How Long Does the Market Reaction Last? Insights from Intraday Data.

How Long Does the Market Reaction Last? Insights from Intraday Data.

While traditional studies typically assessed market reaction to USDA using daily price changes, more recent studies have been able to zoom in using intraday data.  These studies shown in table 8 focus not just on market reaction, but overall market dynamics around USDA report releases.  For example, Lehecka, Wang, and Garcia (2014), examined corn market reaction to USDA reports over 2009-2012, when the reports were released prior to market trading hours, and showed that the strongest price reactions to the releases was found immediately after markets opened and persisted for about ten minutes.  They did not find evidence of systematic under- or over-reaction in prices.  On the other hand, they found some subtle reactions in the last trading session before the release, suggesting that traders adjusted their market exposure in anticipation of the release.

In 2012, ICE and CME commodity exchanges expanded their trading hours, allowing trades during the release of key government reports.  Kauffman (2013) showed that market volatility around WASDE report releases in 2012 has been substantially higher than in prior years and lasted longer, as shown in figure 13.  Kauffman (2013) concluded that extending trading hours resulted in brief shocks in corn futures price volatility around the release of WASDE reports that may pose a challenge for producers whose risk management strategies are affected by intraday price swings.  Typically, however, the heightened volatility has not lasted more than 30 to 60 minutes, thus it should not affect long-term risk management positions.  On the other hand, Wang, Garcia, and Irwin (2014) argued that introduction of electronic trading has significantly reduced bid-ask-spreads thus reducing order execution costs. However, these bid-ask-spreads were substantially larger during the index trader roll periods and on USDA report release days. The authors argue that the evidence of larger spreads during the roll periods point to a sunshine trading effect, with added liquidity entering the market in anticipation of predictable roll behavior, while the USDA announcement effects identify the importance of unexpected information and adverse selection on order execution costs.

In January 2013, the USDA has moved its report publication time to 11 am central to allow markets the best change to absorb news during trading hours.  Using intraday data over 2009-2014, Adjemian and Irwin (2018) compared market reaction to USDA reports across two regimes.  The authors found that “when agricultural futures markets are permitted to discover prices freely in response to USDA reports the adjustment process is not instantaneous, as continuously-traded futures markets experience heightened volatility and trading volume in response to news relative to what was observed during the era of trading halts. Moreover, markets appear to now have a more difficult time distinguishing between the newsworthiness of USDA reports, at least in the very short-run, but these differences persist only for a handful of trading minutes. After that, announcement shocks at major agricultural markets resemble those that were observed when (trading) timeouts were in effect.” (p. 1169)

Another change to the USDA report release rules was implemented in July 2018 when USDA decided to stop sharing reports with media members inside the lockup area ahead of the official publication time to prevent an unfair advantage their customers may have relative to the general public within the two seconds following the report release due to the proximity of their servers to USDA.  Adjemian and Irwin (2020) examined the impact of this change on the market reaction and did not find evidence that removing media members from lockup has increased announcement time price volatility using either daily or intraday data.

Several studies explored market dynamics around report releases using high frequency intraday data.  For example, Shang, Mallory, and Garcia (2018) estimate that from 2008 to 2011 government news tended to increase the market bid-ask spread (BAS) for about an hour (through the channel of compensation for adverse selection), although average liquidity costs did not generally rise notably on USDA announcements days. Focusing on 2013–2016 data, after continuous trading of crop news began, Fernandez-Perez et al. (2019) documented that the BAS rises just preceding USDA announcement time and declines gradually over roughly the next twenty trading minutes. These studies decompose the BAS to draw inference about the path of information asymmetry and liquidity provision during announcement days. Couleau, Serra, and Garcia (2020) find that announcement days in these markets are characterized by price jumps clustered around the report release—consistent with the arrival of new information and that eliminating the timeout led to more price jumps around the announcement, and wider bid-ask spreads.

Overall, these studies confirm the findings of the previous studies using daily data and provide a more detailed look at the market reaction and dynamics around report releases by zooming in on a high frequency level.  The evidence of the increased market volatility, price jumps and higher BAS following the release of USDA reports, indicates that these reports contain valuable information that helps markets find a new equilibrium price.