New Model for Economic Consensus Estimates


News agencies and specialized newsletters have traditionally surveyed a limited number of economists to ascertain the market consensus for various economic indicators.  However last week, crowd sourcing data provider, Estimize added Economic Indicators to its platform in order to obtain a more comprehensive picture of what the market expects will be reported for regular government economic releases.

Value of Consensus Estimates

Determining the “consensus estimates” for corporate earnings announcements, FDA drug trial results, political elections, or government economic releases has historically been an important part of the investment research process as this has helped investors determine what “the market” has discounted into existing securities prices.

Consequently, once an investor knows what the market consensus is for a particular event (like a corporate earnings release), then he/she has a good idea how investors will respond if the actual earnings release for that stock is better or worse than that market expectation.  A better than expected earnings report will generally lead to an improvement in the price of that particular stock while a worse than expected earnings report will lead to a decline in the price of that stock.

Of course, it is important to remember that different types of assets like stocks, bonds, or commodities will respond differently to these expectations surprises as misses will have different implications for the specific security.

In addition, some investors use consensus estimates to help then position trades before the announcement of a specific event – particularly if they expect an outcome that is markedly different than what the consensus view.

Traditional Way to Discover Market Consensus

In the past, various news agencies or newsletter providers have determined the consensus estimate for any formal event or release by periodically surveying a limited number of “experts” like sell-side analysts or economists and calculating the mean or median estimate of that survey.

This practice has been well established for most of the data reported in quarterly corporate earnings announcements.  In fact, over the years a number of consensus earnings databases and calendars have been established enabling investors to get a good handle on current market expectations and how these have changed over time.

However, consensus estimates for monthly and quarterly government economic releases have not been tracked as rigorously.  Of course, most major newswires including the Wall Street Journal, Thomson Reuters, Bloomberg and CBS MarketWatch all conduct regular surveys of the major government economic releases a week or two before the release date.

Typically, these surveys include the estimates of a few dozen well-known Wall Street economists at that point in time.  Unfortunately, none of these surveys track how these forecaster’s estimates have changed over time.  This is important because most economists revise their forecasts as they see evidence from other data releases about actual economic conditions.  For example, many economists might publish a preliminary forecast for monthly retail sales, which they later revise after auto sales, same store sales, and other private sales data is made available.

Of course, there are a few more formal surveys of government economic releases that are commercially available to customers, including Consensus Economics, Focus Economics, and Blue Chip Economic Indicators.  These products typically focus on longer-term estimates like quarterly GDP, but they do collect estimates for monthly time series like retail sales, housing starts and unemployment.  The good news about these consensus surveys is they generally include estimates from a much larger universe of economists.  However, the frequency of these surveys (at best monthly) means the estimates included might be preliminary and not take advantage of the most up-to-date data.

New Estimize Offering

Strangely, very little has changed in the way investors have collected economic consensus estimates for decades, that is until last week when Estimize announced that they were adding economic indicators to their crowd-sourcing financial data platform.

The new service will include market estimates on 17 different US economic indicators, including:

  • Nonfarm Payrolls
  • Unemployment Rate
  • Average Weekly Earnings
  • GDP
  • Producer Price Index
  • Consumer Price Index
  • Retail Sales
  • Purchasing Manager’s Index
  • Existing Home Sales
  • New Single Family Home Sales
  • Case-Shiller Home Price Index
  • Housing Starts
  • Durable Goods Orders
  • ISM Non-Manufacturing Index
  • Initial Claims for Unemployment Insurance
  • Manufacturing New Orders
  • Industrial Production


Estimize expects to add other US indicators and international economic time series as this service gains traction.

Estimize CEO, Leigh Drogen explained the reason for adding this new service, saying, “The currently available Economic Indicator estimate data sets are extremely poor, to put it lightly. They are literally a poll, a few weeks before the report, from 30 economists, with no revision history, and no forward looking estimates. One has to wonder how this has been the case for so long.  We’re going to change all of that now, by using the philosophies that have made Estimize so successful to this point. We believe a more accurate, and more importantly, more representative data set will be produced, with greater depth of information by collecting data the way we believe it should be collected.”


Investors have had to put up with getting consensus estimates for economic indicators in the same old way for decades – by conducting a point in time survey of a limited number of Wall Street economists.  However, last week’s announcement by Estimize that it was going to start collecting forecasts on US economic indicators by leveraging the “wisdom of the crowds” could completely change this practice.

It will be extremely interesting to see if Estimize can attract the same volume of forecasts that it currently gets for corporate earnings, and whether this consensus can be as accurate in aggregate as its earnings data has been.  However, there is no reason to believe that the consensus estimate for economic indicators will be any less accurate than what they are currently finding for corporate earnings.  In fact, given the real-time nature of the Estimize platform, we have every reason to believe that the Estimize economic consensus might be even more accurate than its earnings data as forecasters will be able to include estimates with the most recent data releases.

If this proves to be true, then we don’t expect too many serious investors will continue relying on obtaining their consensus economic indicator estimates from traditional sources in the future.  It will be exciting to see how this all pans out.



About Author

Leave A Reply