Research - 06.06.2024 - 12:00
The validity of day trading as a long-term consistent and uncorrelated source of income for traders and investors is a matter of debate. In their paper, the researchers sought to answer this central question. Using a large dataset that covered more than 7,000 US stocks traded from 2016 to 2023, the research aimed to assess how effective day trading was in producing consistent and uncorrelated returns. To their knowledge, this is the first published paper on the topic using a comprehensive database of stock prices at an intraday frequency. We sat down with Professor Andrea Barbon to learn more.
Andrea Barbon, what inspired you to look into this topic of day trading?
I, along with my research partners have been captivated by the recent surge in day trading for quite some time. Day trading is a dynamic form of investing where individuals buy and sell securities within the same day. As more people invest their time and skills into trading, we wanted to explore whether this trend towards day trading makes sense and can be profitable or if it was simply driven by the desire to get rich quickly without pursuing a traditional career. Our objective was to determine if there are any straightforward day trading strategies that can consistently generate profits.
Moreover, we noticed a lack of academic research on intraday investment strategies in the US stock markets and wanted to investigate this further.
What were you expecting to find?
My research interest revolves around technology and data science applied to financial markets, including blockchain technology and Artificial Intelligence models in asset pricing.
We started by examining a popular day trading strategy introduced in the 1990s by Toby Crabel, known as the Opening Range Breakout (ORB) strategy. The ORB strategy involves taking positions based on the price range established in the first 5 to 60 minutes of the trading day. Our goal was to assess if the strategy was still effective in producing an economically significant trading edge for active day traders.
What were the main results? What did you find?
Our research found that this strategy continues to provide a trading edge, but the key is to focus only on "Stocks in Play", which are stocks with significant intraday trading activity, often those which see a market reaction to fundamental news. By doing so, we created a strategy that generates statistically and economically significant returns, with almost zero correlation to the main equity benchmark for these stocks.
Did anything surprise you?
We were pleasantly surprised by our initial positive results, especially given the simplicity of the strategy's implementation. This outcome was unexpected and encouraging for further research.
Is this good news for those who wish to pursue day trading on a regular basis?
It's crucial to determine whether day trading is a meaningful source of income or if it represents a societal loss in terms of wasted working hours. Our initial findings suggest that while day trading has potential, it is challenging for the average person to rely on it as a primary income source. Understanding this can help society better allocate time and resources.
Where will you be focusing next?
Our future research will test the profitability of other popular day trading strategies and explore the intrinsic value of technical analysis. In collaboration with Concretum Research, we are developing a unique tool named R-Candles, available as a free web app at r-candles.com. This trading simulator allows users to test discretionary technical trading strategies using random events from a large sample of stocks over a long period, without knowing the stock names or event dates. By solely analyzing daily price movements (usually exhibited as candles, a graphical representation of the price range traded by a stock in one day), users can simulate long and short trades while also managing open positions by adding profit targets and stop losses.
Despite being in beta for the last three months, R-Candles already has over 1,000 active users, though it currently only offers daily price data, which isn't ideal for day trading. We are now working on upgrading the tool to include intra-day data, aiming to attract more users who can explore day trading, understand its potential, and improve their skills without financial risk. From an academic perspective, once we gather a substantial user database and their performance data, we can test hypotheses about day trading and technical analysis, enhancing our scientific understanding of these phenomena.
Andrea Barbon is an assistant professor at the Swiss Institute of Banking and Finance at the University of St.Gallen. His research work focuses on the impact on asset prices of demand shocks, arising from fire sales by institutional investors, and quantitative easing programs. Moreover, he is interested in the application of recent machine learning models in the context of asset pricing.
Image: Adobe Stock / Xavier Lorenzo