The data that you find in
economic textbooks is very neat and clean but the data as it actually arrives
in the market can be anything but neat and clean. We’re talking here not only about
the imperfections of economic data gathering, which relies heavily on statistical
sampling methods, but also about how the market interprets individual data
reports.
In “Reality Check: Expectationsversus Actual”, I introduce the idea of consensus expectations as one of the keys to understanding
how markets interpret economic news and data. In my following post, we will
look at important data-reporting conventions and how they can affect market
reactions. When currencies don’t react to the headlines of a data report as you
would expect, odds are that one of the following elements is responsible, and
you need to look more closely at the report to get the true picture.
Understanding and revising data history
Economic data reports don’t
originate in a vacuum - they have a history. Another popular market adage
expressing this thought is “One report
does not make a trend.” However, that saying is mostly directed at data
reports that come in far out of line with market estimates or vastly different
from recent readings in the data series. To be sure, the market will react
strongly when data comes in surprisingly better or worse than expected, but the
sustainability of the reaction will vary greatly depending on the
circumstances. If home sales are generally slowing, for instance, does a
one-month surge in home sales indicate that the trend is over, or was it a
one-off improvement due to good weather or a short-term drop in interest rates?
When you’re looking at upcoming
economic data events, not only do you need to be aware of what’s expected, but
it also helps to know what, if any, trends are evident in the data series. The
more pronounced or lengthy the trend is, the more likely the reactions to
out-of-line economic reports will prove short lived. The more ambiguous or
fluid the recent data has been, the more likely the reaction to the new data
will be sustained.
The other important element to
keep in mind when interpreting incoming economic data is to see whether the
data from the prior period has been revised. Unfortunately, there is no rule
preventing earlier economic data from being changed. It’s just one of those odd
facts of life in financial markets that what the market thought it knew one day
(and actually traded for several weeks based on that understanding) can be
substantially changed later.
When prior-period data is
revised, the market will tend to net out the older, revised report with the
newly released data, essentially looking at the two reports together for what
they suggest about the data series. For example, if a current report comes out
worse than expected, and the prior report is revised lower as well, the two
together are likely to be interpreted as very disappointing. If a current report
comes out better than expected, but the prior period‘s revision is negative,
the positive reaction to the most recent report will tend to be restrained by
the downgrade to the earlier data.
As you can imagine, there are
many different ways and degrees in which current/revised data scenarios can
play out. A general rule is that the larger the revision to the prior report,
the more significance it will carry into the interpretation of the current
release. The key is to first be aware of prior-period revisions and to then
view them relative to the incoming data. In general, current data reports tend
to receive a higher weighting by the market if only because the data is the
freshest available, and markets are always looking ahead.
Getting to the core
A number of important economic
indicators are issued on a headline basis and a core basis. The headline reading is the complete result
of the indicator, while the core reading is the headline reading minus certain
categories or excluding certain items. Most inflation reports and measures of
consumer activity use this convention.
In the case of inflation reports,
many reporting agencies strip out or exclude highly volatile components, such
as food and energy. In the United States, for instance, the consumer price
index (CPI) is reported on a core basis excluding food and energy,
commonly cited as CPI ex-F&E.
(Whenever you see a data report “ex-something,” it‘s short for “excluding” that
something.) The rationale for ignoring those items is that they‘re prone to
market, seasonal, or weather-related disruptions. Energy prices, for example,
may spike higher on geopolitical concerns or disasters that disrupt refinery
output, like Hurricane Katrina in 2005. Food prices may change rapidly due to
drought, frost, infectious diseases, or blights. By excluding those items, the
core reading is believed to paint a more accurate picture of structural,
long-term price pressures, which is what monetary-policy makers are most
concerned with.
Looking at consumer activity
reports, the retail sales report in the United States is reported on a core
basis excluding autos (retail sales ex-autos), which are heavily influenced by
seasonal discounting and sales promotions, as well as being relatively
large-ticket items in relation to other retail expenditures. The durable goods
report is also issued on a core basis excluding transportation (durable goods
ex-transportation), which mostly reflects aircraft sales, which are also highly
variable on a month-to-month basis as well as being extremely big-ticket items
that can distort the overall data picture.
Markets tend to focus on the core
reading over the headline reading in most cases, especially where a known
preference for the core reading exists on the part of monetary-policy makers.
The result can be large discrepancies between headline data and the core
readings, such as headline retail sales falling 1 percent on a month-to-month
basis but rising 0.5 percent on a core ex-autos basis. Needless to say, market
reactions will be similarly disjointed, with an initial reaction based on the
headline reading frequently followed by a different reaction to the core.
