Individual Investing

Given wide bid-ask spreads, do retail option traders systematically bear large losses? In their January 2024 paper entitled “An Anatomy of Retail Option Trading”, Vincent Bogousslavsky and Dmitriy Muravyev characterize retail option trading in the U.S. by exploiting data from a trading journal that attracts retail investors by offering advanced tracking/performance verification tools. When users subscribe and link their brokerage accounts, their trades (past and future) are automatically imported by the journal and verified. Using data for 5,182 traders who initiated 2.1 million trades worth about $20 billion (of which about $8.8 billion involve options) during January 2020 through December 2022, they find that: Keep Reading

In their 2023 book, The Missing Billionaires: A Guide to Better Financial Decisions, authors Victor Haghani and James White seek “to give you a practical framework, consistent with the consensus of university finance textbooks, for making good financial decisions that are right for you. Good decisions will take account of your personal circumstances, financial preferences, and your considered views on the risks and expected returns of available investments. …You will likely get the most out of this book if you have already accumulated a decent amount of financial capital or if you are young with a healthy measure of human capital. …The book is written from the perspective of a US individual or family…” Based on their many years of wealth management experience and portfolio systems development, they conclude that:

Keep Reading

What trading experience makes individual day traders quit trading? In their November 2023 paper entitled “Why Do Individuals Keep Trading and Losing?”. Fernando Chague, Bruno Giovannetti, Bernardo Guimaraes and Bernardo Maciel study the life cycle of individual traders who repeatedly open and close stock or futures positions on the same trading day. They focus on gross daily profit for individuals who begin day-trading during the sample period, day-trade for at least 30 trading days and then quit day-trading during the sample period. They ignore trading costs, costs of any trading courses taken and taxes. Using anonymized daily trade data for all Brazilian day traders during 2012 through 2018, they find that: Keep Reading

What is the best way to exploit U.S. federal government tax code allowing capital losses to offset current or future capital gains and up to $3,000 of current regular income? In his August 2023 paper entitled “Tax-Loss Harvesting: A Primer”, Harry Mamaysky discusses many features of tax loss harvesting, selling securities at a loss and replacing them with different (non-wash sale) but statistically similar stocks. In 10-year simulations, he assumes:

• Statistically similar, non-wash sale assets are available to replace tax-loss sales.
• The capital gain tax rate during the life of the strategy is 30%, with liquidation capital gain tax rate either 0% (charity or inheritance) or 20%.
• The portfolio has no cash inflows from initial purchase through terminal date, with proceeds from tax loss sales allocated equally across pre-existing/replacement stocks and all stocks held at respective average cost bases.
• Annual pairwise return correlation between stocks is 0.40, in line with historical evidence.
• In some simulations, realized tax losses carry over to terminal portfolio liquidation. In others, realized tax losses offset capital gains from other accounts.

Based on these assumptions, he concludes that:

Keep Reading

What modeling techniques help avoid biases/overfitting in use of machine learning to predict stock returns? In his July 2023 paper entitled “Less is More? Reducing Biases and Overfitting in Machine Learning Return Predictions”, Clint Howard explores how modeling choices affect machine learning as applied to predicting next-month stock returns, as follows:

• He considers 11 machine learning methods encompassing ordinary least squares, elastic net, random forest, gradient boosted regression trees, deep neural networks with one to five layers, an ensemble of the five neural networks and an ensemble of all methods.
• Initially, he uses the first 18 years of his sample (March 1957 to December 1974) for model training and the next 12 years (January 1975 to December 1986) for validation. Each December, he retrains with the training sample expanded by one year and the validation sample rolled forward one year.
• He trains all 11 machine learning models either on all firm/stock data together or separately on distinct groups of large, medium-sized and small firms, with size-based predictions subsequently merged.
• For each of the two sets of predictions each month, he sorts stocks into tenths, or deciles from highest to lowest predicted excess return and reforms a hedge portfolio that is long (short) the tenth, or decile, of stocks with the highest (lowest) predicted excess returns.

He calculates breakeven portfolio frictions (zero alpha) for multi-factor models of stock returns, including a 6-factor (market, size, book-to-market, profitability, investment, momentum) model. Using a database of 206 monthly firm/stock characteristics during March 1957 through December 2021, he finds that: Keep Reading

Can state-of-the-art artificial intelligence (AI) applications such as GPT-4, trained on the text of billions of web documents, provide sound financial advice? In their June 2023 paper entitled “Using GPT-4 for Financial Advice”, Christian Fieberg, Lars Hornuf and David Streich test the ability of GPT-4 to provide suitable portfolio allocations for four investor profiles: 30 years old with a 40-year investment horizon, with either high or low risk tolerance; and, 60 years old with a 5-year investment horizon, with either high or low risk tolerance. As benchmarks, they obtain portfolio allocations for identical investor profiles from the robo-advisor of an established U.S.-based financial advisory firm. Recommended portfolios include domestic (U.S.), non-U.S. developed and emerging markets stocks and fixed income, alternative assets (such as real estate and commodities) and cash. To quantify portfolio performance, they calculate average monthly gross return, monthly return volatility and annualized gross Sharpe ratios for all portfolios. Using GPT-4 and robo-advisor recommendations and monthly returns for recommended assets during December 2016 through May 2023 (limited by availability of data for all recommended assets), they find that:

Keep Reading

Should individuals who trade zero-days-to-expiration (0DTE) S&P 500 Index options expect to make money? In their March 2023 paper entitled “Retail Traders Love 0DTE Options… But Should They?”, Heiner Beckmeyer, Nicole Branger and Leander Gayda examine performance of retail 0DTE S&P 500 Index option trades. They focus on effects of the introduction of daily expirations for such options in mid-May 2022. Using daily S&P 500 Index option trade data from CBOE, including trader-type transaction codes, during January 2021 through February 2023, they find that: Keep Reading

Does a constant real annual withdrawal rate of 4% of household savings at retirement, derived from U.S. asset return experience, really protect against financial ruin? In their September 2022 paper entitled “The Safe Withdrawal Rate: Evidence from a Broad Sample of Developed Markets”, Aizhan Anarkulova, Scott Cederburg, Michael O’Doherty and Richard Sias consider data from 38 developed countries to assess safe withdrawal rates. This sample mitigates survivorship/easy data biases of the U.S. experience by including multiple left-tail instances of trading halts, wars, hyperinflation and other extreme events. They use this data to model retirement portfolio performance via stationary block bootstrap simulation, with longevity risk incorporated from U.S. Social Security Administration mortality tables. Their base case examines joint investment-longevity outcomes for a couple retiring in 2022 at age 65 using a 60% domestic stocks-40% bonds (60-40) portfolio strategy. They also look at other fixed stocks-bonds allocations and investment strategies pursued by target-date funds. Using monthly (local) real returns for domestic stocks, international stocks, bonds and bills as available for 38 developed countries during 1890 through 2019, they find that: Keep Reading

How does elimination of broker commissions on stock trades affect individual investors? In their September 2022 paper entitled “Fee the People: Retail Investor Behavior and Trading Commission Fees”, Omri Even-Tov, Kimberlyn George, Shimon Kogan and Eric So examine how retail investors respond to selective elimination of trading commissions (fees) on the international trading platform eToro. Specifically, they compare individual trading behaviors and performance:

1. Overall before and after fee removal.
2. In countries with fees removed versus countries with fees unchanged.
3. For non-leveraged long trades (fees removed) versus leveraged and short trades (fees unchanged).

Using individual trader transaction data and associated demographics from eToro during fee removals from October 9, 2018 through November 6, 2019, they find that: Keep Reading

Do brokers who accept payments for order flow (PFOF) pass this income through to customers in the form of cheaper trade execution? In his June 2022 paper entitled “Price Improvement and Payment for Order Flow: Evidence from A Randomized Controlled Trial”, Bradford Lynch compares execution quality for trading randomly selected U.S. common stocks with at least $10 million daily average dollar volume and a minimum price of $5.00 at the market at random times during normal market hours with the following three brokers:

• A broker that utilizes direct access to exchanges (Interactive Brokers).
• A broker that utilizes wholesale brokers and extensive use of PFOF (Robinhood).
• A broker that utilizes wholesale brokers and modest use of PFOF (TD Ameritrade).

He opens and closes each position the same day with holding time at least five minutes. He uses randomized order sizes representative of retail trades ($1,000 or $4,000). He measures execution quality relative to the national best bid and offer (NBBO) at the time the order is placed, with price improvement based on buys (sells) executed below the ask (above the bid), as follows: (1) proportion of trades with price improvement; (2) price improvement per share as a percent of share price; (3) effective half-spread divided by quoted half-spread; and, (4) execution speed (time between order placement and first execution). Using the specified trade and quote date for about 250 trades per broker during the 20 trading days starting May 25, 2022, he finds that: Keep Reading