Lucena Research

01/28/2020

Cointegration is an excellent time series analysis geared to identify and exploit a high conviction trade on two time series that diverge from their expected normal behavior.

Read more about a scientific approach to arbitrage trading in foreign exchange here: How to Use Arbitrage for Strategies

A Scientific Approach to Arbitrage Trading in Forex Strategies. Cointegration is an excellent time series analysis geared to identify and exploit a high...

01/22/2020

On February 20th join CEO Erez Katz and Peter McGinnis of Equifax for a webinar to showcase how a birds-eye view of the US consumer can deliver impactful insights for: Retail, Auto, Home Builder, and other Large Cap Stocks.

Erez and Pete will share the research process and uncover the most actionable yet surprising findings followed by a Q&A discussion. Reserve your spot here. https://hubs.ly/H0mHky50

01/21/2020

How to minimize in your .

As new datasets enter the predictive analytics world, streamlining their evaluation and deployment is essential for quantitative investment strategies.

01/16/2020

Lucena CEO Erez Katz discusses how and data can be used to prices. Read the full article here. https://hubs.ly/H0mCxdy0

Neural Networks (CNNs and RNNs) are deep learning algorithms that operate on sequences. Here is how time series data and CNNs predict stocks.

01/14/2020

Mean Reversion: Discover lagging securities in anticipation of their movement. Read the full article here. https://hubs.ly/H0mzVGg0

How to discover lagging securities that have yet to follow the trend, in anticipation that they too will eventually join the party and catch up.

01/08/2020

How Can Against

The price of gold has been perceived as inversely correlated to the US dollar. When building investment strategies, gold has been used as a hedge...

01/07/2020

Is your data predictive enough to forecast stock prices or KPIs? A look into how we at Lucena validate data before risking capital. https://hubs.ly/H0lcS1n0

The pitfalls big data providers and data consumers should be aware of before using data for investment decisions. Data can be misleading and costly to...

12/30/2019

Identifying uncorrelated best of breed strategies and combining them into a multi-strategy fund is not a new concept. At Lucena, we have taken it one step further by automating the process of building such a fund from the ground up.

By validating sources and then creating Model Portfolios powered by and we can ultimately combine uncorrelated models into a sustainable algorithmically traded market neutral multi-strategy fund.

Find out more here: Creating a Multi-Strat Fund Using Big Data and Machine Learning

For a diversified portfolio, we combined quantitative research and big data to maximize the value of uncorrelated models into a sustainable long/short fund

12/20/2019

Dynamic retraining is an important aspect of AI and Big Data for quantitative investing. Why? Because applications are exponentially more difficult when used in the context of decisions.

Data is naturally noisy and can at times be intentionally deceiving to obfuscate other investors' true intentions.
CEO Erez Katz addresses how dynamic retraining can be most useful in combating common challenges when using AI and to forecast .

Machine learning for stock forecasting presents many unique challenges. Dynamically retraining models is crucial for success in quantitative investing.

12/18/2019

In 2019 we began assessing as an overlay to our CNN approach to time series forecasting using image classification and object detection. We've uncovered significant advantages to over traditional CNNs.

Find out more on how CapsNets overcome some of the biggest drawbacks of traditional CNNs. Especially in the context of deep learning for stock forecasting.
https://hubs.ly/H0mh_TN0

Capsule Networks, a new tech for image classification and time series forecasting. The benefits compared to CNNs and potential to forecast stock prices

12/16/2019

How to Qualify the Best Machine Learning Models for Investment Research

One of the biggest challenges in quantitative research for investment is to identify the best machine learning model from an array of trained models.

12/13/2019

How to use and to forecast for decisions https://hubs.ly/H0mc7ty0