Quantpedia: Strategy Research Guide

By Sean MackeyApr 4, 202513 min read

Quantpedia is a research database, not a trading platform: its team reads academic finance papers, keeps the small share that pass its selection rules, and rewrites each surviving strategy as plain-language trading rules with the paper's performance and risk figures attached. The site now lists more than 1,000 strategies, links to more than 2,000 papers, and carries more than 800 out-of-sample backtests built with QuantConnect. This guide explains how strategies get into the database, how to use the Screener and read a strategy page, what the Prime, Premium and Pro plans actually include at current prices, and how to take an idea from Quantpedia to a costed backtest on your own market with Quant Charts, where Quant, LuxAlgo's coding agent, writes the Pine Script for you.

Key points:

Quantpedia curates, it does not manage money. Strategies are extracted from published research; the company states it runs no client funds.
Two numbers per strategy. The in-sample figures come from the source paper; the out-of-sample backtest, where one exists, comes from QuantConnect. The gap between them is the useful signal.
Plans differ by depth, not by topic. Prime covers about a hundred essential strategies, Premium the full database, Pro the backtests, portfolio reports, API and live-strategy tools.
A paper's rules are a specification, not a result. Re-test them on the exact symbol and timeframe you would trade, with commission and slippage set, before believing any number.

How Strategies Get Into Quantpedia

Quantpedia's process page describes a filter rather than a factory. The team monitors research portals, journals, universities and conferences daily, and evaluates each candidate paper on whether the strategy can actually be implemented, how long its backtest period is, and whether the overall approach is sound. Black-box strategies are removed automatically, a clearly stated performance characteristic is required, and papers that also report risk figures such as maximum drawdown or volatility are preferred. Only a small percentage of strategies pass. Those that do are rewritten as trading rules in plain language, tagged with performance and risk characteristics, the instruments and markets traded and the backtest period, categorised by keywords, and linked to similar strategies already in the database.

A subset is then implemented in QuantConnect's framework, producing out-of-sample backtests with equity curves, trading statistics and source code. That cooperation was announced in July 2018 with roughly twenty strategies covered; the pricing page now puts the out-of-sample set at around 800 strategies, with the code and backtest for roughly 400 of them refreshed monthly. Quantpedia describes itself as a focused quant-research company that manages no client funds or accounts, which is worth keeping in mind when reading performance figures: nothing on the site is a track record.

Using the Screener

The Screener is the front door. Its filters are Keywords, Markets Traded, Instruments Available, Rebalancing Period, Crisis Hedge and QuantConnect Code, so you can ask for, say, monthly-rebalanced strategies on ETFs with an out-of-sample backtest attached. The results grid has several views: Paper Overview, Paper Description, Paper Risk & Return, and, for strategies with a QuantConnect implementation, OOS Return Stats, OOS Risk Stats and OOS Risk & Return. Each row carries the strategy ID, title, rebalancing period, markets, the source paper's performance and volatility, and keywords; the first entry, Asset Class Trend-Following, is a monthly strategy across equities, bonds, commodities and REITs tagged momentum, asset class picking and trend-following. A free sign-up lets you browse the free strategies, with a viewing limit before a subscription is required.

Two habits make the Screener more useful. First, filter on Rebalancing Period and Instruments Available before anything else: a strategy that rebalances monthly across liquid ETFs is a different research task from one that trades single stocks daily, and most retail research capacity suits the former. Second, switch to the OOS views once you have a shortlist, because the paper's numbers and the replication's numbers are rarely the same, and the difference is the first thing to investigate.

What Each Plan Includes

Quantpedia sells three individual plans, each in three-, twelve- and thirty-six-month terms. The table summarises the feature comparison and the prices on the plan pages as of September 2026.

Feature	Prime	Premium	Pro
Price per 12 months (3 and 36 months)	$399 ($299 / $699)	$599 ($449 / $1,199)	$799 ($599 / $1,599)
Strategies	100+ essential strategies: tactical asset allocation, simple market timing, seasonality	All 900+ Premium strategies and thousands of papers	All 900+ Premium strategies and thousands of papers
Database updates	Irregular	About 10 to 15 new strategies per month	About 10 to 15 new strategies per month
Out-of-sample screening	No	No	About 800 strategies; code and backtest for about 400 refreshed monthly
Portfolio modelling	40 common ETFs and about 100 essential strategies	40 common ETFs and about 100 essential strategies	200+ liquid ETFs, 40 cryptocurrencies, about 400 strategies, about 100 third-party factors
Reports	7 essential Pro reports	7 essential Pro reports	All of about 40 Pro reports, with new ones added monthly
Own equity curves	Up to 3	Up to 3	Unlimited, plus custom benchmarks
API, Market Overview, AI chatbot	No	No	Yes

The practical reading is that Prime is a tactical-allocation and seasonality subscription, Premium is the encyclopaedia itself, and Pro is the analysis layer: the out-of-sample backtests, the portfolio reports and the API. The Pro report list runs from the basics, correlation, equity-crisis analysis and ETF replication, through multi-factor models, hundred-year portfolio and market-state analysis, VaR, Monte Carlo, Kelly and Optimal F sizing, Markowitz, CPPI, risk parity and volatility targeting, to cross-sectional and time-series momentum, seasonality, trend and reversal edges, rebalancing and strategy grading. If your question is which published anomalies exist, Premium answers it; if your question is how a portfolio of them would have behaved, that is Pro.

LuxAlgo Seasonality Widget on Quant Charts bucketing price change by calendar period — The Library's Seasonality Widget on Quant Charts. Prime's seasonality strategies describe calendar effects in general; the widget measures whether your particular symbol honours one, by hour, weekday, day of month, month or day of year.

Reading a Strategy Page

A strategy page has two layers of evidence. The top layer is the paper: the rules in plain language, the markets and instruments, the backtest period, and the return and volatility the authors reported. The bottom layer, where it exists, is the QuantConnect replication: an equity curve on data the authors did not use, trading statistics, and the source code. Quantpedia's own description of the purpose is that users can check whether out-of-sample performance matches the in-sample information from the paper, and that check is the whole point of the page.

The Library's entry on the in-sample and out-of-sample split explains why the comparison matters: parameters tuned on one stretch of history always look good on that stretch, so data the strategy has never seen is the cleanest test of whether an edge is real, and the size of the performance drop between the two is a rough gauge of how much of the original result was curve fit. Some degradation is normal. A Sharpe ratio that softens is one story; one that halves, or a win rate that collapses toward a coin flip, is another. Read the replication's drawdown as carefully as its return, and note the backtest's end date, because a strategy replicated only through a benign period has not been tested at all on the conditions that break it.

Quantpedia's own blog models the sceptical reading. A recent post on a paper by Matti Suominen and Erik Hjalmarsson reports that equity time-series momentum, one of the most heavily backtested anomalies, tends to break down and reverse when valuation measures such as the Shiller CAPE ratio, aggregate dividend yields and the yield-curve slope reach historical extremes. Another post uses the Pro API to pull a peer group of 21 trend-following strategies and shows that an average return hides the dispersion inside the group, so a strategy can trail the average and still sit comfortably within the normal range of outcomes. Both are reminders that the database is a source of hypotheses, and the site itself says so.

Pro Tools: Portfolios, Live Strategies and the API

Pro's portfolio modelling accepts ETFs, cryptocurrencies, Quantpedia strategies, third-party factor series and your own uploaded equity curves, and runs the report set against them, with custom benchmarks built from the same ingredients. In June 2026 Quantpedia added a Live Strategies section: users connect a live or paper-trading algorithm running on QuantConnect by entering the unique strategy hash QuantConnect generates, and the strategy then appears in a reporting interface that embeds QuantConnect's live report of equity curve, margin use, holdings and trading statistics, with live strategies combinable into portfolios. At launch it carried two of Quantpedia's own strategies running on its QuantConnect servers.

For anyone who wants to paper trade a replication, Quantpedia's 2022 case study is the honest guide: it recommends minute-resolution data so that market orders fill when intended rather than at the next open, warming indicators up with historical data before the first live bar, and keeping any custom signal data refreshed rather than loaded once as a static file. It also repeats the point that the whole exercise is meant to reveal whether a backtest was overfit before real money is involved.

From a Quantpedia Idea to a Costed Backtest

Quantpedia gives you a specification: a rule set, a rebalancing period, a universe, and two sets of performance figures. It does not give you a test on the instrument and timeframe you would actually trade, and the paper's universe rarely matches yours. That test is the job of Quant Charts. Describe the rule to Quant in plain language, as concretely as the paper states it, for example a monthly rotation into the asset with the highest twelve-month return, or a long position when the twenty-day average sits above the two-hundred-day. Quant writes the Pine Script as a strategy, you inspect it under Code and click Run, and the chart marks the entries and exits while a Backtest Summary reports net profit, trade count, win rate, maximum drawdown and profit factor.

Adding indicators on Quant Charts. Library studies load in a click for the exploratory reading; Quant writes the strategy script when a Quantpedia rule set is ready to test.

Two settings decide whether that summary means anything. The strategy Properties hold commission and slippage, alongside initial capital, order size, pyramiding and margin; set them to your broker's terms before reading net profit, because the Library's page on execution cost modelling shows how quickly a frequently rebalanced edge disappears once costs are realistic. And the docs are explicit that results depend on the data they run on: re-run the strategy on the exact symbol and timeframe you intend to use, because results from another market or interval do not transfer. Once the script runs, parameter changes happen in Inputs without re-generating anything, which is how you check parameter stability: a lookback that only works at one value is a curve fit, whatever the paper said. If you started with an indicator, the Backtest button asks Quant to rewrite it with entries and exits.

LuxAlgo Seasonality Chart on Quant Charts showing the averaged seasonal curve for a symbol across the calendar year — The Library's Seasonality Chart on Quant Charts averages each calendar day's change into one seasonal curve, the same construction behind the calendar anomalies Quantpedia catalogues, applied to your own instrument.

The validation vocabulary is the Library's. Hold out the most recent block as described in the split entry and spend it once; roll the split forward with walk-forward analysis if the strategy has tunable parameters; and read the overfitting entry before trusting any backtest whose in-sample and out-of-sample figures diverge sharply. Quantpedia's out-of-sample replication does one such test for you on the paper's universe; Quant Charts lets you do the same on yours.

Where Each Tool Stops

Quantpedia stops at research and reporting: it publishes rules and figures, and, for Pro users, monitors strategies that run on QuantConnect. Quant Charts stops at the costed backtest and the chart: the LuxAlgo platform does not place orders for you, and the platform is for research and testing, with the Journal for recording what you did with the result. Execution, whether paper or live, belongs to a brokerage or to a platform such as QuantConnect that connects to one.

A Research Workflow

Screen narrowly. Filter by Rebalancing Period, Instruments Available and QuantConnect Code so every result is something you could run and something with a replication.
Read both layers. Note the paper's return and volatility, then the out-of-sample figures and the replication's end date; investigate the gap.
Write the specification down. Entry, exit, universe, rebalancing period and any filter, in one paragraph you could hand to a coder.
Hand it to Quant. Describe the rule, inspect the Code, click Run on the exact symbol and timeframe you would trade.
Cost it. Set commission and slippage in Properties, then reread net profit, maximum drawdown and profit factor.
Stress the parameters. Vary the lookbacks in Inputs; keep only edges that survive a neighbourhood of values and a held-out period.
Paper trade before capital. Use a brokerage simulator or QuantConnect's paper trading, following Quantpedia's own tips on resolution and warm-up.

Conclusion

Quantpedia earns its place by doing the unglamorous part of quantitative research at scale: reading the papers, discarding the black boxes, extracting rules in plain language, and, for a growing subset, paying for an independent out-of-sample replication. Its plans are priced by depth, from Prime's tactical-allocation shelf to Pro's backtests, reports and API, and its own blog is candid that averages mislead and trends break at extremes. What it cannot do is test a rule on your market with your costs. That is what Quant Charts is for: Quant writes the strategy, the Backtest Summary judges it with commission and slippage in place, and the Library's validation concepts tell you whether to believe the number. Used together, the database supplies the hypotheses and the chart supplies the verdict.

Key Takeaways

Quantpedia curates academic strategies into plain-language rules; it manages no money and publishes no track record.
Compare the paper's in-sample figures with the QuantConnect out-of-sample replication; the gap measures overfitting.
Prime is tactical allocation and seasonality, Premium is the full database, Pro adds backtests, reports, API and live-strategy monitoring.
Re-test any rule on Quant Charts on the exact symbol and timeframe, with commission and slippage set.
Vary parameters in Inputs and hold out recent data before trusting a result.

FAQs

What is Quantpedia?

Quantpedia is a subscription research database that extracts trading strategies from academic finance papers, rewrites their rules in plain language with the paper's performance and risk figures, and, for a subset, adds out-of-sample backtests built with QuantConnect. It is a research company, not an asset manager or a broker.

How much does Quantpedia cost?

As of September 2026 the twelve-month prices are $399 for Prime, $599 for Premium and $799 for Pro, with three-month and thirty-six-month terms also offered. Prime covers about a hundred essential strategies, Premium the full database of more than 900, and Pro adds out-of-sample screening, full portfolio modelling, about 40 reports, unlimited uploaded equity curves and the API.

What does the out-of-sample backtest on a strategy page mean?

It is a replication of the paper's rules run in QuantConnect on data the authors did not use, with an equity curve, trading statistics and source code. Comparing it with the paper's in-sample figures shows how much of the original result was likely curve fit.

Can Quantpedia trade strategies for me?

No. Quantpedia publishes research and, for Pro users, can monitor live or paper strategies that you run yourself on QuantConnect by linking the strategy hash. Execution happens at a brokerage or on a platform connected to one.

How do I test a Quantpedia strategy on Quant Charts?

Describe the rules to Quant in plain language, inspect the generated Pine Script under Code and click Run on the symbol and timeframe you would trade. The Backtest Summary reports net profit, trade count, win rate, maximum drawdown and profit factor; set commission and slippage in the strategy Properties first.

Should I trust a strategy's published return?

Treat it as a hypothesis. Quantpedia's own blog notes that time-series momentum breaks down at valuation extremes and that peer-group averages hide wide dispersion. Re-test on your market with costs, hold out recent data, and vary the parameters before drawing conclusions.