Third-Party Research & Methodology Only

This section shares summaries of third-party academic research and descriptions of quantitative models. The content represents the findings of the original researchers, not the opinions or recommendations of Foxholm Financial. Foxholm Financial does not publish hypothetical or backtested performance metrics on its quantitative research pages. All content is restricted to methodology, signal construction, factor logic, and risk architecture. SEC rules require that investment advisers not present misleading performance data, and our methodology-only approach reflects that standard and the firm's fiduciary obligations.

Momentum Scoring Model

Trend-Based Trading Strategy Signal Generation

Robert Stowe, AAMS® Investment Advisor

A momentum scoring model ranks stocks by the strength of their recent price trends, then uses that ranking to identify the strongest candidates for buying and the weakest candidates for selling or avoiding. The underlying principle is simple: stocks that have been going up tend to keep going up, and stocks that have been going down tend to keep going down, at least over the medium term.

This momentum effect is one of the most studied and most persistent patterns in financial markets. Jegadeesh and Titman documented it in their landmark 1993 paper, showing that buying recent winners and selling recent losers produced consistent profits over 3-to-12-month horizons. The effect has been found across stocks, bonds, currencies, and commodities, and across markets in dozens of countries over more than a century of data.

Conceptual Framework

Momentum challenges the idea that markets are perfectly efficient. If all available information were instantly reflected in prices, there would be no reason for recent winners to keep winning. Two families of explanation have emerged for why the effect persists:

Behavioral explanations: Investors react to new information slowly (underreaction), causing prices to adjust gradually rather than instantly. Confirmation bias leads investors to seek information that supports their existing views, delaying full price adjustment. Herding behavior creates self-reinforcing trends as investors follow each other.
Risk-based explanations: Momentum stocks may be riskier in ways that the standard risk models do not capture. The profits from momentum could be compensation for bearing this hidden risk rather than a genuine market inefficiency. Momentum strategies are vulnerable to sharp, sudden reversals (momentum crashes), and the risk of these crashes may justify the average return.

Core Assumptions

Momentum scoring models make several assumptions about how price trends behave. Each introduces risk when reality diverges:

Trends persist over the medium term: The model assumes that price momentum measured over the past 3 to 12 months predicts positive continuation over the next 1 to 6 months. This is well-supported by academic evidence, but the strength of the effect varies across time periods and market environments.
Past returns contain information: The model treats recent price performance as a signal about future returns. This is the opposite of the efficient market assumption that past prices contain no useful information. The momentum effect suggests that prices incorporate information gradually rather than instantly.
The ranking is meaningful: Momentum scoring assumes that the relative ranking of stocks by momentum is informative: a stock in the top 10% of momentum has a better outlook than one in the top 30%. The model does not predict absolute returns, only relative ones. It works as a sorting mechanism rather than a precise forecast.
Transaction costs do not consume the premium: Momentum strategies require regular rebalancing (typically monthly) as rankings change. The stocks with the strongest momentum are often mid-cap or smaller stocks with wider bid-ask spreads, which increases trading costs. The strategy is only profitable if the momentum premium exceeds the cost of maintaining the portfolio.

Signal Construction

A momentum scoring model follows a structured pipeline from defining the universe of stocks to generating a ranked signal.

Step 1

Define Universe

Step 2

Calculate Momentum

Step 3

Apply Quality Filters

Step 4

Rank & Score

Step 5

Portfolio Construction

Universe Definition

The first step is defining which stocks are eligible. A typical momentum universe includes liquid stocks above a minimum market capitalization (often $500 million or $1 billion) to ensure that the positions can be traded without excessive market impact. Illiquid stocks, recent IPOs, and stocks with insufficient price history are typically excluded.

Momentum Calculation

The core calculation is straightforward: the total return of each stock over a lookback period, typically 12 months with the most recent month excluded. The most recent month is skipped because very short-term returns (the past few weeks) tend to reverse rather than continue, a phenomenon called short-term reversal. Including the most recent month would contaminate the momentum signal with this opposing effect.

Many implementations go beyond simple price return and incorporate additional momentum dimensions:

Risk-adjusted momentum: Dividing the price return by the stock's volatility over the same period. This gives credit to stocks that achieved their returns with less price variation, producing a smoother signal.
Earnings momentum: Measuring the direction and magnitude of recent earnings surprises and analyst estimate revisions. Stocks where earnings are beating expectations often show stronger and more persistent price momentum.
Multi-timeframe momentum: Combining momentum measured at different horizons (e.g., 1-month, 3-month, 6-month, 12-month) into a single composite score. This captures both fast-moving and slow-moving trends.

Quality Filters

Raw momentum signals can be misleading. A stock that doubled because of a speculative bubble has high momentum but poor prospects. Quality filters screen out stocks whose momentum is likely driven by unsustainable factors. Common filters include minimum earnings quality (positive operating cash flow), reasonable valuation (not extreme price-to-earnings ratios), and adequate liquidity (minimum average daily trading volume).

Ranking and Scoring

After calculating the momentum measure and applying quality filters, stocks are ranked from strongest to weakest. The ranking is typically converted to a percentile score (0 to 100) or a z-score to normalize across different time periods. A composite score can combine multiple momentum measures (price momentum, earnings momentum, risk-adjusted momentum) using fixed weights or a statistical weighting scheme.

Risk Architecture

Momentum strategies have well-documented risk properties that distinguish them from other systematic approaches. Understanding these risks is essential because momentum can produce extended periods of strong gains followed by sudden, severe losses.

Model Risk

The most dangerous risk in momentum investing is the momentum crash: a sudden, sharp reversal where recent losers surge and recent winners plunge. The most notable example occurred in March 2009, when momentum portfolios lost over 40% in a single month as beaten-down financial stocks rallied violently. These crashes tend to happen at market turning points, when the economy shifts from recession to recovery and the previous winners (defensive stocks) suddenly underperform the previous losers (cyclical stocks).

Momentum crashes are infrequent but severe enough to wipe out years of accumulated gains. The risk is particularly dangerous because it is concentrated in time: it occurs precisely when investors are least prepared, at the transition between market regimes. Standard risk models based on normal market conditions underestimate the severity of these events.

Known Limitations

Limitations to Consider

Momentum crashes: Sudden reversals can produce losses of 20-40% or more in a single month. These events are rare but devastating and tend to occur at market turning points.
High turnover: Momentum rankings change frequently, requiring regular portfolio rebalancing. Monthly turnover rates of 50-100% are common, generating significant transaction costs and potential tax consequences.
Capacity constraints: The strategy works best with smaller and mid-cap stocks, where price trends are stronger but liquidity is lower. As the amount of capital following momentum strategies grows, the trading impact can erode the premium.
Regime sensitivity: Momentum performs well in trending markets but struggles during choppy, range-bound periods when trends form and reverse quickly. Extended periods of sideways markets can produce a series of small losses.
Crowding risk: Because momentum is a well-known factor, many institutional investors and ETFs follow similar signals. When many participants hold similar positions and try to exit simultaneously, the selling pressure can amplify losses.

Practical Considerations

Choosing the Lookback Period

The standard academic specification uses 12 months of returns minus the most recent month (often written as "12-1" momentum). However, alternative lookback periods capture different aspects of momentum. A 6-month lookback responds faster to trend changes but is noisier. A 3-month lookback captures intermediate momentum but may overlap with short-term reversal effects. Many practitioners use a composite of multiple lookback periods to produce a more robust signal.

Rebalancing Frequency

Momentum portfolios are typically rebalanced monthly, which balances signal freshness against transaction costs. More frequent rebalancing (weekly) captures trend changes faster but increases costs. Less frequent rebalancing (quarterly) reduces costs but allows the portfolio to drift as momentum rankings change. A common middle ground is monthly rebalancing with a "buffer zone": holdings are only replaced when their momentum rank drops below a threshold, reducing unnecessary trading for stocks whose rank changed only slightly.

Combining Momentum with Other Factors

Momentum pairs particularly well with value because the two factors are negatively correlated. Value strategies buy cheap, beaten-down stocks (which have negative momentum), while momentum strategies buy expensive, rising stocks (which have low value scores). When combined in a portfolio, each factor tends to offset the other's worst periods, reducing overall volatility. Research by Asness, Moskowitz, and Pedersen (2013) showed that this negative correlation holds across asset classes and countries.

Adding quality filters to momentum also improves risk-adjusted returns. Stocks with strong momentum and strong fundamentals (high profitability, low leverage, stable earnings) tend to sustain their trends longer than stocks with strong momentum but weak fundamentals. This intersection of momentum and quality is sometimes called "quality momentum."