AWARD WINNING QUANT ANALYTICS PLATFORM
AS SEEN ON
Research, Backtest, and Trade
Your Investment Strategies
Research, backtest, optimize and deploy multi-asset strategies on one platform, with AI agents working every stage of the pipeline alongside you.
$248,600
Equity
+148.6%
Net Profit
91.2%
PSR
+2.4%
Return
$6.4M
Volume
Strategy Equity
$248,600
Asset Volume
Drawdown
-7.4%
Exposure
84%
Capacity
$42.6M
Orders
Symbol
Avg. Price
Quantity
Market Price
Market Value
NVDA184.20450185.06$83,277
SPY603.88200605.41$121,082
TSLA262.55-325261.10-$84,858
GLD318.90150318.02$47,703
Research Pipeline
Move every idea through a single, automated pipeline, from generation and research to backtesting, paper trading, and live deployment. A specialist AI assistant staffs each stage, does the work, and decides what earns promotion.
Research a suite of ideas in parallel, tracked in a central location, with the progression of each idea gated by evidence: statistical validation of the research, backtests with realistic reality modeling, then out-of-sample paper trading that must land inside the backtest's statistical envelope.
Research Multiple Ideas at Once
Validate with Statistical Tests
Deploy to Live Trading
Monitor Live Risk and News
Generating Idea
Fear Greed Skew Rotation
Unified Quant Infrastructure
Finding alpha is a challenging task. As the universe of data expands rapidly and the pace of technological development accelerates, you need every advantage the market has to offer.
Cloud Research
Our cloud-based research terminals attach to terabytes of financial, fundamental, and alternative data, preformatted and ready to use.
Alternative data is linked to the underlying securities, tagged with the FIGI, CUSIP, and ISIN to facilitate building strategies.
Access popular machine learning and feature selection libraries to quantify factor importance. We can install custom packages on request.
Train Machine Learning Models
Visualize and Explore Data
Fast Cloud Cores
||
Code
research.ipynb|Python 3.11.2
Momentum Regime Rotation — Research
Scratch pad for the Nasdaq 100 alpha models. Pull history through QuantBook, check the regime filters against realised volatility, and chart indicators before wiring anything into main.py.
In [1]:
# QuantBook Analysis Tool
# For more information see https://www.quantconnect.com/docs/v2/research-environment
qb = QuantBook()
spy = qb.AddEquity("SPY")
history = qb.History(qb.Securities.Keys, 360, Resolution.Daily)
In [2]:
# Indicator analysis over the research window
bbdf = qb.Indicator(BollingerBands(30, 2), spy.Symbol, 360, Resolution.Daily)
bbdf.drop('standarddeviation', axis=1).plot(figsize=(12, 4))
Out[2]:
In [3]:
history.loc['SPY'].tail()
Out[3]:
| close | high | low | open | volume | |
|---|---|---|---|---|---|
| time | |||||
| 2023-06-26 | 433.21 | 435.60 | 432.47 | 434.55 | 71242300 |
| 2023-06-27 | 438.29 | 438.71 | 433.62 | 434.14 | 68015400 |
| 2023-06-28 | 438.55 | 439.32 | 436.63 | 437.60 | 59437100 |
| 2023-06-29 | 439.20 | 440.13 | 436.90 | 437.71 | 62204800 |
| 2023-06-30 | 443.28 | 443.90 | 440.05 | 441.02 | 84338900 |
In [ ]:
FROM RESEARCH TO PRODUCTION
Backtesting
With minimal-to-no code changes, move from research to point-in-time, fee, slippage, and spread-adjusted backtesting on lightning-fast cloud cores. Perform multi-asset backtesting on portfolios comprised of thousands of securities with realistic margin-modeling.
Import custom and alternative data linked to underlying securities for realistically modeling live-trading portfolios and avoiding common pitfalls like look-ahead bias.
Our technology has been battle-tested with thousands of unit and regression tests, and more than 15,000 backtests are performed on QuantConnect daily.
$1,008,763.83
Equity
-$8,711.03
Fees
$242,445.68
Holdings
$5,772.61
Net Profit
18.120%
PSR
0.88 %
Return
$2,847.38
Unrealized
$5,894,552.90
Volume
Strategy Equity$1,008,763.83
1m3m1yAll
Exposure0.122
1m3m1yAll
Stockplot GILD$74.00
1m3m1yAll
Assets Sales Volume$5.89M
Capacity$78M
1m3m1yAll
Benchmark
1m3m1yAll
Select Chart
Strategy Equity
Assets Sales Volume
Exposure
Capacity
Benchmark
Portfolio Turnover
Stockplot GILD
Ranking
PSR: 18.1%Rank: 23
0%100%
Research Guide
Likely Not Overfit
Likely Overfitting
Likely Not Overfit
FROM RESEARCH TO PRODUCTION
Parameter Optimization
Our parameter sensitivity testing allows you to run thousands of full backtests on our scalable cloud compute, completing weeks of work in minutes.
Visualize all the iterations of parameters on heatmaps to quickly understand your strategy's sensitivity to parameters for robust out-of-sample trading.
Explore further by opening each result and seeing its trades and backtest logs to understand the source of your alpha.
18
Completed
0
Failed
0
Running
0
In Queue
00:00:04
Average Length
00:01:16
Total Runtime
18
Total
5.42 QCC
Consumed
Strategy Equities18 runs
10m1h1d1wAll
Sharpe Ratio (ema-fast, ema-slow, ema-medium)
0%
100%
Drawdown (ema-fast, ema-slow, ema-medium)
0%
100%
Compounding Annual Return (ema-fast, ema-slow, ema-medium)
0%
100%
BacktestsOptimization Id: O-ef30a0bd82eee04b939086916c1b4160
Name
64.016%
1.213
11.247
3.8%
70
140
90
42.475%
0.732
6.987
6.1%
70
140
75
23.053%
0.274
3.409
6.6%
70
120
90
50.575%
0.906
8.336
3.6%
70
120
75
0%
0
0
0%
70
100
90
32.132%
0.504
5.538
6.4%
70
100
75
Filters
Columns
Status
Completed
Configuration
Optimization Strategy
Grid Search : Max of Sharpe Ratio
Parameters
EMA-FAST
Min: 50 Max: 70
EMA-SLOW
Min: 100 Max: 140
EMA-MEDIUM
Min: 75 Max: 90
Constraints
Parameter Chart
Strategy Equities
Sharpe Ratio (ema-fast…
Drawdown (ema-fast, …
Compounding Annual …
Server Statistics
HOST
O4-12x 35
FROM RESEARCH TO PRODUCTION
Institutional-Grade Live Trading
Since 2012, QuantConnect has deployed more than 375,000 live strategies to a managed, co-located live-trading environment. Our platform processes more than $100B in notional volume per month.
Execute trades directly through our 20 integrations or to EMSX Net's 1,300 liquidity providers.
Our live feeds include US SIP, CME, FX, and major crypto exchanges. Other live feed options are available upon request.
Low Latency Dedicated Infrastructure
Redundant 10GB Fiber Internet
Integrations to 20 Destinations
Support/Momentum Regime Rotation
PaperBrokerage
62.410%
PSR
$1,284.50
Unrealized
$214.86
Fees
$10,472.31
Net Profit
10.5%
Return
$110,472.31
Equity
$48,902.10
Holdings
$1,284,650.00
Volume
Strategy Equity$110,472.31
10m1h1d1wAll
Stockplot SPY$460.00
10m1h1d1wAll
Assets Sales Volume$1.28M
Orders5 open
Symbol
Avg. Price
Quantity
Market Price
Market Value
SPY
443.18
1,250
474.86
$593,575
BTCUSD
38,412.60
1.85
41,208.44
$76,235
QQQ
372.44
-400
381.02
-$152,408
NVDA
186.26
800
185.90
$148,720
MSFT
432.49
-450
432.33
-$194,548
Select Chart
Strategy Equity
Assets Sales Volume
Stockplot SPY
Server Statistics
CPU
4.4%
RAM
561 MB / 512 MB
Host
LIVE-150-a07155520
Up Time
1d 07:07:09
FROM RESEARCH TO PRODUCTION
Bring Agentic AI to QuantConnect with Mia
Empower Mia — your agentic AI assistant — to design, backtest, optimize, and live-trade your quantitative strategies on QuantConnect through a streamlined, AI-ready workflow. Built for professional quant teams, Mia delivers the reliability, flexibility, and security needed for real-world trading systems.
Natural-language strategies
Auto backtesting
Smart code edits
Parameter optimization
Projects
Momentum Regime Rotation
Running
SPX Overnight Edge
Running
Nasdaq Vol Carry
Completed
Earnings Drift Basket
Completed
Crypto Trend Follower
Completed
Sector Rotation ETF
Completed
Mean Reversion Pairs
Completed
Gold Macro Filter
Completed
Small Cap Quality
Completed
FX Carry Overlay
Completed
Options Vol Surface
Completed
Index Rebalance Alpha
Completed
High Yield Credit Beta
Completed
Commodity Term Structure
Completed
Low Beta Defensive
Completed
Treasury Curve Carry
Completed
Build a momentum rotation strategy on the Nasdaq 100 and backtest it
Build a momentum rotation strategy on the Nasdaq 100 and backtest it
Three candidates came out of the screen. Momentum rotation scores best
on trend persistence, so I'll build that one.
Pulled three years of daily history and tested the regime filter: it keeps
the book flat through 71% of the drawdown days. Here is the research cell:
In [1]:
qb = QuantBook()
universe = qb.AddUniverse(Universe.Nasdaq100)
history = qb.History(universe, 756, Resolution.Daily)
history.close.pct_change().groupby(level=0).std().nlargest(5)
Out[1]:
symbol
NVDA 0.0311
TSLA 0.0287
AMZN 0.0221
MSFT 0.0164
AAPL 0.0182
Backtested 2022-2023 — 18.1% PSR on a 0.88% return, against a 3.2% max
drawdown. Chart is in the Backtest panel. Want me to paper trade it?
Yes, paper trade it for a month
Deployed to paper with the same universe and a 21-day rebalance.
Fills are showing in the Paper Trading panel.
Paper cleared its first week inside tolerance — promoting the allocation
to live now.
Ideas
Rotate monthly into the top decile of Nasdaq 100 names ranked on six-month trend persistence, sized down when the volatility regime turns.
512
Screened
3
Shortlisted
Research
0.71
Regime hit
3Y
Window
Backtest
18.1%
PSR
+0.88%
Return
1.21
Sharpe
Paper Trading
$100,842
Equity
+0.31%
Day
Live Trading
SPY474.86+1,284
QQQ381.02-486
NVDA185.90+932
Multi-Asset Portfolio Modeling
We accurately model multi-asset portfolio strategies, tracking real-time strategy equity across complex portfolios in backtesting and live trading. You can easily access the margin remaining for your strategy and size positions to reduce cash.
Equity
US Stock and ETFs since 1998, managing corporate actions, from tick to daily resolutions.
Equity Options
US Equity Options at minute resolution since 2010, with realistic portfolio modeling.
Indexes
US Cash Indexes since 1998 from tick to daily resolution bars on NDX, SPX, and VIX.
Index Options
US Index Options since 2012 from minute to daily resolutions, with portfolio modeling.
Futures
US Future markets at tick to daily resolutions since 2009, for the most liquid 70 contracts.
Future Options
Future Option markets at minute to daily resolutions since 2012, for the most liquid 70 contracts.
Forex
Interbank and market maker brokerage spreads, with realistic cashbook and margin lending.
CFD
Derivative CFD assets for leading brokerages for international traders with realistic spreads.
Crypto
Thousands of cryptocurrency pairs from six exchanges with cash and margin account modeling.
Rich Library of Alternative Data
Orthogonal signals are critical to building a robust strategy. We serve a rich library of alternative data with more than 40 distinct vendors covering millions of potential strategies.
Each dataset is processed with a uniform timestamp and delivered to your strategy point-in-time to avoid selection bias. With one simple line of code, your alternative data automatically links to underlying assets and tracks corporate actions through time.
Data is ready to be used in live trading, delivered in real-time in our co-located live-trading environment. Prefer to do your research on-premises? Export and download the data through our Datasets Marketplace.
Explore alternative data with a single line of code
Build vs Buy: Your Firm's Dream Quantitative Research Platform
Have you ever thought about creating your own algorithmic trading platform but aren't sure about the expenses? Use our Build vs. Buy cost calculator to accurately assess the costs of developing a professional-grade quantitative trading platform. Compare these costs with using a ready-made solution like QuantConnect to make an informed decision.
Open-Source Algorithmic Engine
LEAN is the algorithmic trading engine at the heart of QuantConnect. More than 180+ engineers contributed to the development of this lightning-fast, open-source platform. It provides modeling that surpasses the best financial institutions in the world. LEAN can be run on-premises or in the cloud.
Local Development, Cloud Backtesting
Code locally in your favorite development environment, then synchronize your projects to the cloud to work on the go with QuantConnect's IDE.
lean backtest "My Project" —debug
lean cloud backtest "My Project"
lean cloud live "My Project"
Installation: pip install lean
Enterprise-Grade Performance At Any Scale
Built on open-source LEAN and continuously rebased on it, LEAN Enterprise optimizes the layers your algorithm never touches: data decompression, memory allocation, and the hot paths under every bar. Your code doesn't change; the engine underneath it does.
Up to 40% faster execution with substantially higher data throughput.
More than 2x faster on large-scale, high-frequency datasets.
Up to ~45% faster indicator computation, accelerating research and optimization cycles.
An Inspired and Connected Global Community
QuantConnect has a global community of 540K quants, researchers, data scientists, and engineers. Collectively we are the biggest quant research community in the world with more than 1,200 strategies shared through the forums, a vast library of public quant research.
Every month, the brightest quantitative minds use our platform to generate research. On an average month 50,000 QuantConnect users create 2,500 new algorithms and write 1M lines of code.