Bivar Canada MomentumSystematic momentum with daily regime filter, TSX (Toronto Stock Exchange)

A long-only momentum strategy on Canadian equities (TSX) with equal weighting, market regime overlay (cash when TSX below MA75 days), top 80% by market cap (computed independently of the fundamental filter below), and requires NetIncome(TTM) > 0. 319 months of backtesting. Transaction costs of 40bps RT included. No look-ahead bias. Data: EODHD adjusted prices.

Performance Summary

Sharpe and Sortino above deduct a 3%/yr risk-free rate.

II. Market Regime Filter

The TSX Composite (^GSPTSE) is compared to its 75-day moving average daily. When below MA75, portfolio moves to 100% cash. Red shading = defensive periods.

III. Equity Curve

Log-scale equity curve. Both strategy and benchmark start at 100. Strategy includes 40bps round-trip transaction costs applied monthly on actual turnover. Dates are the month the return was earned (realization month), not the month the signal was computed.

IV. Drawdown

V. Monthly Regime

Each cell shows whether the strategy was invested (green) or in cash (red) during that month.

VI. Monthly Returns

VII. Annual Returns

VIII. Rolling 3-Year CAGR

Every point shows the annualised return of the prior 36 months. Green shading = strategy beats TSX, red = underperforms.

IX. Crisis Analysis

Strategy performance during major market downturns. The MA75 regime filter protects capital during sustained downtrends.

XV. Risks & Caveats

• Transaction costs. 40bps round-trip on actual monthly turnover, based on real IBKR median bid/ask spreads measured across this strategy's historical holdings. Costs on regime transitions (moving fully in or out of the market) are approximated from the same turnover-based model rather than measured separately, which likely understates the true cost of those transitions by a small margin.
• Parameter selection. MA75, N=10, market-cap floor (top80%), and 12M (no skip) momentum were chosen via sensitivity analysis on the same dataset used to report performance here — no out-of-sample holdout. The companion parameter-robustness work documents how performance behaves across a range of reasonable values for each choice, not just the final one.
• Benchmark is TSX Composite (^GSPTSE), price return only (no dividends). Strategy uses adjusted prices including dividends. Table columns above are labeled accordingly (TSX Composite Price Return / Relative Performance vs TSX Price Return) rather than "Alpha," since the strategy and benchmark are not on a like-for-like total-return basis.
• Backtested performance is not a forecast. Live results may differ from the backtest due to parameter uncertainty, implementation costs, liquidity constraints, and model degradation over time.
• Resource concentration. TSX is heavily weighted in energy and materials. Top 10 momentum stocks may be sector-concentrated in commodity cycles. Performance is not driven by a handful of outlier months: the ten best months (out of 319) are spread across 2003–2023, not concentrated in a single regime.
• Data source: EODHD adjusted prices, market cap (shares × price), and quarterly financials (NetIncome, filing-date-gated to avoid look-ahead). Adjusted for splits, dividends, and corporate actions.

XVI. Fundamental Filter

XVII. Complete Position History

Every month since January 2000. "Date" is the month the return was earned; "Signal Date" is the prior month-end the holdings were actually selected from (using only data available through that date).