Start freeSign in

Sonar Sciences Quant & Research Team

Quant & Research Team, Sonar Sciences
The research desk of Sonar Sciences

The Sonar Sciences Quant & Research Team is the house research desk. It writes and maintains the research section: the pillar guides, the glossary and the methodology articles, under the site's claim register and review gates. Owner decision, 2026-08-07: all final articles carry this team byline rather than a personal one, to avoid security and access issues around individual accounts.

Reviewed by Sonar Sciences Quant & Research Team

117 articles
ArticleWhy arithmetic and geometric returns differStrategy research fundamentals3 minArticleBar replay or automated backtestingPlatform comparisons3 minArticleCan you backtest in a spreadsheetPlatform comparisons5 minArticleCloud or local compute for backtestingPlatform comparisons4 minArticleDatabases or flat files for market dataPlatform comparisons4 minArticleWhat is a deflated Sharpe ratio?Strategy validation & overfitting9 minArticleWhat paid market data buys over free feedsPlatform comparisons6 minArticleHow trading costs compound over timeStrategy research fundamentals3 minArticleHow does backtesting workStrategy research fundamentals5 minArticleHow is drawdown calculatedStrategy research fundamentals4 minArticleHow leverage changes risk of ruinStrategy research fundamentals5 minArticleHow long should you paper trade a strategyStrategy research fundamentals5 minArticleHow many trades does a backtest needStrategy validation & overfitting4 minArticleHow to align timestamps across data sourcesCross-venue market data & signals5 minArticleHow to annualize returns and volatilityStrategy research fundamentals3 minArticleHow to archive a retired strategyFrom research to publishing4 minArticleHow to audit someone else's backtestStrategy validation & overfitting6 minArticleHow to benchmark a backtesting engine for correctnessPlatform comparisons3 minArticleHow to bootstrap confidence intervals for a backtestStrategy validation & overfitting3 minArticleHow to build a point-in-time datasetCross-venue market data & signals6 minArticleHow to calculate position sizeStrategy research fundamentals4 minArticleHow to choose a backtesting platformPlatform comparisons6 minArticleHow to choose a benchmark for a strategyStrategy validation & overfitting4 minArticleHow to choose an out-of-sample splitStrategy validation & overfitting5 minArticleHow to choose candle timeframes for researchCross-venue market data & signals4 minArticleHow to clean tick dataCross-venue market data & signals3 minArticleHow to compare backtesting engines fairlyPlatform comparisons3 minArticleHow to compare execution quality across brokersPlatform comparisons3 minArticleHow to compare charting tools for researchPlatform comparisons4 minArticleHow to compare crypto data providersPlatform comparisons5 minArticleHow to compare two equity curvesStrategy research fundamentals6 minArticleHow to detect a regime change in dataCross-venue market data & signals3 minArticleHow to detect bad prints and outliersCross-venue market data & signals5 minArticleHow to detect data leakage in strategy featuresStrategy validation & overfitting7 minArticleHow to detect interpolated candlesCross-venue market data & signals4 minArticleHow to detect look-ahead bias in a backtestStrategy validation & overfitting3 minArticleHow to detect regime dependence in resultsStrategy validation & overfitting4 minArticleHow to detect stale quotes in a feedCross-venue market data & signals3 minArticleHow to detect survivorship bias in a datasetStrategy validation & overfitting5 minArticleHow to document a trading strategyFrom research to publishing4 minArticleHow to estimate slippage from dataCross-venue market data & signals4 minArticleHow to evaluate a market data vendorPlatform comparisons4 minArticleHow to find gaps in historical dataCross-venue market data & signals5 minArticleHow to handle exchange outages in a datasetCross-venue market data & signals3 minArticleHow to keep a strategy research logFrom research to publishing5 minArticleHow to make a backtest reproducibleFrom research to publishing4 minArticleHow to measure strategy capacityStrategy validation & overfitting6 minArticleHow to model transaction costs in a backtestStrategy validation & overfitting4 minArticleHow to prove when a result was producedFrom research to publishing4 minArticleHow to read a backtest reportStrategy research fundamentals6 minArticleHow to read a strategy tester report criticallyPlatform comparisons5 minArticleHow to read an equity curveStrategy research fundamentals6 minArticleHow to reconcile data from multiple venuesCross-venue market data & signals3 minArticleHow to report a backtest honestlyFrom research to publishing4 minArticleHow to run a permutation test on a strategyStrategy validation & overfitting5 minArticleHow to run an independent strategy reviewFrom research to publishing5 minArticleHow to separate research code from execution codeFrom research to publishing4 minArticleHow to set an embargo period in validationStrategy validation & overfitting3 minArticleHow to store market data for researchCross-venue market data & signals4 minArticleHow to stress test a strategy across market regimesStrategy validation & overfitting3 minArticleHow to test parameter stabilityStrategy validation & overfitting4 minArticleHow to trial a research platform in a weekPlatform comparisons6 minArticleHow to validate exit rules separatelyStrategy validation & overfitting4 minArticleHow to verify a data vendor's historyCross-venue market data & signals5 minArticleHow to version a datasetFrom research to publishing4 minArticleNotebooks or scripts for strategy researchPlatform comparisons3 minArticleOpen source or commercial backtesting toolsPlatform comparisons5 minArticlePaper trading or backtestingPlatform comparisons5 minArticlePython or Pine Script for strategy researchPlatform comparisons4 minArticleREST or websocket market dataPlatform comparisons5 minArticleRetail and institutional data feeds comparedPlatform comparisons4 minArticleCan you separate skill from luck in trading resultsStrategy research fundamentals6 minArticleVectorized or event-driven backtestingPlatform comparisons3 minArticleWhat belongs in a performance disclosureFrom research to publishing4 minArticleWhat counts as a trading strategyStrategy research fundamentals3 minArticleWhat counts as investment adviceFrom research to publishing3 minArticleWhat is a good Sharpe ratioStrategy research fundamentals4 minArticleWhat is risk per tradeStrategy research fundamentals3 minArticleWhat the MT5 strategy tester does and does not testPlatform comparisons3 minArticleWhy a backtest needs a cost modelStrategy validation & overfitting5 minArticleWhy a single data feed biases a backtestCross-venue market data & signals4 minArticleWhy a smooth equity curve is suspiciousStrategy validation & overfitting4 minArticleWhy a track record needs an audit trailFrom research to publishing4 minArticleWhy adding rules usually adds overfittingStrategy validation & overfitting3 minArticleWhy backtest start dates matterStrategy validation & overfitting4 minArticleWhy broker feeds differ from exchange feedsCross-venue market data & signals5 minArticleWhy the same candle differs across venuesCross-venue market data & signals4 minArticleWhy consolidated prices hide venue behaviourCross-venue market data & signals3 minArticleWhy drawdown statistics need long samplesStrategy validation & overfitting3 minArticleWhy every public claim needs a sourceFrom research to publishing4 minArticleWhy execution research needs tick dataCross-venue market data & signals4 minArticleWhy feature checklists mislead platform choicesPlatform comparisons4 minArticleWhy financial promotion rules matter to quantsFrom research to publishing3 minArticleWhy funding rates matter in perpetual backtestsCross-venue market data & signals3 minArticleWhy identity checks exist in financeFrom research to publishing3 minArticleWhy liquidity varies by session and venueCross-venue market data & signals3 minArticleWhy losses hurt more than gains helpStrategy research fundamentals3 minArticleWhy one out-of-sample pass is not enoughStrategy validation & overfitting4 minArticleWhy optimized parameters fail out of sampleStrategy validation & overfitting4 minArticleWhy optimizing for return alone overfitsStrategy validation & overfitting3 minArticleWhy perpetual and spot prices divergeCross-venue market data & signals6 minArticleWhy pre-registering a test keeps you honestFrom research to publishing3 minArticleWhy public track records skew positiveFrom research to publishing4 minArticleWhy publication needs a review gateFrom research to publishing3 minArticleWhy random entries can look profitable in backtestsStrategy validation & overfitting4 minArticleWhy market regimes belong in your sampleCross-venue market data & signals4 minArticleWhy reported volume can misleadCross-venue market data & signals3 minArticleWhy reproducibility matters in trading researchFrom research to publishing4 minArticleWhy backtest results differ across platformsPlatform comparisons4 minArticleWhy screenshots are not a track recordFrom research to publishing3 minArticleWhy Sharpe ratios shrink out of sampleStrategy validation & overfitting4 minArticleWhy stationarity matters for backtestsStrategy validation & overfitting3 minArticleWhy strategy changes need version historyFrom research to publishing5 minArticleWhy the same strategy fills differently at two brokersPlatform comparisons3 minArticleWhy timezone errors corrupt daily candlesCross-venue market data & signals4 minArticleWhy trade frequency changes what results meanStrategy validation & overfitting7 minArticleWhy you must count every backtest you runStrategy validation & overfitting4 min