Summary
This report combines the market-neutral portfolio that we previously constructed which uses all 3 entry points of the REACTION signals (t0 open + 60min, t0 close, t1 open) with a filter based on large moves in SPY preceding the signal (which was introduced in our past research). The filtered portfolio outperforms the unfiltered portfolio with an average annual return of 13.7% (on constant capital of $50M per entry point; $150M in total), a Sharpe ratio of 2.63, a maximum drawdown of -3.2%, and a correlation to SPY of 0.08 between 2016-2021H1. These results are inclusive of conservative trading cost assumptions and robust to stock liquidity.
Read as PDF
Introduction
In a previous report, we found that large positive moves in SPY prior to a REACTION signal are an indication to avoid short signals and we found that large negative moves in SPY prior to a REACTION signal are an indication to avoid long signals. In this report, we apply this finding to the combined market-neutral portfolio we previously created that uses all 3 signal entry points (t0 open + 60 min, t0 close, t1 open). Specifically, we screen out short signals that are preceded by a large positive move in SPY and screen out long signals that are preceded by a large negative move in SPY. The motivation of this research is to determine if applying this filter to the REACTION signals improves portfolio performance.
We proceed by reviewing the portfolio construction steps, presenting the results of a portfolio using all three entry points that is screened for signals preceded by large moves in SPY, and discussing the findings.
Portfolio Construction with SPY Screen
Leapday’s REACTION signals are available for 3 entry points after an event day, defined as the first day the market is open subsequent to a company’s earnings announcement. In previous reports, we built market-neutral portfolios which traded the strongest one-third of the 3d hold signals (ordered by signal magnitude) for each entry point: (1) 60 minutes after the open on the event day (t0 open + 60min); (2) just before the market close on the event day (t0 close); (3) before the market open on the next trading day after the event day (t1 open). The specific assumptions used in our portfolio construction for every entry point can be found in each individual report.
The t1 open report also included a construction of a market-neutral portfolio that combined all 3 entry points. Because the return streams of each entry point are relatively uncorrelated, the combined portfolio produced superior performance with metrics that were stronger than any individual entry point or any pair of entry points.
In this report, we take the combined portfolio and apply the findings from previous research which indicated that large positive moves in SPY prior to a signal are an indication to avoid short signals and large negative moves in SPY prior to a signal are an indication to avoid long signals. The assumptions for each individual entry point are identical to the previous three reports referenced at the beginning of this section and the only difference is the following rules are applied to screen out certain signals:
t0 open + 60min signals
If the return of SPY between the t-1 close and t0 open is greater than or equal to 0.50%, do not trade short signals.
If the return of SPY between the t-1 close and t0 open is less than or equal to -0.50%, do not trade long signals.
t0 close signals
If the return of SPY between the t-1 close and t0 close is greater than or equal to 0.50%, do not trade short signals (this introduces a minor lookahead bias since the t0 close is released 10 minutes before the close).
If the return of SPY between the t-1 close and t0 close is less than or equal to -0.50%, do not trade long signals (this introduces a minor lookahead bias since the t0 close is released 10 minutes before the close).
t1 open signals
If the return of SPY between the t0 close and t1 open is greater than or equal to 0.50%, do not trade short signals.
If the return of SPY between the t0 close and t1 open is less than or equal to -0.50%, do not trade long signals.
Results
Below we present results for portfolios that trade the strongest one-third of signals for each of the three entry points, using the 3d hold for each entry point. One portfolio is set up without a market direction filter applied while the other uses a market direction filter as specified in the previous section. To recap, the market directional filter screens out long signals preceded by SPY returns less than or equal to -0.5% and screens out short signals preceded by SPY returns greater than or equal to +0.5%. Both portfolios begin with a starting capital of $150M.
Metrics associated with the chart above are presented below in Tables 1 - 4.
Before screening out signals based on moves in SPY preceding the signal, a portfolio (with a starting capital of $150M) that trades all 3 REACTION entry points has an average annual return of 12.3%, a Sharpe ratio of 2.18, and a max drawdown of -3.2% between 2016 and 2021H1. After applying the market direction signal, portfolio performance is improved to an average annual return of 13.7%, a Sharpe ratio of 2.63, and a max drawdown of -3.2%. These results include conservative trading cost assumptions. A review of the yearly results shows that 2018 was the only year in which the market direction signal did not improve performance on a risk-adjusted basis. These results highlight the potential of extracting additional value out of the REACTION signals beyond simple standalone strategies.
Discussion
We have demonstrated that building a market-neutral portfolio that combines all three entry points of Leapday’s REACTION signals produces strong annual returns, high Sharpe ratios, and small drawdowns with results that are robust to trading costs and various market regimes. We have also shown that screening out short signals that are preceded by a large positive return in SPY and screening out long signals that are preceded by a large negative return in SPY boosts performance even further. We believe this incremental improvement in results exists because investors overreact to company specific news that is contrary to the prevailing market sentiment on that day.
This report further highlights the potential for investors to boost performance beyond the simple standalone strategies presented in our previous reports. In this report, we boosted returns by filtering out signals preceded by big moves in SPY in the opposite direction of the signals. A natural extension of this filter is to overweight trades preceded by big moves in SPY in the same direction of the signal, as our previous research indicated that those signals outperform. Similarly, researchers may filter or overweight trades based on other signals and market moving information they monitor. In addition, the entry and exit rules used in this portfolio are simple. The portfolio performance can likely be improved by optimizing the holding period to incorporate additional information gauged from the onset of the trade until the optimal exit.
Conclusion
In this report we begin with a portfolio that combines the t0 open + 60min 3d signals, the t0 close 3d signals, and the t1 open 3d signals and trades the top one third of each entry point’s signals based on signal magnitude and holds those trades for 3 days using constant capital of $50M per entry point ($150M in total). This portfolio has an average annual return of 12.3% between 2016-2021H1, a Sharpe ratio of 2.18, a maximum drawdown of -3.2%, and correlation to SPY of 0.07.
Next, we apply a market direction filter which was explored in a prior research note. This filter uses large positive moves in SPY prior to a REACTION signal as an indication to avoid short signals and large negative moves in SPY prior to a REACTION signal as an indication to avoid long signals. By applying this filter, portfolio performance improves to an average annual return of 13.7%, a Sharpe ratio of 2.63, a maximum drawdown of -3.2%, and a correlation to SPY of 0.08. Like our previous reports, these metrics are conservative given the trade cost assumptions, portfolio position sizing constraints that we use, and the cyclical nature of earnings events which allows investors to deploy capital elsewhere during periods of low earnings activity.
This report highlights the opportunities for quantitative researchers and portfolio managers to combine the Leapday REACTION signals with additional signals, filters and data to create unique portfolios with performance metrics that are improved beyond the standalone strategies presented in previous reports.
Disclaimer
This material is solely for informational purposes and is not an offer or solicitation for the purchase or sale of any security, nor is it to be construed as legal or tax advice. References to securities and strategies are for illustrative purposes only and do not constitute buy or sell recommendations. The information in this report should not be used as the basis for any investment decisions. We make no representation or warranty as to the accuracy or completeness of the information contained in this report, including third-party data sources. The views expressed are as of the publication date and subject to change at any time. Hypothetical performance has many significant limitations and no representation is being made that such performance is achievable in the future. Past performance is no guarantee of future performance.