1. The Core Distinction: Relative Strength vs. Raw Momentum
To build a robust stock selection framework, you must first separate the signal from the noise. Raw momentum is a price-based metric that measures an asset’s own velocity and acceleration—typically calculated as a percentage change over a lookback period (e.g., 3, 6, or 12 months). It answers the question: “How fast is this stock moving?” Relative Strength (RS), in the context of stock picking, is not the same as RSI (Relative Strength Index). Instead, RS is a comparative measure: it divides a stock’s price performance by a benchmark’s performance (e.g., the S&P 500 or a sector ETF) over the same period. It answers: “Is this stock moving faster or slower than everything else?”
The critical flaw in using raw momentum alone is that it ignores the market tide. During a broad bull market, a stock can have a 20% trailing return yet still be a laggard, gaining only half of what the index gained. Conversely, in a correction, a stock may have a negative absolute return but a high RS because it fell 5% while the market fell 15%. The former is a weak selection; the latter is a sign of institutional sponsorship. Combining the two creates a dual-filter system. Raw momentum ensures you are trading in the direction of positive absolute price action (avoiding falling knives), while relative strength ensures you are selecting the leaders of the current market cycle, not merely the participants.
2. Why the Combination Works: The Behavioral Finance Rationale
The efficacy of this hybrid approach is rooted in two well-documented behavioral biases: anchoring and herding. When a stock exhibits high raw momentum, investors anchor to its recent price history, expecting the trend to continue, which drives further buying. However, this effect is amplified when combined with high RS. Stocks with high RS are typically receiving disproportionate capital inflows from institutional funds. Fund managers, benchmarked against indices, cannot afford to hold a stock that is outperforming their benchmark by a wide margin—they face career risk by being underweight. This forces a herding effect, pushing the RS leader higher.
Moreover, the combination filters out “vicious” momentum traps. A stock can have high raw momentum due to a short squeeze, a one-time news event, or low liquidity. However, high RS sustained over a long lookback (e.g., 52 weeks) indicates a structural shift in business fundamentals relative to peers. When you require both high absolute returns (momentum) and high benchmark-relative returns (RS), you are effectively selecting securities that have both positive earnings revisions and are accruing a scarcity premium. This confluence reduces the probability of owning a stock that is merely bouncing off a cyclical low.
3. The Mathematical Framework: Calculating Dual-Factor Scores
To implement this systematically, you must standardize the data to avoid scale mismatches. Raw momentum is often expressed as a percentage, while RS can be a ratio (e.g., 1.15) or a percentile rank. The most robust method is to use a cross-sectional Z-score for both metrics.
- Step 1: Momentum Score (M-Score). Calculate the 126-trading-day (6-month) price return, excluding the most recent month (skip the last 20 trading days) to avoid the short-term reversal effect. The formula:
M = (P_t / P_{t-126}) - 1. - Step 2: Relative Strength Score (RS-Score). Calculate the stock’s return over the same 126 days divided by the benchmark’s return over the same period. Alternatively, calculate the stock’s return minus the benchmark’s return. For a more robust metric, use a regression-based Beta-adjusted alpha, but a simple geometric excess return (Stock Return – Benchmark Return) is highly effective.
- Step 3: Standardization. For each stock in your universe, compute the Z-score for both M and RS-Score:
Z = (X - μ) / σ, where μ is the cross-sectional average and σ is the standard deviation. - Step 4: Composite Score. Combine the two Z-scores:
Composite = (0.5 * Z_M) + (0.5 * Z_RS). Many practitioners overweight RS (e.g., 0.6/0.4) because RS is a more stable predictor in trending markets. Sort the universe by this composite score and select the top decile.
The critical technicality is the lookback period alignment. Using a 12-month lookback for momentum (ignoring the last month) captures the “time-series” effect. For RS, using a 6-month lookback for the excess return calculation captures the acceleration of relative outperformance. Mixing different lookbacks—e.g., 12-month momentum with a 1-month RS—creates noise. Standardize your lookbacks: use a 6-month (126-day) acceleration period for both raw price change and excess benchmark return.
4. Sector-Relative vs. Market-Relative RS: The Hidden Layer
A common pitfall is comparing a stock’s RS against a broad index like the S&P 500. This creates a systemic bias toward cyclical sectors (Technology, Consumer Discretionary) during bull markets and defensives (Utilities, Healthcare) during bear markets. To refine the selection, you must layer two RS calculations.
- Level 1: Market RS (Beta-adjusted). This is the standard RS against the S&P 500 or a global equity benchmark. It tells you if the stock is a market leader.
- Level 2: Sector RS. This measures the stock’s performance relative to its own Industry Classification Benchmark (ICB) or GICS sector ETF. For example, if a semiconductor stock is down 5% over the quarter, but the SOX Semiconductor Index is down 15%, the stock has positive internal sector RS. This indicates it is the dominant player in a struggling space.
When combining momentum for stock picks, prioritize the sector-relative RS for the primary entry trigger, and use market RS as a tie-breaker. Why? A stock can have top-tier market RS simply because its sector is exploding (e.g., AI stocks in 2023). By requiring internal sector RS to be high, you filter out the “rising tide” effect. You want the stock that is leading its own group, not just riding a cyclical wave. A high composite score (momentum + sector RS) that also shows a market RS above 1.0 is the highest-conviction setup.
5. The Optimal Holding Period and Rebalancing Frequency
The combination strategy demands specific rebalancing cadence. Research (Jegadeesh & Titman, 1993; Moskowitz & Grinblatt, 1999) suggests that momentum profits are concentrated in the first 3–6 months. However, RS profiles are stickier. A stock that outperforms its benchmark for one quarter is likely to continue for the next quarter due to institutional flow inertia.
A practical framework involves a dual-frequency rebalance.
- Monthly Review (Ranking): Every 20 trading days, recalculate the M-Score and RS-Score for your entire universe.
- Quarterly Rebalance (Holdings): Do not immediately sell a stock just because its daily momentum dips, but do hard-sell if its RS drops below 1.0 (i.e., the stock begins to underperform the benchmark on a trailing 60-day basis).
- The 13-Week Rule: Do not buy a stock unless it has made a 13-week (quarterly) high. This aligns with the raw momentum filter. For the RS filter, ensure the stock’s 63-day (quarterly) excess return over the benchmark is positive.
This creates a self-cleaning loop. Raw momentum (price) must be in a confirmed uptrend (13-week high). RS must be confirmed by relative price strength (excess return). Rebalancing too frequently (weekly) increases transaction costs and whipsaws from volatility. Rebalancing too rarely (annually) misses the decay point where RS reverses.
6. Integration with Technical Volume Confirmation
Price-based RS and momentum are lagging indicators—they tell you what has happened. To ensure the future trajectory of the trend, you must add a volume confirmation ratio. High RS with declining volume suggests a lack of institutional participation, making the move prone to failure.
Integrate the Up/Down Volume Ratio (or On-Balance-Volume trend) relative to the benchmark. A stock with high composite momentum and strong RS must exhibit an Accumulation/Distribution Line that is rising at a rate commensurate with its price rise. If prices are making new highs (momentum) and the stock is outperforming the index (RS), but volume is contracting on up days and expanding on down days, the RS is likely to fade.
Specifically, calculate the Price Volume Trend (PVT) for both the stock and the benchmark. Calculate the differential: PVT_Stock - PVT_Benchmark. If this differential is positive and rising over the last 63 days, volume is confirming the RS move. This acts as a final compliance filter: do not select a stock for your high-momentum list that scores in the top decile for RS but in the bottom decile for volume differential. Volume provides the fuel; momentum is the engine; RS is the heading. You need all three aligned.
7. Risk Mitigation: The Volatility-Adjusted RS
One of the primary criticisms of relative strength is that it is a measure of return without regard to the risk taken to achieve that return. A stock with a beta of 2.0 that returns 20% while the market returns 10% appears to have an RS of 2.0. However, if the standard deviation of its weekly returns is 10%, the risk-adjusted excess return is poor.
To improve the quality of the stock picks, apply a volatility penalty to the RS score. Calculate the Information Ratio (IR):
IR = (Annualized Stock Return - Annualized Benchmark Return) / (Annualized Standard Deviation of the Excess Return)
A high IR indicates that the stock is outperforming on a consistent basis with low volatility of excess returns. Using the IR as the RS component of the composite score eliminates high-beta lotteries. For example, a biotech stock that triples on a drug approval but has huge weekly swings will have a low IR, disqualifying it despite high absolute momentum. Conversely, a consumer staples stock with steady 1% weekly gains against a flat market will have a high IR, making it a superior risk-adjusted pick.
You can implement this by dividing your excess return by the standard deviation of weekly returns over the same 126-day period. This not only selects for direction (RS) but also for the quality of the trend. This systematic smoothing helps you stay on the “right side” of the trade longer.
8. The Macro Overlay: When to Disable the Composite
No momentum/RS algorithm is infallible; they fail violently during regime shifts (e.g., 2008, March 2020). A crucial aspect of a high-quality article on this topic is knowing when the signal breaks down. The combination of relative strength and momentum assumes a degree of trend persistence. During a sharp V-shaped recovery, momentum is a disaster because the fastest-falling stocks (best RS leaders) are typically cyclical laggards being bought.
Implement a macro volatility filter. If the 20-day realized volatility of the S&P 500 exceeds a threshold (e.g., the 90th percentile of its trailing 1-year range) and the 20-day EMA of the index is below the 100-day EMA, the reliability of the composite score falls below 45%. In this regime, your composite should be inverted for tactical hedging or you should hold cash.
Additionally, apply the 12/1 Momentum Rule to the benchmark itself. Only be fully invested in RS/Momentum stocks when the benchmark index is above its 10-month simple moving average. If the benchmark is below this level, reduce position sizing by 50% and require a much higher RS threshold (e.g., top 2% percentile) before initiating longs. This protects you against “catch the falling knife” scenarios where a stock has high RS but the overall tape is unwinding violently.
9. Practical Implementation: A Step-by-Step Screening Protocol
To take this from theory to practice, your screener logic must be explicitly coded with specific thresholds. Here is a high-accuracy protocol:
Step A: Base Liquidity Filter. Screen for stocks with a minimum Average Dollar Volume of $50 million over the last 3 months to eliminate micro-caps manipulable by small flows.
Step B: Price Floor. Exclude stocks trading under $10. Penny stocks with high RS are statistically unreliable due to fundamental insolvency risks overriding technicals.
Step C: Raw Momentum Acceleration (Time-Series). Select stocks where the 120-day rate of change (excluding the latest 10 days) is positive AND the 120-day rate of change is greater than the 240-day rate of change. This confirms acceleration (raw momentum is increasing).
Step D: RS-Decile Ranking (Cross-Sectional). Rank the remaining stocks by their 126-day Excess Return (Stock Return – SPY Return) divided by the standard deviation of that excess (Information Ratio). Keep only the top 20% of this ranking.
Step E: Sector Leadership Check. Ensure the stock’s 63-day return is greater than the median 63-day return of its specific sub-industry grouping. This avoids picking the laggard in a hot sector.
Step F: Scoring Output. Assign points: 1 point for raw momentum acceleration, 1 point for top-quartile IR, 1 point for being a sector leader. Final Picks = All stocks with 3 points. This rank-ordering ensures you aren’t just picking a high-number composite but a confluence of favorable conditions.
10. Common Pitfalls and How to Avoid Look-Ahead Bias
When combining RS and momentum, the most costly error is look-ahead bias in ranking. If you calculate momentum using closing prices, you must execute your trades at those exact closing prices (or the next open immediately following). Many practitioners inadvertently calculate rankings using day ( t ) closing data but enter trades at day ( t+2 ) average prices, which destroys the edge.
To avoid this, use a strict T+1 execution rule.
- Ranking Day (D): Use closing data to generate the composite score.
- Execution Day (D+1): Enter limit orders at the open or during the first hour only. All fill prices must be captured for backtesting purity.
Additionally, avoid survivorship bias. When backtesting this dual strategy, ensure you include delisted stocks. Stocks with high momentum/RS often get acquired (disappearing from databases) or drop off the exchange due to insolvency. If you only use the current list of S&P 500 components, you will overestimate returns by 3–5% annually.
11. The Role of Earnings and Fundamental Confirmation
While the strategy is strictly price-based, a quality overlay involves fundamental confirmation to avoid earnings bomb traps. High RS and momentum can be inaccurately high for a 2-month period preceding a catastrophic earnings disappointment. To filter this, calculate the Estimate Revision Ratio: the count of upward EPS estimate revisions minus downward revisions over the last 60 days, divided by the total revisions.
Do not buy a stock with a high composite momentum/RS score if the Revision Ratio is negative. The technicals are often leading a fundamental deterioration, but the risk is asymmetric. If the stock has top RS and momentum but declining analyst estimates, the volatility is too high for a stable equity curve. Sell or skip.
When the RS/Momentum stock also demonstrates a positive Revision Ratio, you are buying the “price-leading-earnings” dynamic. This confirms that the price action is driven by institutional buyers privy to upcoming fundamental beats, providing a sustainable tailwind for the next 3–6 months—the optimal holding period for this systematic combination.
