Since January 2008, we've offered 47 posts in which we've made 52 total predictions. Using the plus-minus statistic from hockey and basketball, where we gain a point if we're right and lose a point if we're wrong, and score a zero for when the outcome of a prediction in question is uncertain, we have now upped our plus-minus score from our most recent score three months ago by 1 point to +18.

That's pretty remarkable, given that most of our predictions involve forecasting the level of the S&P 500.

The table below provides our track record since 16 July 2009, including any updated results for older predictions:

Political Calculations' Plus-Minus Score Update, 15 October 2009
Date	Prediction	Outcome	+/- Score
1 July 2009	We forecast a range for the S&P 500 in July 2009 between 925 and 945.	The month's not over, so the jury's still out. The month is long over, and we nailed it! The S&P 500 averaged 935.82 in July 2009.	+1
27 July 2009	We jumped the gun before getting all the data that July 2009 had to offer and preliminarily forecast the S&P 500 to average somewhere between 991 and 1007 in August 2009.	August 2009 came in with an average value for the S&P 500 of 1009.74. We hate to score a miss, but let's face it, we were off by 0.2%! We'll once again leave it to our readers to judge for themselves how a near-miss should be used to evaluate our forecasting chops.	-1
3 August 2009	Now armed with all of July 2009's data, we up our forecast average value of the S&P 500 in August 2009 would be between 1005 and 1020.	Well, what a difference a few extra days of data makes!...	+1
13 August 2009	We make fun of 47 economists for predicting that the U.S. recession would end in the third quarter, based on the "Cash for Clunkers" program. We point out that the dividend futures data for the S&P 500 has been saying the recession would be over in 2009Q3 for months, long before C4C even became legislation!	Too soon to tell. This prediction looks pretty likely, but we'll have to wait for the NBER to get around to declaring the date of the end of the recession.	+0
31 August 2009	With August 2009 just about done, we anticipated that the S&P 500 would average between 1005 and 1020 in September 2009. And then we anticipate that a significant amount of noise would enter the stock market between 10 September and 16 September, which we initially anticipated would make stock prices dip. In a 10 September 2009 update to the post however, we reverse that forecast and instead predict a surge in stock prices, based on what we found when we dug deeper into the data.	We followed up our 10 September 2009 update on 16 September 2009, as we identified the likely source funding the rise in stock prices during this time. Meanwhile, the average level of the S&P 500 in September 2009 was 1044.55, which means that 2 of the 3 predictions recorded in this post were wrong, and one was correct. That nets us a -1 on this batch!	-1
8 September 2009	We suggest that teen employment figures might soon begin to improve provided no further minimum wage increases are in the works.	Too soon to tell. Since the data comes out monthly, it may not be until early January (when December 2009's jobs data is released) where we'll have an answer.	+0
16 September 2009	After confirming the run-up in stock prices above where our forecasting methods would place them, we anticipate that a selling opportunity is lurking in the near future.	That selling opportunity came by 30 September 2009, right before the S&P dove back down close to our forecast range for September 2009 in early October, before rebounding.	+1
30 September 2009	We come to the fork in the road for our method of forecasting the direction of the S&P 500. Option A puts the average for the index in October 2009 between 867 and 892. Option B puts the average for the index between 1089 and 1106.	Too soon to tell. Right now, it's looking like investors have chosen Option B!	+0
8 October 2009	We take a closer look at what we believe drove stock prices up in September 2009, finding that those forces are still at work in October. In a chart, we identify two approximate dates where the upward ride provided by our suspect source of market noise might come to an end: 15 October 2009 and 17 November 2009 (both one year after the actual indicated dates shown on the chart.)	Too soon to tell. At this writing, we would say that 17 November 2009 looks more likely, mainly because of the sustained fall-off in the year-ago 10-Year/3-Month spread in U.S. Treasuries. We'll find out within the next several days if there's anything to the large drop in the year-ago spread that we see after 15 October.	+0

What we like about this approach is that if our predictive ability is no better than the random outcome determined by a coin toss, our plus-minus score will drift toward a value of zero over time. If we're better at making predictions than simple randomness would suggest, then our plus-minus score will grow higher in value over time. If we're wrong, then our plus-minus score will fall in value. If we're really bad, then our plus-minus score will plunge into deep negative territory!

Overall, since January 2008, our net scores are as follows:

• 10 no decisions or equally mixed outcomes (such as a multiple part prediction where one part is correct and one part is incorrect.)
• 10 incorrect predictions. This includes all our near-misses (there are three at this writing.)
• 28 correct predictions.

The numbers above don't add up to the number of predictions made due to the effect of multiple part predictions.

Previously on Political Calculations

• Plus-Minus for Predictions - 16 April 2009
• Our Plus-Minus Is Now Seventeen! - 16 July 2009

Labels: forecasting, track record

- posted by Ironman at 12:08 PM | Permalink | Home |