PREDICTING INTRADAY S&P 500 RETURNS FOLLOWING U.S. MONETARY POLICY ANNOUNCEMENTS

Date of Award

5-2026

Thesis Subject

Dissertation/Thesis

Degree Name

Bachelor of Science (BS)

Advisor(s)

Andrey Ermolov

Abstract

This study explores the impact of monetary policy announcements by the U.S. Federal Reserve on the intraday returns of the SPY ETF – an exchange-traded fund that tracks the S&P 500 index (the 500 largest U.S. companies). This paper specifically focuses on the predictability of those intraday returns after the announcements. Using Tick Data 1-minute SPY ETF prices from February 1997 to September 2025 and regressing 1-minute price returns between 1:59 pm and 4:00 pm on monetary policy announcement days, we found that return predictability is highest 26 minutes after the announcement. This price reaction trend lasts approximately 49 minutes, creating a predictable trade window. Regressing the same 1-minute price returns on Wednesdays with no monetary policy announcements, we found weak predictability, suggesting that monetary policy announcements are a significant driver of SPY ETF returns on Wednesdays between 1:59 pm and 4:00 pm and that there are no predictable price trends on Wednesdays without such announcements. The results are consistent with earlier research showing that S&P 500 returns exhibit intraday reactions to macroeconomic announcements. However, this study expands on existing research by narrowing down SPY ETF returns predictability to a single regression model that can be used in a trading algorithm to identify potential trade opportunities following U.S. Federal Reserve monetary policy announcements.

Decades ago, trillions of dollars’ worth of securities were traded on exchange floors, where dealers shouted out bids and offers to match buyers and sellers. Today, the New York Stock Exchange floor is silent, and you can only hear the sound of working computers. With the advancement of trading algorithms, in-person trading has been replaced by computerized trading. This has led to the creation of quant hedge funds, such as Citadel, which use quantitative models in their trading algorithms to analyze multiple markets and execute orders in fractions of a second. These algorithms use models based on specific factors to determine the target price and the exact time of trade execution. With the advent of AI, electronic trading is expected to account for a higher share of Wall Street’s trading operations. It will likely contribute to massive layoffs, as banks anticipate cutting 200,000 jobs over the next five years. Understanding these algorithms and being able to add value in their quantitative models has thus become essential to Wall Street quant traders, and this is where this research comes into play. Profitable algorithms require identifying market inefficiencies created by market events that move prices, such as the U.S. Federal Reserve monetary policy announcements. These inefficiencies occur in the time between the event happening and investors and traders fully reacting to it by buying the financial asset. During this short time frame, the markets form a trend moving upward or downward, which can be taken advantage of if a model can predict it. That is exactly how quantitative hedge funds make profits. Previous papers suggest systematic predictability in equity markets around monetary policy announcements. Bomfim (2003) documents greater short-run stock market volatility, with positive surprises in the target Federal funds rate having a larger effect on volatility than negative surprises. Bernanke and Kuttner (2005) found that, on average, a hypothetical unanticipated 25-basis-point cut in the Federal funds rate target is associated with about a 1% increase in broad stock indexes. Bernanke and Kuttner (2005)

and Wongswan (2009) found that monetary policy surprises partially explain this reaction. Instead, Bernanke and Kuttner (2005) found that the impact of monetary policy announcements comes either through its effects on expected future excess returns or on expected future dividends. More specifically, tightening policies lower prices by raising the expected equity premium (due to higher risks and interest costs). These results show that financial assets react to monetary policy announcements, but do not discern a clear trend pattern that could be exploited with a trade. Additionally, despite extensive documentation of pre-announcement drifts and immediate price jumps on equity indexes, the intraday price movements in the 90 minutes following monetary policy announcements remain unexplored. Therefore, the motivation for this research is twofold. First, identifying intraday return patterns following monetary policy announcements would provide direct evidence of return predictability. Second, if such return patterns are revealed, it would suggest a time period for predictability. Even if the markets react strongly in the seconds following the monetary policy announcements, it could create a trend that lasts throughout the afternoon. Therefore, combining predictable intraday patterns with a time horizon for a predictable trend would provide sufficient information to build a quantitative model. Such a model could be incorporated into a trading algorithm that predicts price trends to enter and exit trades. In this research, we examine the impact of the U.S. Federal Reserve monetary policy announcements – referred to as FOMC announcements – on the intraday returns of the SPY ETF using 1-minute price data from February 1997 to September 2025. This paper focuses on the 1- minute returns following the 2 pm FOMC announcements and explores their predictability within one and a half hours of the announcement using multi-factor regression analyses. It identifies recurring price patterns stemming from the market’s reaction to FOMC decisions, thereby

informing the predictability of returns. Finally, it narrows the predictability of SPY returns to a single regression model that can be incorporated into a trading algorithm to determine entry and exit price points, along with a dynamic price adjustment based on real-time market data.

This document is currently not available here.

Share

COinS