CGI EMA Crossover Strategy

That's the good news - in Canada there is no fixed F&O lot for cash equities. You buy exactly the number of shares your risk math calls for. With C$100,000 capital, a 2% risk budget (C$2,000), and a stop C$4 away, you'd buy 500 shares - no rounding to a chunky lot. The standard board lot is 100 shares, but odd lots are tradeable too. If you use Montreal Exchange options or share futures instead, those are 100 shares per contract. This granularity makes precise position sizing much easier than with Wipro's 1600-share lot.

Yes, but with adjustments. Hourly or 15-minute charts generate more signals but also more whipsaws. For shorter timeframes, consider faster EMAs (5/13 instead of 9/21), tighter stops, and smaller position sizes. The daily timeframe is recommended for beginners as it's less noisy and requires less active monitoring. Note also that CGI's single-stock liquidity is thinner intraday than US large-caps, so very short timeframes can suffer from wider spreads.

If the 9 EMA crosses back below the 21 EMA the day after your entry (a quick reversal), your stop loss should protect you from large losses. This is a whipsaw. Don't revenge trade - accept the small loss and wait for the next clean signal. Consider adding an ADX filter to avoid signals during ranging conditions.

For most traders, the TSX cash/share market is the right answer - it's the most liquid venue for CGI and lets you size precisely. CGI options exist on the Montreal Exchange (100 shares/contract) and give defined risk, but they trade with wider spreads and fewer strikes than US options, so check the bid-ask first. Share (single-stock) futures are listed but typically too illiquid to be practical. If you want leveraged or sector exposure, the XIT ETF is far more liquid than CGI single-stock derivatives. Match the instrument to your capital, your risk tolerance, and - importantly in Canada - to the liquidity actually available.

EMA crossover trades are swing trades typically lasting 5-15 days. Hold until: (1) an opposite crossover occurs (the 9 EMA crosses back), (2) the stop loss is hit, (3) the target is reached, or (4) 10+ days pass without meaningful progress (time stop). Don't convert swing trades into investments - if the signal fails, exit and wait for the next opportunity.

Three filters help: (1) an ADX filter - only trade when ADX > 22, indicating trending conditions; (2) an EMA separation rule - wait for 0.5-1% separation between the 9 and 21 EMA before acting; and (3) the price-EMA relationship - avoid trading when price is oscillating around the 50 EMA. If unsure, reduce position size rather than skipping entirely. CGI's calmer, lower-beta profile means its trends are often cleaner, but tight ranges still produce whipsaws.

Use options when: (1) capital is limited (an option costs less than the full share position), (2) you want defined maximum risk, or (3) you're uncertain about signal quality and want to cap downside. But in Canada, first check CGI option liquidity - if the bid-ask spread is wide or strikes are sparse, the spread you pay can erase the strategy's edge, and the cash market or XIT is better. Use shares when the trend is clear, you want full participation without premium decay, and you're comfortable with the (smaller) capital outlay relative to a leveraged future.

Check the S&P/TSX IT index (XIT) trend before trading CGI - crossovers aligned with the sector trend have higher probability, though remember the index is driven by higher-beta names like Shopify. Compare CGI's relative strength versus the index - outperformance adds confidence. Watch Open Text (the closest business-model peer) for confirmation, and Constellation/Shopify for overall sector tone. A genuine IT-services signal (CGI plus Open Text moving together) is more reliable than CGI moving alone on an M&A headline.

Key filters: RSI(14) > 50 for longs, < 50 for shorts. MACD histogram positive/negative aligning with the crossover direction. ROC(10) > 0 for longs. Volume > 1.3x average on the crossover day. Create a checklist scoring 0-4 based on how many filters pass. A score of 3-4 = high confidence; 0-2 = low confidence, skip or reduce size.

Avoid new positions 5-7 days before CGI earnings (September fiscal year, so quarters report roughly late January, late April/early May, late July, and early November) due to elevated uncertainty and binary risk. After earnings, wait for the dust to settle (1-2 days), then fresh crossover signals can carry extra momentum from the earnings catalyst. Crossovers immediately after a positive report often have strong follow-through as institutions adjust positions. Watch for M&A and large bookings announcements too - CGI is highly acquisitive and deals can move the stock outside earnings windows.

Use walk-forward analysis rather than simple in-sample optimization. Test multiple periods (8/18, 9/21, 10/26, 12/26) on 3-year training windows, validate on 1-year forward periods, roll and repeat. Focus on Sharpe ratio and max drawdown, not just returns. Parameters consistent across periods are robust; wildly varying parameters indicate curve-fitting. Be especially careful with CGI - a steadier, lower-turnover name produces fewer crossover samples, so don't over-tune on a small dataset.

Train classification models (Random Forest, XGBoost) on features like crossover angle, EMA separation rate, volume ratio, ADX, RSI, and sector momentum. Target variable: crossover success (hit target before stop). The output probability allows position sizing - higher probability = larger position. Validate rigorously to avoid overfitting; monitor and retrain periodically. With a thinly traded single name, guard hard against small-sample overfitting and keep human oversight on every recommendation.

Limit total IT-sector exposure to 25-30% regardless of signals - this matters in Canada because the listed tech universe is small and correlated. Implement strategy risk budgeting - allocate, say, 5% of portfolio risk to the CGI system and pause when the budget is consumed. Use the Kelly Criterion (practical half or quarter Kelly) for position sizing. Diversify across uncorrelated strategies (e.g., Canadian banks or energy mean reversion) to smooth the equity curve.

Track rolling win rate vs backtest, rolling profit factor, drawdown vs maximum expected, execution slippage, and signal frequency. Red flags: >5 consecutive losses, drawdown exceeding 10%, significant deviation from backtest metrics, or unusual execution issues (which, for CGI, often means thin-liquidity slippage). Conduct a monthly review with detailed analysis and quarterly recalibration if needed.

Calculate a regime indicator (ADX or a volatility measure). In a trending regime (ADX > 25), use faster EMAs (8/18) to capture moves quickly. In a ranging regime (ADX < 20), either use slower EMAs (12/30) to filter noise or avoid trading entirely. Backtest the regime switching to ensure it improves on static parameters before implementing - and confirm the improvement survives walk-forward testing on CGI's limited sample.