our Alpha API Brochure
Transcription
our Alpha API Brochure
NewsHedge ALPHA API: Actionable Signals for Algorithmic Trading New technologies. New catalysts. A whole new tool for capturing alpha. Problem. Machine-readable technologies have forever changed the way stocks trade. Individual asset prices are increasingly more susceptible to news, rumors, TV mentions, and Twitter posts. So what does it all mean? It means that this new trading environment requires a new approach. And new tools for capturing alpha. Solution. To help you react to these opportunities, we’ve developed the NewsHedge ALPHA API, a self-learning, self-adjusting market-scanning technology that uncovers notable price and volume velocities happening within individual assets, including velocities occurring instantly (milliseconds), or over a few minutes. 2 How It Works. Leveraging powerful algorithms and direct access exchange feeds, NewsHedge continuously tracks and measures tick data for generating trading behavior metrics on over 6,000 individual assets. The NewsHedge analysis engine uses this complex grid of historical trade metrics and compares it to how an asset is currently trading in real time. By comparing how an asset has statistically traded in the past to how it’s currently trading, NewsHedge is able to earmark anything notable and unusual—allowing you to capture opportunities and manage risks in a systematic, probable manner. REAL-TIME TICK DATA HISTORICAL TICK DATA 6000+ individual assets Milliseconds through 30 days 3 Why You Need It. Because you need smart, actionable data to successfully power: • Algorithmic trading • Quantitative models • Big-Data solutions • Trend-analysis models • Risk management systems Because you can create deep learning strategies using NewsHedge in conjunction with: • Machine-readable news • Social sentiment • Twitter velocity data • Chart indicators / Market internals • Proprietary data sets 4 NewsHedge In Action. On October 14, 2015, a typical trading day, NewsHedge generated a total of 65 events. 44 (67%) of these events were tied to market-moving news. Results by Catalysts: 25 Intraday, market-moving news (8 events moved greater than 6% by EOD) 21 Unusual trading behavior, but no known catalyst 7 Pre-market news 6 After-hour news from prior trading day (carryover) 5 Trading halts 1 Trading halt, news pending Click HERE to download the complete case study. Click HERE to visit our daily case studies blog. 5 Which Assets Does NewsHedge Cover? NewsHedge tracks individual assets which trade at or above $1.90 intraday, with a minimum daily trading volume of 320,000 shares. NewsHedge is currently covering: NASDAQ: 2,343 NYSE: 2,455 AMEX: 1,098 For the full comprehensive list of Component Sectors, U.S. Equity Sectors, Country Indexes, and Foreign Sector Indices, please visit www.newshedge.com/assets. Note: Coverage can be customized to meet even the strictest of demands. 6 Available Data Fields. NewsHedge provides Identifiers for mapping securities, Financials to support financial modeling, and unique exchange level trade data for generating accurate, probable trading decisions. Identifiers: Asset name Ticker BSYM entity ID BSYM composite ID Industry Sector Financials: Market cap Average daily volume Shares float % owned by insiders % owned by institutions Current short interest Short interest total Short interest prior total Short percent change Days to cover ratio NewsHedge: Move direction Move strength Trigger price Trigger timestamp Recent trade volume Recent trade count Recent dark volume Trading halt 7 Using NewsHedge Data Fields. Move direction: Describes the price direction as “up” or “down”. This can be used as a directional trigger. Move strength: “Jogging”, “Running”, or “Spiking” are terms used to describe the proprietary measurements of speed and velocity of the detected price change. This can be used as a rush trigger, filter mechanism, directional trigger, or decision maker. Jogging - Medium price move within a 5-minute period. Running - High price move within a 2-minute period. Spiking - Drastic price move happening instantaneously, or within a 40-second time period. Trigger price: Price which triggered the alert. This can be used as a reference point. Trigger timestamp: Unix timestamp when the Trigger price was captured. This can be used as a reference point. Recent trade volume: Total trade volume from the 10 seconds that preceded the alert. This can be used as a filter mechanism or decision maker. Recent trade count: Number of trades executed in the 10 seconds that preceded the alert. Compare this to recent trade volume and/or recent dark volume to enhance filtering and decision making. Recent dark volume: Quantity of “recent trade volume” that was reported to the Nasdaq and NYSE Trade Reporting Facilities (TRF). This can be used as a filter mechanism or decision maker to react upon dark pool trades and block orders. Trading halt: Exchange specific messages for “status” and “reason”. This can be used as a rush trigger, filter mechanism, directional trigger, or decision maker. 8 Intraday Momentum Strategy: S&P 500 This backtest model examines the ability to profit on trade direction once a “jogging” or “running” price velocity alert is generated. We started with a beginning balance of $1M, used $80K (rounded) per trade, and leveraged position limits which never exceeded 95% of the total portfolio value. Commissions are calculated at $0.003 per share and are included in the P&L. No overnight holdings. Time Horizon: January 01, 2014 - February 08, 2016 Benchmark: S&P 500 index (SPY) Buy if: Move direction = up and Move strength = jogging or running and Recent trade volume >= 12,480 and Recent trade count >= 6 and Recent dark volume < 4,099 40% and Bid price > Trigger price and Current Time - Trigger timestamp < 5 minutes and (No trade has already been triggered by this alert) Trading Universe: S&P 500 individual stocks (rebalanced) Trade Time Conditions: On at 9:46:00am. Off at 3:48:00pm (EST) Sell if: Move direction = down and Move strength = jogging or running and Recent trade volume >= 22,480 and Recent trade count >= 15 and Recent dark volume < 10,099 and Ask price < Trigger price and Current Time - Trigger timestamp < 5 minutes and (No trade has already been triggered by this alert) Close if: Position achieves 3.35% gain or Position goes negative 1.65%, or Position is still open 5-min prior to close. 9 Intraday Momentum Strategy: S&P 500 continued. Summary: Total Alerts: 1,156 Total Trades: 584 Total Profit/Loss: $247,791 Winning Trades: 303 Losing Trades: 281 Ratio win/loss: 1.69 % Winning Trades: 51.88% % Losing Trades: 48.12% Avg. Winning Trade: $2,003 Avg. Losing Trade: $1,278 Total Returns: 24.78% Benchmark Returns: 0.91% Credits: Jason Stone & Jeff Pate - Trinity Christian College Max Drawdown: ($13,940) Alpha: 0.96 Beta: 0.48 Sharpe: 2.97 Sortino: 5.12 Volatility: 0.18 10 Intraday Mean Reversion Strategy: Nasdaq 100 This backtest model examines the ability to fade price overreactions—often brought on by news headlines—using “running” or “spiking” price velocity data as the main trigger. We started with a beginning balance of $1M, used $80,000 (rounded) per trade, and leveraged position limits which never exceeded 95% of the total portfolio value. Commissions are calculated at $0.003 per share and are included in the P&L. No overnight holdings. Time Horizon: January 01, 2014 - February 08, 2016 Benchmark: Nasdaq 100 index (QQQ) Buy if: Move direction = down and Move strength = running or spiking and Recent trade volume >= 33,020 and Recent trade count >= 28 and Recent dark volume < 17,099 and Bid price > Trigger price and Current Time - Trigger timestamp < 36 seconds and (No trade has already been triggered by this alert) Trading Universe: Nasdaq 100 individual stocks (rebalanced) Trade Time Conditions: On at 9:46:00am. Off at 3:48:00pm (EST) Sell if: Move direction = up and Move strength = running or spiking and Recent trade volume >= 26,020 and Recent trade count >= 23 and Recent dark volume < 14,299 and Ask price < Trigger_price and Current Time - Trigger timestamp < 36 seconds and (No trade has already been triggered by this alert) Close if: Position achieves 2.90% gain or Position goes negative 1.75%, or Position is still open 5-min prior to close. 11 Intraday Mean Reversion Strategy: Nasdaq 100 continued. Summary: Total Alerts: 1,005 Total Trades: 884 Total Profit/Loss: $256,961 Winning Trades: 417 Losing Trades: 467 Ratio win/loss: 1.68 % Winning Trades: 47.17% % Losing Trades: 52.83% Avg. Winning Trade: $2,174 Avg. Losing Trade: $1,391 Total Returns: 25.70% Benchmark Returns: 10.80% Credits: Jason Stone & Jeff Pate - Trinity Christian College Max Drawdown: ($18,911) Alpha: 0.71 Beta: 0.72 Sharpe: 2.15 Sortino: 4.89 Volatility: 0.27 12 NewsHedge Delivers Speed and Redundancy. In Your Preferred Language. NewsHedge uses the famously fast and accessible ØMQ library for network communications. This library supports a huge number of programming languages and is well known for handling the boilerplate and edge-cases common in networked messaging. (It’s often called “ZeroMQ” for googling purposes) • Available in C, C++, Objective C, Python, Java, .NET, Ruby, PHP, and more • Fast, direct TCP connection bypassing web latencies • Separate server push and search query protocols • Architecture independent • True, real-time (milliseconds) intra-minute reporting • Fully customizable parameters to suit your specific needs 13 Who Is NewsHedge? NewsHedge was founded in 2012 out of our own need for real-time U.S. equity risk solutions. Since then, NewsHedge has grown into a diverse team of experienced market professionals who have been quietly helping numerous systematic trading firms react to market-moving events. We’ve also invented new methods for distributing financial information—see U.S. Patent No. 9,135,662. Today, our solutions help power some of the world's top quant groups, prop firms and hedge funds. We’re also beginning to form licensing partnerships with some of today’s most respected financial applications so they can distribute our automated marketmoving alerts. 14 Management. Kevin Evenhouse - Founder, CEO https://www.linkedin.com/in/kevinevenhouse Prior to founding NewsHedge, Kevin owned a capital-management practice with trade operations conducted on the CME Group’s exchange floor. He’s successfully traded a wide range of financial instruments, developed numerous automated trading strategies, and patented new methods for disseminating financial information. Kevin’s knowledge, passion, and imagination for finding solutions to the financial industry’s challenges are the key driving forces behind NewsHedge. Drew Dormann - Co-Founder, CTO https://www.linkedin.com/in/drewdormann Drew brings over 17 years of diverse programming experience to the NewsHedge organization. He’s worked on numerous performance-critical applications, and his background includes building games for Microsoft, Sony, and Midway Games, and most recently developing complex trading systems for companies such as TradeLink and Jump Trading. He’s one of only 255 Gold Badge holders for C++ through stackoverflow.com. 15 Legal Disclaimer. THE MATERIALS IN THIS PRESENTATION ARE FOR INFORMATIONAL PURPOSES ONLY. THESE MATERIALS ARE PROVIDED “AS IS” WITHOUT ANY EXPRESS OR IMPLIED WARRANTY OF ANY KIND INCLUDING, WITHOUT LIMITATION, WARRANTIES OF MERCHANTABILITY, OWNERSHIP, NON-INFRINGEMENT OF INTELLECTUAL PROPERTY, OR OTHER HARMFUL COMPONENTS OR FITNESS FOR ANY PARTICULAR PURPOSE, INCLUDING ANY WARRANTY FOR THE USE OR THE RESULTS OF THE USE OF SERVICES OR PRODUCTS WITH RESPECT TO THEIR CORRECTNESS, ACCURACY, COMPLETENESS, RELIABILITY, PERFORMANCE, OR CONTINUED AVAILABILITY. NEITHER ALGO INNOVATIONS LLC. (dba NewsHedge) NOR ITS THIRD PARTY PROVIDERS SHALL BE RESPONSIBLE FOR INVESTMENT DECISIONS, DAMAGES, OR OTHER LOSSES RESULTING FROM USE OF THIS INFORMATION. LIMITATION OF LIABILITY: NEITHER ALGO INNOVATION LLC. (dba NewsHedge) NOR ITS THIRD PARTY PROVIDERS NOR ANY PERSON THROUGH WHOM ALGO INNOVATIONS LLC. (dba NewsHedge) MAKES THIS INFORMATION AVAILABLE SHALL BE LIABLE FOR ANY DAMAGES OF ANY KIND ARISING FROM OR RELATING TO THESE TERMS, THE INFORMATION PROVIDED, OR THE USE OF, OR RELIANCE ON, ALGO INNOVATIONS LLC. (dba NewsHedge) DATA PRODUCTS, WHETHER IN CONTRACT OR TORT OR OTHERWISE. 16 Get In Touch. Enough about us, let’s talk. [email protected] Click HERE to watch NewsHedge intro video. Click HERE to read our recent press release. Click HERE to learn about our intellectual property. 17