Overnight Trading On The S&P 500: Returns And Drawdowns In SPY

When discussing the stock market, most people tend to focus on what happens during regular Wall Street trading hours. After all, that is when most trading takes place.

Yet one often overlooked aspect is that prices can also change between one session’s close and the next session’s open. How much do these moves actually contribute to overall returns?

In this article, we will explore that question by comparing the historical performance of the regular trading session, also known as the cash session, with overnight performance in SPY, the ETF that tracks the S&P 500 Index. We will then break down overnight results by day of the week and examine how profits and drawdowns change when we limit trading to Monday, Tuesday, and Wednesday nights.

Why Use the SPY ETF to Analyze the S&P 500?

For our analysis, we will use SPY, one of the world’s most liquid ETFs. Because it tracks the S&P 500 Index, it provides a useful reference for studying the behavior of the U.S. stock market.

We chose this instrument for a reason. Rather than using instruments commonly employed in systematic trading, such as S&P 500 futures, we are focusing on an ETF, which has a simpler structure and does not require us to manage contract expirations. Futures are more complex instruments, and rolling from one contract to the next can introduce price discontinuities known as rollover gaps. If not handled properly, these gaps can distort the results of the analysis.

The goal of this article is less about developing a trading system ready for live trading and more about analyzing market behavior and identifying potential recurring patterns.

Each trader can then assess whether and how to capitalize on any edge identified in the analysis, choosing the instrument that best suits their needs while considering factors such as accessibility, trading costs, and capital efficiency.

Cash vs. Overnight Sessions: Comparing Returns in SPY

Let’s begin with Figure 1, which compares the performance of two strategies: one that trades exclusively during the cash session and another that holds positions only overnight.

By cash session, we mean the period from 9:30 a.m. to 4:00 p.m. Eastern Time. The strategy buys $10,000 worth of SPY at the market open and closes the position at the end of the same session, holding no position overnight.

The overnight session, by contrast, runs from one trading session’s close to the next session’s open. In this case, the strategy buys SPY at 4:00 p.m. and sells it at 9:30 a.m. on the next trading day, capturing price changes between the two sessions, including any opening gaps.

Figure 1. Comparison of cash-session and overnight trading results in the SPY ETF.

At first glance, the results appear clear: the overnight session substantially outperformed the cash session, generating a total profit of approximately $15,000, compared with roughly $5,900 during regular trading hours.

A closer look at the two curves, however, reveals two quite different periods.

Before 2010, the gap between the two equity curves gradually widened. In other words, while the overnight session contributed positively to market gains, the cash session delivered much weaker results, at times accumulating substantial losses.

From 2010 onward, that difference appears less pronounced. Both sessions contributed positively to stock market gains, and the cash session performed considerably better than it had during the previous decade. Another interesting observation emerges during two specific downturns: the pandemic-related crash in 2020 and the market decline in 2022. In both cases, the cash session held up relatively better than the overnight component, which experienced sharper declines.

Based on the overall results, however, the overnight session offers the more promising starting point for further analysis. Let’s focus exclusively on this component and examine how its returns are distributed across the days of the week.

The goal is to determine whether the observed edge is spread evenly across the week or tends to be concentrated on particular days.

Overnight Trading in SPY: Results by Day of the Week

In Figure 2, we have broken down the overnight strategy’s results by day of the week, producing five separate equity curves.

The strategy works as described above, except that each curve includes only trades initiated on a specific day. For example, the Monday curve shows the results of buying SPY at Monday’s close and selling it at Tuesday’s open. Similarly, the Friday curve includes trades opened at Friday’s close and closed at the following Monday’s open, covering the weekend as well.

The results reveal clear differences across the days of the week. Monday’s overnight session stands out as the most profitable historically, generating a total profit of more than $7,300, well above the results for any other day.

The Tuesday, Wednesday, and Friday overnight sessions also produced positive overall results, although their profits were considerably smaller, at approximately $3,400, $2,500, and $2,700, respectively.

Thursday is the clear exception: it is the only day of the week with a negative overall result, recording a loss of approximately $900.

These results show that the edge observed in the overnight session is not evenly distributed throughout the week but varies by day.

At this point, a natural question arises: what would happen if we traded only on the days that delivered the most promising results?

For example, we could limit trading to the Monday, Tuesday, and Wednesday overnight sessions. We would exclude Thursday because of its negative results, and Friday, despite its positive performance, to avoid the additional risks associated with holding a position over the weekend.

Figure 2. Cumulative results of the overnight strategy in SPY, broken down by day of the week.

Monday Through Wednesday Overnight Trading: Effects on Profits and Drawdowns

In Figure 3, we compare the original overnight strategy with a more selective version that trades only on Monday, Tuesday, and Wednesday nights, excluding the Thursday and Friday sessions.

The results are particularly interesting. Despite taking fewer trades, the modified strategy retains most of the original strategy’s total profit, with approximately $13,300 versus $15,000.

The most significant improvement, however, concerns risk. Looking at the two equity curves, we can see that the more selective version follows a much smoother path, with substantially smaller drawdowns.

This is reflected in the ratio of net profit to maximum drawdown, which rises from 3.7 for the original strategy to 9.6 for the modified version, more than two and a half times the original value.

In other words, trading less frequently allows us to achieve nearly the same profit with considerably less risk exposure. This result highlights the importance of identifying a statistical edge and understanding when to put it to work.

Figure 3. Comparison of the overnight strategy in SPY without a day-of-week filter and the version with entries limited to Monday through Wednesday.

The Overnight Risk Premium: One Possible Explanation for the Results

Our analysis shows that a substantial share of historical U.S. stock market returns has occurred during the overnight session, with marked differences across the days of the week.

But why does this happen? One possible explanation is the so-called overnight risk premium. Investors who hold positions overnight remain exposed to economic news, geopolitical events, and potential opening gaps, without always being able to respond promptly. The positive returns observed could therefore represent compensation for taking on this additional risk.

If that were the main explanation, however, we might also expect particularly strong results from the Friday overnight session, which spans the entire weekend and involves longer exposure to unexpected events. Instead, although this session was profitable, its performance fell well short of the Monday overnight session.

The risk premium therefore remains one possible interpretation, but it is unlikely to explain all the differences observed, particularly those between individual days of the week.

What we can say, however, is that simply filtering trades by day of the week preserved most of the historical profits while significantly reducing drawdown.

Once again, we have seen how even a straightforward analysis of market behavior can provide a useful starting point for identifying new systematic trading opportunities.

Until next time, happy trading!

Benzinga Disclaimer: This article is from an unpaid external contributor. It does not represent Benzinga’s reporting and has not been edited for content or accuracy.