skip to Main Content
×Close search
Search
Quantitative Investing Blog - Lucena Research

Perpetually Monitored Strategies Derived From Alternative Data and AI

In the spirit of validating an ever growing list of data providers, we have developed several strategies to perpetually monitor how their respective signals translate to above benchmark performance. Here is how we've combined multiple Alt Data factors to create winning investment strategies.

Erez Katz

Erez Katz, Lucena Research CEO and Co-founder

 

Alternative-Data-Based Long and Short Portfolios and Backtests in action

 

At Lucena, we connect investment professionals with actionable and validated alternative data. Over the past two years, we've added new datasets to our list of validated and researched data partners.
 
It's no secret that backtests are not sufficient enough to evaluate the potency of a signal. To further advance the reliability of our research, we've established Model Portfolios and Smart Data Feeds to perpetually evaluate the signals over time. Here I will highlight how using best of breed data combined with advanced AI can create winning investment strategies that even exceeded our expectations. 
 

Lucena's Model Portfolio

Lucena delivers Model Portfolios algorithmically powered by big data and AI. The concept is as follows:

  1. 1. Carry forward the very same execution rules of a backtest into the future where decisions are made and published before market opens.
  2.  
  3. 2. Simulate as authentically and realistically as possible in order to provide unbiased assessment of how predictive alternative data is in a real-life scenario.
  •  - Account for slippage and transactions cost.
  •  - Apply short borrowing cost when applicable.
  •  - Account for dividends, split, and reverse splits 
  •  - Rich execution guidelines such as  OCO (Order Cancel Order), Stop Loss & Target Gain, allocation guidelines.
  1. 3. Generate a comprehensive performance attribution report on demand for real-time assessment.

Below, I provide a high level synopsis of some of our Model Portfolios.

 

How Does the Model Portfolio Trade?

A multi-factor model is nothing more than a multi-criteria filter or a scan. In essence, we apply machine learning classification to identify which factors are most selective of constituents with impending price action either bullish or bearish. 

The model portfolio scans for securities daily and identifies which are most primed for entry (long or short). Once positions are identified, the model creates  and publishes trading orders pre-market based on its allocation restrictions and available cash.

Upon trade execution after market opens and based on live intra-day prices, transactions are created and the positions are then held until either their time elapsed in which they close on MOC (market on close) or an intra-day stop condition is triggered intra-day (stop loss or target gain).  Although we don't consider TCA (transaction cost analysis), this is as close as we get to a real life scenario. In addition, we are fully integrated with Interactive Brokers for a complete true assessment from a live brokerage perspective.

The idea is to provide full transparency by which all trades are determined and published pre-market. Furthermore, all trades are netted of transactions cost, and short borrowing cost.

 

Lucena Model Portfolio: Analyst Consensus

 

Analyst Consensus is a long only trading strategy that identifies stocks from the Russell 1000 with bullish analyst recommendations. The strategy further sub-selects investment positions using Lucena's proprietary technical and fundamental factors. 

Backtest 

Analyst Consensus Model Portfolio

 
 
 
 
 
 
 
 
 
 
 
 
 
Past performance is not indicative of future returns. View the full backtest report.

Perpetual Performance 

Analyst Consensus has generated healthy returns since its inception on January 2019. The goals of the strategy are to outperform the S&P 500 in both total returns and lower volatility.

 
Model Portfolio
 
Analyst Consensus Live Model Portfolio and backtest
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Past performance is not indicative of future returns. View the full live portfolio.
 

Lucena Model Portfolio: BlackDog

BlackDog is a conservative, long only strategy that follows a risk parity (RP) approach. The strategy invests in liquid Exchange Traded Funds (ETFs) each representing hundreds of individual assets.

Backtest

Lucena Research long only ETF strategy

 

 

 

 

 

 

 

 

Past performance is not indicative of future returns. View the full backtest report.

Perpetual Performance 

Blackdog 2x has provided a nice alternative to the traditional 60/40 portfolio.   Since its inception in April 2014. It has consistently outperformed its benchmark, AQR Capital risk parity fund.

Model Portfolio

ETF long only strategy

 

 

 

 

 

 

 

 

Past performance is not indicative of future returns. View the full live portfolio.

Lucena Model Portfolio: Dynamic Short-Only

The Dynamic Short Only strategy is a combination of seven  different Event Studies designed specifically to identify short opportunities. The Event Studies are based on Lucena's proprietary technical and fundamental factors. The backtest combines short positions from each model and equally allocates them based on  available cash and buying power.

Backtest

Dynamic Short Only investment strategy

 

 

 

 

 

 

 

 

Past performance is not indicative of future returns. View the full backtest report.

Perpetual Performance

Beyond beating shorting the S&P (SH as a benchmark), the Dynamic Short Only strategy has generated alpha since its inception in January 2019. The strategy’s key objectives are to outperform the ProShares Short S&P 500 (SH) in returns. 

Model Portfolio

 
Short strategies from fundamental and technical factors
 
 
 
 
 
 
 
 
 
 
 
 
 
Past performance is not indicative of future returns. View the full live portfolio.

Lucena Model Portfolio: Spdr ETF Hedged

Spdr ETF Hedge is a long/short market neutral strategy.  The core holdings are made of large sector spdr ETFs for which allocation is optimized for a max Sharpe (using Markowitz MVO).  The optimized core is further hedged with short constituents from bearish event studies. The strategy’s goal is to account for sector rotation while protecting the portfolio from sudden and protracted market correction.

 
Backtest
 
Longshort market neutral strategy
 
 
 
 
 
 
 
 
 
 
 
 
 
Past performance is not indicative of future returns. View the full backtest report.

Perpetual Performance

 
Spdr ETF Hedge has generated consistent low vol returns since its inception in January 2019. The Strategy’s key objectives are to capitalize on sector rotation cycles while preserving returns by hedging for the underlying core portfolio with short only constituents.
 
Model Portfolio 
 
ETF Hedge
 
 
 
 
 
 
 
 
 
 
 
 
 
Past performance is not indicative of future returns. View the full live portfolio.

Lucena Model Portfolio: Market Neutral Tech Hedged

 
Market Neutral Tech Hedged is a long/short market neutral strategy.  The core holdings are made of large technology firms hedged by shorting constituents from an ensemble of bearish event studies. The goal of the strategy is to preserve return while creating the best short hedge for the underlying core holdings.
 
Backtest
 
Market Neutral tech hedged portfolio
 
 
 
 
 
 
 
 
 
 
 
 
 
Past performance is not indicative of future returns. View the full backtest report.
 
Perpetual Performance 
 
Market Neutral Tech Hedged has provided healthy returns since inception in November 2018. The Strategy’s key objective is to protect high-flying technology names  from excess volatility while maximizing risk-adjusted returns.
 
Model Portfolio
 
Longshort market neutral tech hedge
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Past performance is not indicative of future returns. View the full live portfolio.

Lucena Model Portfolio: TieBreaker

 

TieBreaker is an actively managed long/short strategy. It invests in US-based large-cap equities and tilts its long/short exposure based on market risk. Its main objectives are to accrue returns over time while maintaining low beta, low volatility, & low correlation to the S&P 500. 

Backtest

 
Us-based large-cap equities strategy
 
 
 
 
 
 
 
 
 
 
 
 
 
Past performance is not indicative of future returns. View the full backtest report.
 
Perpetual Performance
 

TieBreaker has provided consistent positive  returns every year since its inception in April 2014. The strategy has demonstrated accrual of  returns over time while maintaining low beta, low volatility, and low correlation to the S&P 500.

 
Model Portfolio
 
Us-based equities longshort strategy
 
 
 
 
 
 
 
 
 
 
 
 
 
Past performance is not indicative of future returns. View the full live portfolio.

Lucena Model Portfolio: Utilities Live

 
The Utilities Live portfolio consists of up to 10 securities from the Russell 1000 that together minimize tracking error against the XLU. Using Lucena's machine learning portfolio replication and  optimization engine (MVO combined with Lucena's Forecaster) the portfolio is optimized every two weeks to maximize Sharpe Ratio. 
 
Backtest
 
Utilities live strategy portfolio
 
 
 
 
 
 
 
 
 
 
 
 
 
Past performance is not indicative of future returns. View the full backtest report.
 
Perpetual Performance
 
Utilities Live has provided excellent returns since its inception in November 2016. The portfolio tracks the XLU, generates healthy dividends and notional returns while still maintaining market relative low volatility.
 
Model Portfolio
 
Russell 1000 securities strategies
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Past performance is not indicative of future returns. View the full live portfolio.
 
 
As can be seen, choosing the right data vendor and deploying data with advanced predictive analytics can translate to investment success. Lucena empowers our buy side customers with a full life cycle solution which covers: validation, enhancement and deployment offerings for alternative data.
 
Bottom line, we don't advocate how to trade but merely provide all the necessary tools and defensible evidence needed to guide asset managers and help them make more informed decisions.
 
 
Find out more about custom strategies and derived signals, contact us.
 

 

More Posts

Never miss a new blog post, subscribe to our newsletter.