Headline Sentiment Word Matching

Article Strategy library · Author: QuantConnect

Summary

This example demonstrates a basic text-driven trading rule using custom economic news headlines and NLTK tokenization. It downloads a headline dataset, tokenizes each day’s text, and checks for user-selected positive and negative words. For the SPY equity position, a positive word prompts full allocation when not invested, while a negative word liquidates an existing holding. The schedule runs the check after the market opens.

The example sets a 2018–2019 date range and initial cash of 100,000, but reports no backtest outcomes. Its sentiment method is a literal word-presence check rather than contextual language analysis: it does not account for negation, word sense, headline relevance, or signal strength. The sample data import and parsing also assume a simple comma-separated format. It is therefore a teaching example for connecting text input to trading actions, not evidence that this sentiment rule has predictive value.

Key ideas

The algorithm tokenizes daily custom news text with NLTK before evaluating signals.
A user-defined positive word triggers a full SPY allocation when the portfolio is not invested.
A negative word causes an invested holding to be liquidated.
The example runs its signal check after the market opens and covers 2018 through 2019.
Literal word matching ignores context and the document reports no performance evidence.

Tags

Equities Sentiment Machine learning Execution

Full text

# NLTKSentimentTradingAlgorithm


# NLTKSentimentTradingAlgorithm









## Source (Apache-2.0)

```python
# QUANTCONNECT.COM - Democratizing Finance, Empowering Individuals.
# Lean Algorithmic Trading Engine v2.0. Copyright 2014 QuantConnect Corporation.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

from AlgorithmImports import *

import nltk
# for details of NLTK, please visit https://www.nltk.org/index.html

class NLTKSentimentTradingAlgorithm(QCAlgorithm):

    def initialize(self):
        self.set_start_date(2018, 1, 1)  # Set Start Date
        self.set_end_date(2019, 1, 1) # Set End Date
        self.set_cash(100000)  # Set Strategy Cash

        spy = self.add_equity("SPY", Resolution.MINUTE)
        self.text = self.get_text() # Get custom text data for creating trading signals
        self._symbols = [spy.symbol] # This can be extended to multiple symbols

        # for what extra models needed to download, please use code nltk.download()
        nltk.download('punkt')
        self.schedule.on(self.date_rules.every_day("SPY"), self.time_rules.after_market_open("SPY", 30), self.trade)

    def trade(self):
        current_time = f'{self.time.year}-{self.time.month}-{self.time.day}'
        current_text = self.text.loc[current_time][0]
        words = nltk.word_tokenize(current_text)

        # users should decide their own positive and negative words
        positive_word = 'Up'
        negative_word = 'Down'

        for holding in self.portfolio.values():
            # liquidate if it contains negative words
            if negative_word in words and holding.invested:
                self.liquidate(holding.symbol)

            # buy if it contains positive words
            if positive_word in words and not holding.invested:
                self.set_holdings(holding.symbol, 1 / len(self._symbols))

    def get_text(self):
        # import custom data
        # Note: dl must be 1, or it will not download automatically
        url = 'https://www.dropbox.com/s/7xgvkypg6uxp6xl/EconomicNews.csv?dl=1'
        data = self.download(url).split('\n')

        headline = [x.split(',')[1] for x in data][1:]
        date = [x.split(',')[0] for x in data][1:]

        # create a pd dataframe with 1st col being date and 2nd col being headline (content of the text)
        df = pd.DataFrame(headline, index = date, columns = ['headline'])
        return df

```

Shown in full with attribution under the source's licence. Licence: Apache-2.0

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.

Trading Equity Headlines with Simple NLTK Word Matching

Summary

Key ideas

Tags