“Grade the Analysts” resurrected – Joseph Spak leads the pack

“Grade the Analysts” was a favorite monthly story during my intermezzo at Thetruthaboutcars.com. Well, at least, it was a favorite of the analysts that competed for who has the sharpest crystal ball  in predicting car sales. There even was occasional drama and intrigue. No wonder: Knowing the consistently right-on analyst can mean real money when betting on the economy.

Now at the Dailykanban, Grade the Analysts is being resurrected. My former Tokyo cohort Craig Trudell volunteered to supply Bloomberg data (domo arigatou!) and I crunch them against sales and SAAR reports supplied by Motorintelligence.

In January, Joseph Spak of RBC takes the top spot, coming within 8% of the true results of the Detroit 3 and the overall SAAR. Jessica Caldwell of Edmunds, a multiple champion in years back, comes in 3rd.

Overall, the analysts were a bit too exuberant. Detroit did not do as expected. The  Seasonally Adjusted Average Selling Rate (SAAR)  was called more or less correctly as transplants made up for the domestics.

The algorithm is the same as we used back at TTAC. Basically, it measures how much the forecasts deviate from the actuals. A few analysts supply only incomplete (“NA”) data. They get punished with a 100% deviation from the actual for the missing data point. The algorithm was reviewed by a few analysts back when, and I see no reason to change it.

Here is where I could use a little help. In addition to the Detroit 3, a few analysts also predict the monthly performance of the transplants. I think this should receive points, but I have not figured out how. If you have suggestions, please send a message.  Suggestions from the analysts will receive preference.

I will try to do this each month, with the aim on crowning the Grade A champion analyst when the year is over. If you want to win, analysts, hand in complete predictions.

This site automatically detects and reports abuse