Are FOMC Announcements Really Informative?

Federal Open Market Committee (FOMC) of Fed (Federal Reserve Board/System) meetings which bring announcements usually followed by press conferences are one of the most important events in the rich calendar of investors’ watch lists. They are always closely watched for possible trading opportunities and are full of volatile moves on both long and short sides in fronts of all asset classes ranging from forex, bonds, and equities to nowadays even crypto markets. In our today’s summary, we will take a closer look at some implications that those kinds of financial phenomena bring.

A very fresh paper from Oliver Boguth, Vincent Grégoire, and Charles Martineau (September 2022) presents us with a nice compilation of equities returns reactions due to volatility from occasions surrounding U.S. central bank. Using an approach to quantify information flow net of price noise following FOMC announcements using unbiasedness regressions that focus on regression R2s rather than slope coefficients used by other scientists, they arrive at some interesting conclusions that we will discuss a bit further.

Authors surprisingly conclude that strength of informativeness relating to price decreases on the announcement days, and they even propose not to reject the hypothesis that FOMC announcements contain no information that is relevant for equity markets. Most price action and fluctuation in indexes and stock prices are attributed to pure noise, which is getting into equilibrium sometimes for as long as two weeks. Bonds and interest rates derivatives, Federal fund futures, or Eurodollars are better digesting often surprising data coming from Fed and their comments and do not fluctuate as much later.

Expectations of market participants are often wrong and do not reflect in price action soon after these events. It looks like there is some kind of truth in those never-aging trading wisdom adages such as “[it is] priced in” and “bad news is good news”.

Authors: Oliver Boguth, Vincent Gregoire, and Charles Martineau

Title: Noisy FOMC Returns

Link: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4131740

Abstract:

Does the substantial trade volume in equity markets around FOMC announcements reflect information flow? Using a new diagnostic to quantify information flow net of price noise, we show that equity prices following FOMC announcements are less informative about future indicative prices than those before announcements, suggesting that announcement returns are dominated by noise. The necessary reversal of the noise is obscured by our finding that price informativeness remains low for about two weeks. Our findings have important implications for existing studies of equity returns around FOMC announcements: they cannot be interpreted as an efficient market transitioning to a new macroeconomic equilibrium.

As always we present several interesting figures:

Notable quotations from the academic research paper:

“[…] the exact channel of why stock markets react so strongly to FOMC announcements is not well understood. Changes to the overnight Fed funds rate propagate to interest rates at economically relevant longer horizons, and therefore directly impact bond prices and the risk-free discount rate (Bernanke and Kuttner, 2005). In addition, FOMC announcements also reveal information about the prospect of future economic growth, which can include information about future cash flows and risk premia (see, for example, Romer and Romer, 2000, Campbell, Evans, Fisher, Justiniano, Calomiris, and Woodford, 2012, Nakamura and Steinsson, 2018). Yet, stock prices seem to be much less affected by other macroeconomic announcement that also provide information about the economy (Lucca and Moench, 2015, Ai and Bansal, 2018) and policy announcements of other central banks (Brusa, Savor, and Wilson, 2020).
Our evidence suggests an alternative explanation for the large stock price reactions: noise. On FOMC announcement days, price informativeness not only fails to increase by more than a typical day’s worth of information, but it actually decreases by nearly the same amount. In other words, even if announcements contained no incremental price-relevant information, the magnitude of noise on these days corresponds to approximately two days’ worth of average information flow. It seems surprising that noise of this magnitude, and its necessary future reversal, has eluded research to date.6 The reason for this is that noise levels remain high, and price changes largely uninformative, for about ten days after the announcement.

It is important to emphasize what we mean by “noise”. At its very core, noise is the deviation of observed prices from their efficient counterparts, or fundamental values. As such, it depends on the assumptions for fundamental values, which will vary with the research question analyzed and the choices made by the researcher. In the case of FOMC announcements, the “noise” we identify is correlated with measures of rebalancing activities and even with surprises in the Federal funds rate. As such, in any research that aims to understand the effects of price pressure or the link between Federal funds rates and equity prices this must not be considered noise. In contrast, from the perspective of many macroeconomic models, a stock price reaction that dissipates after less than 20 trading days can be considered noise.
Our main finding is illustrated in Panel A of Figure 2 for the S&P 500 using a window spanning 10 days prior to the FOMC announcement to 30 days after. Initially, as the independent variable comprises of more and more daily returns, Rt2 increases approximately linearly until just before the event day. On the announcement day, Rt2 drops by about 1.8%, which corresponds to 75% of the magnitude of the typical one-day information flow under the null hypothesis, 1/41 = 2.4% (represented by the dashed red line). Price informativeness does not increase materially over the following four days, and remains well below the diagonal even ten days after the announcement. Informativeness then increases quickly, and returns to that predicted under the null hypothesis after about 15 days.

In a related paper, Kroencke, Schmeling, and Schrimpf (2021) identify “risk shifts”, a component of monetary policy news that is captured by risky asset prices and orthogonal to risk-free rate news. They show that these risk shifts explain 27% of stock market announcement return variation, compared to only 9% explained by risk-free rate news. Interestingly, and in line with our findings, they show that risk shifts predict a price reversal following FOMC announcements.

Closely following Kroencke, Schmeling, and Schrimpf (2021), we analyze the reversal of the risk shift component as well as the residual in Figure 6. For this, we first rescale the risk shift components so that the announcement day return has unit exposure. We then regress the daily equity market returns on days h after the announcement onto the announcement day risk shift fit and the residual, and plot cumulative coefficients in Panel A. By construction, the cumulative coefficients of both the risk shift and residual component start at one, and revert back to zero over the following 12 to 16 days. Panel B shows that the difference in the cumulative coefficients is not significant, and the point estimate suggests that the residual component reverts faster that the risk shift component.

Figure 7 provides some evidence of the reversal associated with noise. The figure plots cumulative market returns following FOMC announcements conditional on the FOMC announcement returns. All returns are centered to zero on the day before the event. In particular, the blue line consists of the quintile of FOMC announcements that saw the largest stock price increases, while the red line represents the quintile of announcements with the lowest returns. While our prior analysis identified noise in individual announcement returns, we do not expect this noise to fully diversify in this non-random sorting.
Mechanically, the two lines diverge drastically at the announcement. In the high return quintile, market prices increase by nearly 2% on average, while they fall by more than 1% in the low return group. The difference in returns just after the announcement is 3%. Following these two sets of FOMC announcements, we observe that returns of the top quintile gradually decrease, and end at just over 1.5% after 30 days. In contrast, returns of the bottom quintile recover all of their initial losses. As a result, the gap narrows drastically to less than 1% after about 20 days. This is consistent with reversal of temporary price components.18


Are you looking for more strategies to read about? Sign up for our newsletter or visit our Blog or Screener.

Do you want to learn more about Quantpedia Premium service? Check how Quantpedia works, our mission and Premium pricing offer.

Do you want to learn more about Quantpedia Pro service? Check its description, watch videos, review reporting capabilities and visit our pricing offer.

Are you looking for historical data or backtesting platforms? Check our list of Algo Trading Discounts.


Or follow us on:

Facebook Group, Facebook Page, Twitter, Linkedin, Medium or Youtube

Share onRefer to a friend

Subscribe for Newsletter

Be first to know, when we publish new content


    logo
    The Encyclopedia of Quantitative Trading Strategies

    Log in

    GET NOW
    BLACK FRIDAY DEAL
    GET NOW
    BLACK FRIDAY DEAL