Thresholds, Tipping Points, Good & Bad News

It is not news that news moves financial markets. This blog will publish research on how, when, why, and which news moves what financial markets.

Traders often will set threshold for action and in aggregate this might explain why some announcements are a “scratch” from a trader’s perspective that is there is a move but not enough to act upon. Once the threshold is breached though, lots of people want in.

The move from disorder to order or the contagion effect has been used to explain financial crises. Perhaps it can also be used to explain markets reactions to news too.

Looking at the impact of the Non-Farm Payrolls (“NFP”) announcements in the U.S. on the Euro (EURUSD) exchange rate, a couple of thresholds appear to be important. Interestingly the threshold is larger for bad news than for good news.

Good news and bad news have different effects also the skew of the expectations is important. These effects have been discussed before but if the news is very good or exceptionally bad there is an extra kicker. So, negative news is more likely to be a non-event than good news from a trading perspective.

The thresholds that were found to be statistically significant were when news was greater than one standard deviation and when news was less than two standard deviations from what was expected (as measured by the economic derivatives auction for NFP).

Does Bad News Matter More Than Good?

It is not news that news moves financial markets. This blog will publish research on how, when, why, and which news moves what financial markets.

Bad news tends to have a bigger impact than good. Running a regression of the returns from futures on the S&P 500 index for 5 minutes before to 25 minutes after the data release on the derivatives-based non-farm payroll news gives:

S&P 500 Index	News Coefficient	Standard Error	t-Statistic	R2
Derivatives-Based News	0.00311	0.00082	3.78	0.3091

The same intraday regression as above is run but splitting the news effect into two (one when the news is a positive surprise and the other when the released number is less than expected):

S&P 500 Index	News Coefficient	Standard Error	t-Statistic	R2
Negative News	0.00331	0.00110	3.01	0.3108
Positive News	0.00285	0.00128	2.22

There is a slight overall improvement in fit (although the standard error increases and adjusted R² falls). Depending on the application of the results the difference in the estimated average effect for positive and negative of 0.00311 and the 0.00331 for good and 0.00285 for bad may be enough to justify the differentiation.