A two-decade empirical study shows that bond and equity market prices are more likely to “jump” on days with U.S. economic data releases. In particular, surprises in news announcements tend to lead to higher volatility and larger price moves. The impact of key data surprises on bond markets seems clearer and simpler. The impact on equity markets depends on the state of the business cycle

Huang, Xin (2015). “Macroeconomic News Announcements, Systemic Risk, Financial Market Volatility and Jumps,” Finance and Economics Discussion Series 2015-097. Washington: Board of Governors of the Federal Reserve System,

The below are excerpts from the paper. Headings, links and cursive text have been added.

News and effects

“This paper studies financial market volatility and jump responses to macroeconomic news announcements [in the U.S.]…based on two decades of high-frequency data…As market responses to news are mostly short-lived, we need to go into intraday and detect these responses using high-frequency data… This paper studies both the equity and the bond markets…The corresponding market responses data are five-minute returns on S&P 500 index futures and 30-year U.S. Treasury bond futures.”

“Twenty-six [U.S.] macroeconomic news announcements are included in this study.”

“In this paper, jump is defined in the formal statistical sense, instead of price movements over a certain threshold. In nature, both jumps and volatility are unobservable. [A statistical method]…makes it possible to separate discrete jumps from continuous volatility…We can turn the latent jumps and volatility processes into observable time series, and study their dynamics during announcement days.”

The importance of news days

“Given each and all the news announcements, we can compute the proportion of the jump days… Around two-third of the days are news days.”

“The proportions of jump days in many news announcement days are statistically significantly larger than those in no-news days for both the equity and bond markets…especially non-farm payrolls, PPI, CPI, retail sales, initial unemployment claims, consumer credit and business inventories. “

“Non-farm payrolls turns out to be the most influential news across different markets…Non-farm payrolls affect the equity and bond markets in different fashions: it moves the equity market mostly by discrete jumps while it moves the bond market mostly by continuous volatility… Retail sales appears to be the second most influential news, as some of its related variables are significant in affecting the volatility and jumps of the equity market, and the volatility of the bond market.”

“The bond market exhibits more sensitivity to news announcements than the equity market does.”

The impact of data surprises

“Not all news announcements bring new information to markets. Markets should respond not to the expected announcements, but to the surprising component…[One can use] a standardized measure [actual release minus median forecast, divided by standard deviation of surprise] to convert the raw released value into a news surprise value.”

“To uncover the market’s expectation for a future announcement value is to use survey data, such as those available from Money Market Services (MMS). The mean or median forecast is then compared to the released value to construct a measure of forecast error or news surprise…Surprises are separated into positive and negative ones.”

“Surprises in news announcements…jointly significantly affect the continuous volatility of the equity market. Positive surprises from survey data also jointly significantly affect both the volatility and jumps of the bond market.”

“Moreover, non-farm payrolls news surprises, including positive and negative surprises, are also significant in affecting bond market jumps, even though its disagreement and other explanatory variables are not.”

The influence of the business cycle

“Specific to the equity prices, there are two competing factors — cash flow and discount rate. The relative effects of a given news announcement on these two factors change over business cycles…[Previous research found] that higher industrial production and lower unemployment rate drive stock price down in a strong economy and up in a weak economy, because in a strong economy the discount rate increases more than the cash flow does, and the reverse happens in a week economy.”

“The data for business cycles are readily available from the NBER website.”

“When the full sample is separated into expansion and contraction subsamples the bond market remains significantly influenced by most variables in the two subsamples, while the equity market is less so…[For example] non-farm payroll surprises are insignificant in explaining equity market volatility in the full sample. But its positive surprise significantly increases the equity volatility in the expansion subsample.”

The influence of the zero lower bound

“The period of zero lower bound (ZLB)…starts from December 16, 2008.”

“For the equity market, ZLB does not appear to constrain responses to news, in terms of either volatility or jumps. In contrast, ZLB does constrain the bond market volatility and jump responses to news. For example, when all news are considered jointly…there are not as many significant variables in the ZLB subsample.”