Start with one path
New user: Strategies -> Backtest -> Compare -> Paper. Existing user: jump straight to Metrics, Methodology, Data Sources, or Risk.
Guide & Methodology
This is the canonical operating manual for Alphrex: workflow, metrics, methodology, data sources, and trust limits in one reference surface.
Recommended reading path
Use this as the operating manual for the research loop.
Keep this collapsed unless you want the workflow framing first. The actual documentation starts immediately below.
Read this page as a workflow reference, not as a long introduction. Start in the product, then use the relevant section only when you need definitions, limits, or trust framing.
Keep the trust ceiling in view
Ticker coverage is tiered: Free is core-only, while Pro extends into best-effort symbols and baskets rather than promising blanket coverage.
Treat the docs as support material
The point is to clarify the loop without blocking you behind a large intro block before the actual documentation starts.
Use the guide tactically. The value is not reading every section in order; it is knowing where the platform trust boundary, validation rules, and workflow handoffs actually sit.
Do not linger at the top
The real reference starts in the section index below. Open this panel only when you need the workflow framing or policy boundary.
Use metrics and methodology first
If you already know the surfaces, the highest-value refresh is Metrics, Methodology, Data Sources, and Risk rather than re-reading page descriptions.
Treat ticker policy as a contract
Provider expansion is telemetry-driven. Free keeps the narrow core contract; Pro widens access but keeps the support promise weaker.
Canonical manual
Use this page when you need operating instructions, metric definitions, or the product trust boundary. If you want the staged theory layer instead of the manual, open Learn. If you only want the compressed trust summary, open Methodology Summary.
Dashboard
Your research command centre
The dashboard gives you a live overview of your research activity, the market, and today's signals in a single view. Nothing here executes trades — it is a read-only summary panel.
Stat tiles
Counts of your backtests, strategies, best Sharpe achieved, watchlist size, paper strategies, and portfolios. Updates on every page load.
Market Snapshot
Live prices and day-change percentages for the watchlist, with interactive market charts. Add symbols to the watchlist desk, click a ticker chip to feature it in Focus view, and change the chart range (6M / 1Y / 3Y / 5Y). Use Focus / 3-Up / 5-Up / All to control layout density. Hover the chart to inspect exact date and price. Click any tile to open a pre-filled backtest for that ticker.
Today's Signals
A compact view of the most recently computed signals — ticker, strategy, LONG / SHORT / FLAT badge, and signal strength bar. Click "View All →" to reach the full Signal Monitor with filters.
Recent Backtests
Your last 10 backtest runs with key metrics. Click any row to open the full result detail page. Use the "Compare →" shortcut to run multi-strategy comparisons.
Strategy Library
Browse and select strategies to test
The library contains all built-in strategies plus any custom strategies you have written. Each card shows the strategy category, difficulty level, a one-line description, and the key parameters you can tune in the backtester.
Category badges
Mean ReversionPrice reverts to a mean; trade the deviation.
MomentumTrend-following; buy strength, sell weakness.
VolatilityTrade volatility expansions or contractions.
FactorSystematic factor exposures such as value, carry, or size.
StatisticalModel-driven frameworks such as Kalman, OU, or cointegration.
Difficulty
Beginner strategies have 1–2 parameters and are easy to interpret. Advanced strategies have more parameters and more complex signal logic — they are not necessarily better; they are just harder to attribute and tune responsibly.
Param chips
The small monospace tags under each card name the tunable parameters. You can override their values in the Backtest page.
Details
Opens the strategy detail page with a full description, signal logic explanation, and parameter reference.
My Strategies
Custom strategies you have written in the Sandbox or Custom Strategy editor. They are scoped to your account only.
Backtest
Run a single strategy on historical data
The backtester runs a chosen strategy on a chosen ticker over a date range and returns an equity curve plus a full metrics table. All results are stored in your account and accessible from the Dashboard.
Ticker
Free is intentionally narrow: liquid core symbols only. Pro widens the surface to extended best-effort symbols and weighted benchmark baskets, but unsupported long-tail coverage is still not an implied promise.
Benchmark
A second ticker used as a buy-and-hold reference. Defaults to the same ticker. SPY is a common baseline for equity strategies. Leave blank to skip benchmark overlay.
Commission (bps)
One-way transaction cost in basis points, applied to each turnover dollar. 5 bps ≈ a typical retail broker. Set to 0 for gross returns. Higher values penalise high-frequency strategies more severely.
Slippage (bps)
Execution slippage in basis points on top of commission. Models the bid-ask spread and market impact on fills. Also applied per-turnover.
OOS Split %
Reserves the last N% of the date range as a terminal holdout. Headline Sharpe/CAGR/MaxDD remain full-sample metrics, while OOS metrics are shown separately so you can judge how much of the story survives the untouched tail.
Validation Folds
Validation adds a second diagnostic layer on top of the terminal holdout. You can run sequential walk-forward with a purge gap or combinatorial purged CV. Both summarize fold/path behavior, but they do not replace the headline full-sample metrics.
Strategy params
Each strategy exposes its own tunable parameters (window length, threshold, etc.) below the main form. Defaults are sensible starting points but you should explore the sensitivity.
Tip: Always run with realistic costs first (5–10 bps commission + slippage). Many strategies that look excellent at 0 bps collapse at realistic transaction costs due to high turnover.
Free-Plan Core Tickers
The practical safe set when validation fails on Free
This is the practical core universe for the Free tier, not an exhaustive entitlement promise. Stay inside these names when you want the clearest supported path. Outside this set, resolution can still work, but Free is not meant to imply blanket coverage.
Core ETFs
SPY, QQQ, IWM, DIA, VTI, EFA, EEM, XLF, XLE, XLK, XLV, ARKK
Core single names
AAPL, MSFT, NVDA, GOOGL, AMZN, TSLA, META, JPM
Rates and real assets
TLT, IEF, HYG, AGG, GLD, SLV, USO
Core crypto
BTC-USD, ETH-USD, SOL-USD
Major indices
^GSPC, ^SPX, ^NDX, ^IXIC, ^DJI, ^RUT, ^VIX
Rule of thumb: If you want non-US suffixes, niche indices, or weighted benchmark baskets, you are outside the narrow Free reliability contract and into Pro best-effort territory.
Metric Reference
What each number in a backtest result means
Sharpe Ratio
Annualised excess return divided by annualised volatility. Computed without a risk-free rate. >1.0 is generally considered acceptable; >1.5 is strong. Treat IS Sharpe with scepticism — always compare to OOS Sharpe.
CAGR %
Compound annual growth rate of the strategy equity curve. Net of commission and slippage.
Max Drawdown %
Largest peak-to-trough decline in equity. This is the number that determines whether you can psychologically hold the strategy through its worst period.
Win Rate %
Fraction of trades with a positive P&L. A strategy can be profitable with a sub-50% win rate if the average winner is large relative to the average loser (high payoff ratio).
Num Trades
Total closed trade slices extracted from the position path. Fractional scaling is handled lot-by-lot, so the count is more trustworthy than a naive “every size change = new round trip” approach, but it is still a product metric rather than an exchange blotter.
OOS Sharpe
Sharpe computed only on the holdout period. The IS/OOS Sharpe gap is the single most important diagnostic: a large gap signals overfitting or data snooping.
WF Avg Sharpe
Average Sharpe across walk-forward test folds. More stable than a single OOS split when the date range is short.
Overfitting warning: IS Sharpe is almost always better than OOS Sharpe. A large gap (>0.5) or a negative OOS Sharpe strongly suggests the strategy is curve-fitted to the training data and should not be traded.
Strategy Comparison
Rank multiple strategies side-by-side
The comparison page runs several strategies on the same ticker over the same period and overlays their equity curves. It is the fastest way to see which strategies actually differ versus which ones are only telling different stories about the same underlying exposure.
Strategy picker
Check two or more strategies from the grid. All cost and date settings apply uniformly to every run — you are comparing strategies on equal footing, not cherry-picking parameters per strategy.
Equity curves
All curves are normalised to start at 1.0. Divergence between strategies is the signal; convergence means they are trading the same underlying bet.
Metrics table
Full IS metrics plus OOS Sharpe and walk-forward average Sharpe for each strategy. Sort mentally by OOS Sharpe, not CAGR, to avoid selecting the most overfit strategy.
Tip: Select 4–6 strategies from different categories (one momentum, one mean reversion, one volatility) to understand whether your best backtest result is genuine or just the lucky draw from a large strategy set.
Paper Trading
Track strategies forward from today
Paper trading creates a forward record from the first live day after creation and tracks hypothetical P&L from that point onward. No capital is deployed — this is a forward-monitoring layer, not a broker bridge.
Create strategy
Choose a built-in strategy or a simple long-only / short-only sleeve plus a ticker. The system freezes the creation timestamp, uses earlier history only to warm up the signal state, then resets live NAV at the first forward observation. You can add multiple paper strategies across different tickers.
Refresh
Fetches the latest price data and updates only the forward record. You can refresh manually or wait for the daily automatic refresh at 18:00 UTC.
Portfolios
Aggregate multiple paper strategies into a portfolio with custom weights. Portfolio equity is the weighted sum of component strategy returns. This lets you track a diversified allocation rather than individual strategies.
Snapshots
The system stores daily live snapshots after the forward start date. Scoreboards rank from that stored live history rather than from a backfilled equity curve.
Execution Playbook
The separate Execution Playbook turns paper sleeves and portfolios into planning records for a future broker bridge: cadence, checks, cash buffers, and handoff specs. It is planning only, never live execution.
Note: Paper strategies do not account for real execution — fills happen at daily close prices with no slippage unless you configure it. This is forward simulation, not live trading.
Signal Monitor
Today's strategy signals across the watchlist
The signal monitor runs all built-in strategies across a curated watchlist of tickers and reports the current signal direction and strength for each combination. Signals are computed from daily bars — they update once per day, not intraday.
LONG
The strategy's logic currently indicates a long position. This means the signal crossed into buy territory as of the most recent daily bar.
SHORT
The strategy currently indicates a short position. Note that many strategies are long-only by design; SHORT signals from these strategies reflect the absence of a long, not an active short recommendation.
FLAT
The strategy is neutral — no position recommended. Most tickers will be FLAT for most strategies most of the time. Use the signal filter to hide FLAT rows and focus on actionable signals.
Strength
A normalised [0, 1] score measuring the persistence and magnitude of the signal. Higher strength = stronger conviction from the model. Use the min strength slider to filter out weak signals.
Refresh
Manually triggers an async signal recompute across the full watchlist. A progress bar shows completion. The page reloads automatically when done. The daily auto-refresh runs at 18:00 UTC.
Tip: Filter to LONG or SHORT with min strength ≥ 0.5 to see only the highest-conviction signals. Then use "Backtest" to verify the strategy's historical performance on that specific ticker before acting.
Scoreboards
Forward performance rankings
Scoreboards rank paper strategies and portfolios by their forward (live) performance — not by historical backtest results. The key distinction is that stored live history starts only after creation, while any earlier lookback is used only to initialize signal state.
Ranking metric
Strategies are ranked by forward Sharpe computed on their stored daily live history. A minimum number of live days is required before a strategy appears on the leaderboard.
Min live days
A guardrail against short-sample mythology. Strategies with fewer live days than the threshold are excluded regardless of how good their short-term performance looks.
Anti-gaming
Warmup history can initialize the signal state, but ranking starts only once the forward record begins. A strong old backtest can motivate promotion into paper, yet it does not directly seed scoreboard return history.
Important: A raw backtest leaderboard would mostly reward overfitting. This platform deliberately does not rank strategies by in-sample backtest returns for this reason.
Methodology
How the engine works — implementation details and limits
Signal timing
Signals are computed from daily OHLCV bars. A signal on day t reflects the strategy logic applied through close of day t, and the backtest applies that position to the next bar return t→t+1.
No execution latency model, no open-vs-close microstructure model, and no intraday fill guarantee.
Backtest return model
Position on bar t is applied to the forward return r(t→t+1). Commission and slippage in bps are deducted via realised turnover, and regime-conditioned overlays can add a dedicated switch-friction penalty on confirmed state changes and de-risking ramps.
Still no full market-impact model, partial-fill simulation, borrow-fee model for shorts, or venue-specific open execution model.
Out-of-sample split
The last N% of the date range is withheld as a terminal holdout. OOS metrics are shown separately, while the headline Sharpe/CAGR/MaxDD remain full-sample metrics for the full run.
Still not a nested train-tune-test optimisation workflow. The holdout is diagnostic evidence, not a proof of robustness.
Walk-forward validation
You can run either sequential walk-forward folds with a purge gap or combinatorial purged CV. Both report fold/path count, average Sharpe, worst-path Sharpe, and positive-path ratio, and both are additive diagnostics on top of the main run.
This is stronger than a single holdout, but still not proof. There is no nested parameter search, no execution simulator, and no full institutional research stack.
Benchmark comparison
Default benchmark is the same-ticker buy-and-hold. You can override with a single ticker or a weighted basket, and result surfaces now also compute constructed alternatives such as a factor-matched optimizer basket and a risk-balanced core benchmark.
Constructed benchmarks are still reference tools, not a claim that the optimizer found the “true” neutral benchmark for the strategy.
Paper scoreboards
Scoreboards rank forward-tracked paper strategies and portfolios from stored live history only. Warmup history can initialize state, but live NAV is reset at the forward start and minimum live-day thresholds still apply.
Still not a brokerage-linked live track record, and the board remains a product ranking surface rather than an institutional performance composite.
Data Sources
Where price data comes from and what can go wrong
Price data is fetched via a waterfall: Yahoo Finance → Nasdaq Data Link → Stooq. Each result carries a source tag visible in the backtest output. Coverage policy is tiered: core liquid symbols are the product contract, while long-tail or international symbols remain best-effort and should be interpreted with more skepticism.
Yahoo Finance
Primary source. Adjusted close prices. Generally reliable for major US equities, ETFs, and crypto. May have gaps for thinly traded names, delisted tickers, or non-US exchanges.
Nasdaq / Stooq
Fallback sources used when Yahoo fails. Data formatting differences are normalised but not guaranteed to be identical in all edge cases.
Invalid data handling
When the system cannot obtain trustworthy market history, the run is marked as non-decision-grade and surfaced with explicit warnings. The product prioritises continuity of workflow, but invalid data should never be treated as evidence.
Caching policy
Autocomplete and support-resolution responses are cached for a short window so repeated research inputs stay fast. Live market snapshot data uses a much shorter cache because it exists for workspace context, not archival precision.
Data source issues the tool does not eliminate: adjustment mismatches (splits, dividends), stale or missing rows, survivorship bias (delisted companies), exchange-specific quirks, and non-US corporate action handling.
Risk Disclosure
What this tool does not and cannot tell you
Not financial advice. Alphrex is a research and simulation tool. Nothing on this platform constitutes investment advice, a trading recommendation, or a solicitation to buy or sell any financial instrument.
Past backtest performance does not guarantee future results. Backtests are model outputs conditional on data quality, cost assumptions, look-ahead-bias absence, and implementation details. A strategy that performed well in backtest may fail to survive:
Regime shifts
A momentum strategy that worked in 2010–2020 trending markets may fail in a choppy or mean-reverting regime. Alphrex now includes regime-conditioned backtests and a persisted cross-asset conditioning study, but those layers still improve skepticism rather than eliminate regime risk.
Crowding
When a strategy becomes widely followed, its edge may erode as more capital competes for the same signal.
Live execution frictions
Real fills differ from modelled fills. Slippage estimates are conservative approximations — illiquid names, large orders, or volatile opens will incur larger real-world costs.
Survivorship bias
The data you have is biased toward companies that survived. Backtesting on today's S&P 500 constituent list implicitly selects for winners.