A Systematic, Factor-Driven Approach to Equity Portfolio Design
(Research & Experimental Study)
Most individual investors approach markets by asking what to buy.
Professional equity funds focus on a different problem entirely:
How do we design a repeatable process that continuously holds the best available stocks, while controlling risk and behavioral errors?
This paper outlines and experimental model on how an individual investor can structure their decision-making like an equity fund manager—using objective ranking, disciplined portfolio construction, and periodic rebalancing—rather than discretionary stock picking.
1. How Equity Funds Actually Operate
Despite differences in branding, most equity funds share the same internal architecture:
- A defined investable universe
- A set of measurable factors
- A ranking and aggregation mechanism
- A fixed portfolio size
- A predefined rebalance schedule
Alpha, where it exists, comes from process discipline, not prediction.
2. Defining the Investable Universe
Before ranking stocks, professional managers first restrict what they are allowed to own.
Typical constraints include:
- Minimum market capitalization
- Adequate liquidity
- Reporting and listing quality
This step removes:
- Illiquid names
- Structurally fragile companies
- Stocks unsuitable for systematic allocation
In institutional investing, exclusion improves robustness.
3. Factor Framework: Decomposing Stock Leadership
Instead of narratives, stock selection is reduced to measurable dimensions that capture different aspects of leadership.
Earnings Strength
- Recent EPS growth – captures acceleration
- Annual / YoY EPS change – validates sustainability
Capital Efficiency
- Return on Equity (ROE) – filters low-quality growth
Market Leadership (Momentum)
- Short-term rate of change (ROC) – identifies current participation
- Intermediate ROC – filters short-term noise
- Long-term ROC – confirms primary trend leadership
Size Constraint
- Market capitalization – ensures liquidity and scalability
Each factor addresses a different risk:
- Momentum without fundamentals
- Growth without efficiency
- Small stocks that cannot absorb capital
4. Independent Ranking: Reducing Bias
A critical design choice is ranking each factor independently, rather than combining raw values.
Process:
- Rank all stocks on each factor separately
- Convert each metric into a relative rank (or percentile)
- Treat all ranks symmetrically
This approach:
- Normalizes scales
- Reduces dominance of any single variable
- Improves stability across regimes
No single metric determines inclusion.
5. Aggregate Ranking: Let the System Decide
Individual ranks are aggregated (simple or weighted average) into a composite score.
Stocks with the best composite ranks represent:
- Strong earnings
- Efficient capital usage
- Sustained market leadership
- Institutional viability
Portfolio selection is an outcome of aggregation—not conviction.
6. Portfolio Construction
Typical professional-style constraints:
- 10–15 holdings
- Equal-weight or volatility-adjusted weights
- Maximum position limits
This balance:
- Preserves factor expression
- Controls idiosyncratic risk
- Avoids index replication
Too few holdings increase fragility.
Too many dilute alpha.
7. Rebalancing: Continuous Portfolio Upgrade
At predefined intervals (monthly or quarterly):
- Re-rank the full universe
- Remove stocks whose ranks deteriorate
- Replace them with higher-ranked candidates
This is not frequent trading—it is systematic renewal.
The portfolio remains invested, but never static.
8. Core Objective
The objective is not prediction or timing.
The objective is to remain continuously invested in the highest-ranked stocks available at any point in time.
This mindset shift—from stock picking to portfolio governance—is what differentiates professional processes from retail behavior.
Appendix A: Practical Implementation Using TradeStation RadarScreen
The above framework can be implemented using common retail platforms.
Below is an illustrative TradeStation EasyLanguage indicator, designed to extract the raw inputs required for ranking and aggregation on a universe such as the S&P 500.
This is research tooling, not a trading signal.
What This Indicator Provides
- Earnings growth metrics (EPS change, YoY EPS)
- Capital efficiency (TTM ROE)
- Multi-horizon momentum (22D, 64D, 1Y)
- Market capitalization filter
- Monthly reference price for rebalance logic
These outputs can be:
- Ranked directly in RadarScreen, or
- Exported to Excel for composite ranking
TradeStation EasyLanguage (Illustrative)
variables:
EPSCHNGYR( 0 ),
oEPSCHNGYRErr( 0 ),
AEPSCHG( 0 ),
oEPSCHNGYoyErr(0),
TTMROEPCT( 0 ),
oROECHNGTTYErr(0),
PriceChange_1Yr(0),
PriceChange_64D(0),
PriceChange_22D(0),
MktCap( 0 ),
oErrorCode( 0 ),
Price(0),
High52Week(0),
DeviationPct(0);
// ---------------- Price & Momentum ---------------- //
Price = Close;
High52Week = Highest(High, 252);
DeviationPct = ((Price - High52Week) / High52Week) * 100;
PriceChange_1Yr = (Close - Close[252]) / Close[252] * 100;
PriceChange_64D = (Close - Close[64]) / Close[64] * 100;
PriceChange_22D = (Close - Close[22]) / Close[22] * 100;
// ---------------- Fundamental Data ---------------- //
oEPSCHNGYRErr = 0;
EPSCHNGYR = FundValue( "EPSCHNGYR", 0, oEPSCHNGYRErr );
oEPSCHNGYoyErr = 0;
AEPSCHG = FundValue( "AEPSCHG", 0, oEPSCHNGYoyErr );
oROECHNGTTYErr = 0;
TTMROEPCT = FundValue( "TTMROEPCT", 0, oROECHNGTTYErr );
oErrorCode = 0;
MktCap = FundValue( "MKTCAP", 0, oErrorCode );
// ---------------- RadarScreen Outputs ---------------- //
if oEPSCHNGYRErr = fdrOk then Plot1( EPSCHNGYR, "EPS_Change_1Y" );
if oEPSCHNGYoyErr = fdrOk then Plot2( AEPSCHG, "EPS_YoY" );
if oROECHNGTTYErr = fdrOk then Plot3( TTMROEPCT, "ROE" );
Plot4( PriceChange_1Yr, "ROC_1Y" );
Plot5( PriceChange_64D, "ROC_64D" );
Plot6( PriceChange_22D, "ROC_22D" );
if oErrorCode = fdrOk then Plot7( MktCap, "MarketCap" );
How an Excel-Savvy Researcher Can Use This
- Apply the indicator to an S&P 500 RadarScreen
- Export all columns to Excel
- Rank each factor independently
- Aggregate ranks into a composite score
- Select top N stocks
- Rebalance at predefined intervals
At this point, the process mirrors institutional equity portfolio construction.
Disclaimer
This document is purely educational and research-oriented.
- It does not constitute financial advice
- It is not a recommendation to buy or sell securities
- The models, indicators, and methodologies described are experimental
Market behavior is uncertain, and results depend on implementation details, costs, timing, and assumptions. Readers are solely responsible for any decisions made using this material.
