Context

Gold, the yellow shiny metal, has been the fancy of mankind since ages. From making jewelry to being used as an investment, gold covers a huge spectrum of use cases. Gold, like other metals, is also traded on the commodities indexes across the world. For better understanding time series in a real-world scenario, we will work with gold prices collected historically and predict its future value.

Content

Metals such as gold have been traded for years across the world. Prices of gold are determined and used for trading the metal on commodity exchanges on a daily basis using a variety of factors. Using this daily price-level information only, our task is to predict future price of gold.

Data

For the purpose of implementing time series forecasting technique , i will utilize gold pricing from Quandl. Quandl is a platform for financial, economic, and alternative datasets. To access publicly shared datasets on Quandl, we can use the pandas-datareader library as well as quandl (library from Quandl itself). The following snippet shows a quick one-liner to get your hands on gold pricing information since 1970s.

import quandl gold_df = quandl.get("BUNDESBANK/BBK01_WT5511")

The time series is univariate with date and time feature

Starter Kernel(s)

-Start with Fundamentals: TSA & Box-Jenkins Methods

This notebook is an overview of TSA and traditional methods

Acknowledgements

For this dataset and tasks, i will depend upon Quandl. The premier source for financial, economic, and alternative datasets, serving investment professionals. Quandl’s platform is used by over 400,000 people, including analysts from the world’s top hedge funds, asset managers and investment banks.

Inspiration

• Forecast gold price

*If you find the data useful your upvote is an explicit feedback for future works, Have fun exploring data!*

#

MORE DATASETs ...

PowerMap U.S. is an innovative trading solutions, specializing in order flow analytics on U.S. Stock market. With its AI-inferred proprietary algorithm trained on market data, TradePulse predicts stock flow on using trade volume and buy intensity providing an additional key metric for decision-making while providing catalogue of alternative dataset on its platform.

Key Features: 💠 AI-driven order flow prediction based on trade volume and buy-side intensity 💠 Proprietary algorithms trained on historical and real-time U.S. equity data 💠 Real-time analytics across major U.S. exchanges (NYSE, NASDAQ, etc.) 💠 Integrated dashboard with visual flow indicators and trend detection 💠 Access to alternative datasets curated for quantitative and discretionary strategies 💠 Customizable signals aligned with trading styles (momentum, mean-reversion, etc.) 💠 Scalable infrastructure suitable for institutional-grade workflows

Primary Use Cases: 🔹 U.S.-focused hedge funds leveraging inferred flow data for intraday alpha 🔹 Quantitative traders integrating buy-side pressure metrics into models 🔹 Execution teams identifying optimal entry/exit points through real-time flow signals 🔹 Asset managers enhancing conviction through AI-derived trade behavior insights 🔹 Research analysts and PMs utilizing alternative datasets for cross-validation of ideas

Contact us for a real time order flow data in different markets. Stay ahead with TradePulse's order flow insights.