Scoring methodology
Every strategy is backtested on the same 15 years of real daily data, then ranked by a single composite score that balances return against risk.
The composite score
A weighted blend of six components, each fed by your real backtest metrics (no random padding):
| Component | Weight | Rewards |
|---|---|---|
| Return (CAGR) | 30% | Higher compound annual growth (sigmoid-normalized so it can't run away) |
| Sharpe ratio | 25% | Return earned per unit of risk |
| Max drawdown | 20% | Capital preservation — deep losses are penalized exponentially |
| Consistency | 15% | Share of winning months |
| Stability | 5% | Durable long-run compounding |
| Volatility | 5% | Lower portfolio volatility |
The leaderboard re-computes this score on the server from your submitted metrics, so it can't be spoofed, and it stores your strategy definition so any entry is reproducible.
How the backtest works
- Data: daily adjusted close (splits + dividends) for 200+ US stocks & ETFs, ~15 years, via Yahoo Finance.
- Costs: a per-side transaction cost is charged on turnover. Taxes are not modeled — see Taxes & turnover for why that matters for high-turnover strategies.
- Benchmark: every strategy is compared to SPY (S&P 500) over the identical window.
- Determinism: the same strategy always produces the same result — fair comparison for everyone.
Taxes & turnover
Every result on StratPick is pre-tax. For a buy-and-hold strategy that barely matters, but for strategies that trade frequently it can be the difference between a great backtest and a mediocre real-world outcome. Here's why.
- Short-term vs long-term gains. In a US taxable account, a position sold after being held one year or less is taxed at your ordinary-income rate (up to 37% federal, plus state). Held longer than a year, the gain qualifies for the lower long-term rate (0/15/20%). Strategies that rotate holdings monthly or quarterly realize mostly short-term gains.
- Turnover is the driver. A rough rule: average holding period ≈ 1 ÷ annual turnover. A strategy with ~150% annual turnover holds the typical name about 8 months — under a year, so gains are short-term. Buy & hold has ~0% turnover; momentum, RSI, and monthly factor rotation can run several hundred percent.
- Wash-sale rule. If you sell a stock at a loss and buy it back within 30 days — common when a name drops out of the top-N and re-enters — the IRS disallows the loss for that trade, deferring the tax benefit. Frequent rotation makes this more likely.
- Account type changes everything. Inside a tax-advantaged account — IRA, 401(k), or Roth — trades don't trigger capital-gains tax, so turnover costs you nothing tax-wise. The tax drag discussed here applies to taxable brokerage accounts only.
- What the backtest still misses. Because taxes aren't modeled, a high-turnover strategy's after-tax return in a taxable account will generally be lower than the figure shown. Two strategies with the same pre-tax return are not equal if one churns and the other holds.
The results panel flags this automatically: build a high-turnover strategy and you'll see a ⚠️ pre-tax / short-term-gains warning above the equity chart. None of this is tax advice — consult a professional for your situation.
Strategy types
Six ways to turn your universe into a portfolio. All are backtested identically and scored the same way.
| Type | What it does |
|---|---|
| Buy & Hold | Buy your selected stocks in equal amounts on day one and hold to the end — the simplest baseline. |
| Fixed allocation | Hold your picks at equal weight, rebalancing back to equal weight monthly or quarterly (trims winners, tops up laggards). |
| Trend (price > moving average) | Own a stock only while its price sits above its N-day moving average; move to cash on that name when it drops below. A simple trend filter. |
| Momentum (hold top-N) | Each rebalance, rank your stocks by their return over a lookback window and hold the top N, rotating as leadership changes. |
| Mean reversion (RSI) | Buy stocks that are oversold (14-day Relative Strength Index below your buy level) and sell once they recover (above your sell level). Bets on short-term bounces. |
| Custom factor model | Score every stock on the factors you weight (see below) and hold the top-ranked names, rotating as scores change. |
Advanced options
| Option | What it does |
|---|---|
| Benchmark | The index your strategy is charted and measured against (SPY, QQQ, IWM, DIA, VTI). It also sets the reference for the Relative-Strength factor and the bear-market filter. |
| Bear-market filter | An optional safety switch. Once a month it checks whether the benchmark is trading below its 200-day moving average — a slow line that marks the market's long-term trend. Above the line, your strategy stays fully invested; below it, the filter moves everything to cash and waits on the sidelines until the benchmark climbs back above. The big, drawn-out crashes generally happen below that line, so sitting them out cushions the worst drawdowns — but it also pulls you out of short dips that quickly recover, so it usually lowers total return. A trade of some upside for a smoother ride. |
| Sector-neutral ranking (factor model) | Ranks each stock against its own sector's peers instead of the whole market, so a single hot sector can't crowd out the portfolio — a more diversified, best-in-breed selection. |
| Inverse-volatility weighting (factor model) | Allocates more capital to steadier (lower-volatility) holdings and less to jumpy ones, smoothing the equity curve versus plain equal weight. |
The custom factor model — the eight factors
One strategy type, Custom factor model, picks stocks by blending measurable characteristics called factors. Each rebalance, every stock in your universe is scored on each factor; the scores are standardized across the universe (a z-score, so they're comparable), multiplied by your slider weights, summed into one composite, and the top-N by that composite are held equal-weight. As a stock falls in the ranking it is rotated out and replaced. Every factor is computed point-in-time from ~15 years of adjusted daily prices — no look-ahead.
| Factor | How it's measured (price only) | The idea |
|---|---|---|
| 12-Month Momentum | Total price return over the last ~12 months (252 trading days). | Winners tend to keep winning over 3–12 month horizons — the momentum effect. |
| 6-Month Momentum | Price return over the last ~6 months (126 days). | A medium-term momentum tilt; quicker to react than the 12-month version. |
| 3-Month Momentum | Price return over the last ~3 months (63 days). | A short-term momentum tilt; most responsive, but noisier. |
| Trend (vs 200-day avg) | How far today's price sits above (or below) its 200-day moving average. | A classic trend gauge — trading above the long-run average marks a durable uptrend. |
| Low Volatility | The inverse of annualized volatility of daily returns over ~6 months (calmer = higher score). | The low-volatility anomaly: historically, steadier stocks have delivered strong risk-adjusted returns — counter to the notion that more risk means more reward. |
| Mean Reversion (oversold) | The inverse of the 14-day RSI (Relative Strength Index) — more oversold scores higher. | Bets that recently beaten-down names snap back toward their average. It pulls against momentum, so blending the two balances chasing strength with buying dips. |
| Relative Strength vs SPY | The stock's 6-month return minus the S&P 500's over the same window. | Favors stocks that are outperforming the broad market, not merely rising with it. |
| Drawdown Resilience | The inverse of the worst peak-to-trough drop over the last ~6 months (smaller drop = higher score). | A resilience / quality tilt — rewards stocks that held up better through recent declines. |
Weights are relative. Set any factor to 0 to ignore it; the rest are blended in proportion — a 50 next to a 25 gives the first factor twice the say. Because scores are standardized each period, a positive score just means "better than the average stock in your universe that period," and holdings are chosen by rank (top-N), not an absolute cutoff.
Why no valuation or quality factors? Every factor here is price-derived on purpose. Fundamental factors (P/E and other valuation ratios, ROIC and other quality measures, growth) would require point-in-time historical financials; scoring 15 years of history with today's fundamentals is look-ahead bias, so they're deliberately excluded until that data is sourced.
Honest limitations (read this)
- Overfitting: a strategy tuned to look great on history often fails live. High score ≠ good future bet.
- Survivorship bias: the ticker universe is today's well-known names; delisted failures aren't included.
- No slippage/liquidity modeling beyond a flat cost; real fills differ.
- Hypothetical results are not real trading and not a prediction.
Frequently asked questions
How does StratPick score a strategy?
Is the leaderboard ranked only by return?
How does StratPick handle overfitting?
Are taxes included?
Where does the price data come from?
Further reading
Background on the concepts behind these factors and metrics:
- Factor investing — weighting stocks by measurable characteristics
- Momentum investing
- Sharpe ratio — risk-adjusted return
- Backtesting and its pitfalls
Books worth reading
If you want the theory behind the game, these are genuine recommendations. Disclosure: some links below may become affiliate links, meaning we could earn a small commission at no extra cost to you if you buy — it never changes what we recommend or what we say about it.
- A Random Walk Down Wall Street — Burton Malkiel. The classic case for humility, market efficiency, and low-cost index funds.
- The Little Book of Common Sense Investing — John C. Bogle. Why the boring buy-and-hold baseline is so hard to beat.
- Your Complete Guide to Factor-Based Investing — Berkin & Swedroe. The evidence behind the very factors in the builder.
- Advances in Financial Machine Learning — Marcos López de Prado. A rigorous look at why backtests overfit and how to guard against it.