backtests use only information available at each year-end (e.g., December 31 data for January portfolio construction) in a given year. There is no look-ahead bias, no annual tweaking, and no discretionary overrides.
Does the backtest adjust for survivorship bias?
Yes.
Constituents are adjusted retroactively to reflect the actual TSX composition as of each year-end. Historical selections include delisted or acquired companies where applicable, and all returns are calculated on a total return basis (price + dividends).
Does the model outperform in all years?
No investment strategy outperforms in every period. What matters is persistence and risk-adjusted return over time.
- Outperformance in 19 of the past 24 years
- Average excess return in outperforming years: +7.4%
- Average shortfall in underperforming years: –3.1%
That asymmetry—limited downside in weaker years, material upside in strong years—creates a positive skew that compounds over time. For example, $10,000 invested in the model in 2000 would have grown to over $158,000, versus $53,000 in the TSX benchmark.
How does the model perform in volatile or adverse markets?
The strategy has demonstrated resilience across a range of market conditions, including:
- The Dot-Com correction (2001–2002)
- The Global Financial Crisis (2008)
- COVID-19 volatility (2020–2021)
Despite not being a “defensive” strategy, downside capture is lower than expected—driven by the model’s preference for high-quality, resilient businesses.
Additional details can be found on our Results page.
Is the performance just a function of tech exposure?
No.
While the model has periodically included technology companies, its sector allocations vary across financials, consumer goods, industrials, and utilities. Stock selection is driven by valuation relative to each firm’s historical range, not sector tilts or thematic bets.
Is the outperformance front-loaded in the early 2000s?
No.
Recent performance remains strong and consistent:
- 5-Year Annualized Return (2020–2024): 16.1% vs. TSX 11.1%
- 10-Year Annualized Return (2015–2024): 11.6% vs. TSX 8.7%
This is not a backtest anchored in one cycle. It held up across rate regimes, commodity cycles, and macro environments—because it’s based on firm-specific fundamentals.
Is the backtest fully transparent?
Yes.
All annual selections and returns are published on the Results page. While the model inputs and weightings are proprietary, the outcomes are disclosed in full—so you can evaluate them without opacity or selective reporting.
How strong are the risk-adjusted metrics?
- Sharpe Ratio: 0.69 (vs. TSX at 0.33)
- Information Ratio: 0.80
These figures reflect not just outperformance, but efficient outperformance. By comparison, most active mutual funds struggle to maintain an information ratio above 0.5.
Is this investment advice?
No.
True North Alpha is a model portfolio strategy intended for educational purposes. It is not personalized advice and does not consider your individual financial objectives, risk tolerance, or constraints. Past performance is not indicative of future returns. Please consult a licensed financial advisor before acting on any investment strategy.
Have you tested the strategy in other markets or universes? How portable is the signal outside of Canada?
Yes. We’ve tested the strategy on U.S. universes, including the S&P 500 and Dow Jones Industrial Average. The results were materially weaker than in Canada.
We believe this reflects structural differences. The TSX has many companies in cyclical sectors (like energy, mining, financials), where valuation changes happen quickly when sentiment or fundamentals shift. It is a more cyclical, less crowded market, with valuation anomalies that tend to persist longer.
In contrast, U.S. indices like the S&P 500 and DJIA are heavily institutionally covered, more efficient, and quicker to price in re-ratings. The model's empirical strength is far more pronounced in the Canadian context, where market inefficiencies remain more exploitable.
Why does backtesting matter? Isn’t it just looking at the past?
Backtesting is how you test whether a strategy actually works—not just in theory, but across real market history.
But it only matters if it’s done right. At True North Alpha, the backtest uses only the data that would’ve been available at the time—no peeking ahead, no hindsight tweaks. And we include all the companies that were actually in the index, even ones that were later delisted or acquired. That avoids what’s called survivorship bias.
The result? A track record that’s consistent, repeatable, and rooted in how the model would have performed in real time.