Posts On Single-Stock ETFs and Modeling Buy-Sell Decisions

My personal blog, Security Trading Analytics, aims to empower visitors who seek examples and demonstrations of quantitative methods for tracking, analyzing, and projecting security prices.   Another blog goal is to present methods and resources that are practical and useful for individuals and small consulting practices that wish to improve their trading, investing, and analysis skills via quant methods.  Additionally, my blog does not shy away from disruptive stances, such as for single-stock leveraged ETFs.

This post highlights a recent Security Trading Analytics post titled “Should I Trade Single-Stock Leveraged ETFs?”.  You can learn here about the origins of single-stock ETFs for public trading in the United States.  Around the date that single-stock ETFs initially became available for public trading in July 2022, the Securities and Exchange Commission (SEC) Office of Investor Education and Advocacy issued a statement highlighting risks associated with these instruments.  While the statement identified some potential single-stock ETF trading risks, it did not present empirical evidence supporting those risks or workarounds to mitigate them.  In contrast, the Security Trading Analytics blog published three posts on single-stock ETFs utilizing backtesting and charting techniques to do exactly that.

Some Key Findings from a Highlighted Prior Post

The following table from the “Should I Trade Single-Stock Leveraged ETFs?” post shows ticker pairs sorted by summed ETF percentage change.

• Outperformance Ratio: For eleven of the seventeen ticker pairs, the summed percentage change was larger for the leveraged ETF ticker than for its matching underlying ticker.
• Skewed Wins: Beyond that, the difference in summed percentage change values was much larger on average for the ticker pairs where the ETF won, compared to the pairs where the underlying security held the upper hand.
• Win Metrics: Ultimately, the leveraged ETF tickers outperformed the underlying tickers both in the sheer number of winning cycles and in the average size of those winning trades.

Some readers might be wondering: what is the summed percentage change metric?  It is simply the sum over the individual trades for a ticker.  Each ticker pair has two summary percentage change values — one for the leveraged ETF ticker and another for the underlying ticker for an ETF.

For the historical price data summarized in the “Should I Trade Single-Stock Leveraged ETFs?” post, there is a total of 217 trades for the full set of seventeen ticker pairs displayed in the first two columns of the preceding table.  Start and end dates for individual trades were based on the POGL model, which is described later in this post.

Dollar change values across trades were analyzed with fixed wagers of $10, $100, and $1000 per trade.  The following screenshot contains summarized dollar change results for both single-stock ETF securities and matching underlying securities across all 217 trades.  The screenshot reveals two key findings.

• The collection of seventeen single-stock ETFs consistently outperformed their corresponding underlying securities by 43.38 percent no matter what the size of the different wager amounts.
• The dollar size of the change was directly related to the size of the wager.  For example, the single-stock ETF trades increased by a factor of ten whenever the size of the wager increased by a factor of ten.  This relationship is also for underlying securities.

Two Predecessor Posts to the “Should I Trade Single-Stock Leveraged ETFs?” Post

This blog’s initial single-stock leveraged ETF post (“What are Single-Stock ETFs and Should I Invest in Them?”) introduced the asset class for investing as opposed to trading.  The post compared ten single-stock leveraged ETFs to their matching underlying securities.

The comparison metric was the percentage change from a leveraged ETF inception date through to January 31, 2025, the last date for which historical price data was collected for the post.  These comparisons were for a simple buy-and-hold percentage change metric.  It was common for different ETFs to have different start dates.  As a result, the number of trading days on which percentage change values were evaluated and compared was different across the ten ticker pairs.  The percentage change values were based on as few as 54 trading days through a maximum of 623 trading days.  This post was conducted at a relatively early date in the history of single-stock ETFs.

The ten underlying securities had an average percentage change value of 80.65%.  The comparable average percentage change value for the ten single-stock ETFs was 109.78%.  These backtested outcomes indicate investors achieved about a 29% advantage with the single-stock ETFs relative to their matching underlying securities.
Digging deeper in the results, the percentage change for the top two single-stock ETFs (PTIR and MSTU) substantially exceeded the percentage change for any other leveraged or underlying security.  This outcome confirms that the potential single-stock ETF advantage could be substantially greater than 29% if there was a way of anticipating when a single-stock leveraged ETF was likely to return exceptional gains.

A subsequent post (“A Model to Lock-in Gains When Trading Single-stock ETFs”) compared four single-stock ETFs to their underlying securities.  The comparison metric in this case was not for a buy-and-hold scenario.  Instead, the most important metric in this case was the average percentage change across a set of trades where the POGL buy-sell model chose the start and end date for each trade.  The start dates were chosen during times when above average returns were likely, and the end dates were chosen to lock-in a substantial portion of any accumulated gains during a trade.

An Overview of the POGL Model

The Proper Order and Gain Lock-in (POGL) model was used to choose start and end dates for trades in both “A Model to Lock-in Gains When Trading Single-stock ETFs” and “Should I Trade Single-Stock Leveraged ETFs?” posts.  This model was initially introduced in the “A Preliminary Analysis of EMA Period Lengths and Price Action in a Buy-Sell Model” post to evaluate the feasibility of choosing start and end dates for trades.

Some key points for the POGL model include the following.

• The POGL Model assesses when to buy a security based on the commencement of a proper order relationship between specific Exponential Moving Average (EMA) values and their underlying closing prices (where the closing price > fast EMA > slow EMA).  The model is based on the assumption that this proper order relationship allows the identification of generally rising price windows.
• The model logic also dictates when to exit a position by attempting to lock in any accumulated gains before price pullbacks substantially diminish those gains or even cause a loss.
• These entry and exit strategies were backtested for individual stocks as well as matched pairs of single-stock leveraged ETFs and their underlying securities.

At the time a security is bought, the POGL model sets an upper price boundary (greater than the buy price) and a lower price boundary (below the buy price).  If prices continue to rise after a security is bought, the closing price will eventually exceed the upper price boundary, prompting the model to dynamically reset both the upper and lower boundaries.  This process repeats as long as prices rise past each new upper price boundary.  In this way, the model hitches a ride on a rising price trend for as long as it exists.  By using closing prices instead of EMA value sets for boundary values, the model gets you out of trades in a quick and unemotional way that reflects current market action.

When the closing price ultimately falls below the lower price boundary, the security is sold at the opening price on the next trading day.  On the day after a sell date, the proper order conditions for buying a security take effect again.

Next Steps

There are at least four next steps as a follow-up to this review.

• Perform parametric studies illustrating frameworks for seeking optimal EMA value sets as well as the best-performing upper and lower price boundaries  for the POGL model.
• Evaluate the POGL model for different asset classes when different market conditions prevail to assess the sensitivity of the model to various asset classes and market conditions.
• Tweak the underlying POGL model assumptions about when to buy and sell securities to improve the profitability of discovered trades.
• Package the historical data acquisition and processing steps so evolving versions of the POGL model become available to a wide audience of potential users.

Your comments are invited.