In April 2009, we began keeping track of how accurate all of our predictions have been since January 2008, providing updates for our readers once a quarter ever since. Today that comes to an end, as we've now completed our two-year experiment in publicly forecasting future economic and market events and reporting the results.
During that time, we've measured the accuracy of our predictions using one of the most unforgiving metrics ever developed for measuring performance: the plus-minus statistic from hockey and basketball. Using this system, we gain a point if we're right, lose a point when we're wrong, and score a zero for when the outcome of a prediction either cannot yet be determined or, in the case where we make multiple predictions, when we have contrary results that cancel each other out. And like a game of horseshoes, close doesn't count where near misses are involved.
Ultimately, our plus-minus score will reveal the number of times we scored hits rather than misses in making predictions of the future. If we're no better at predicting the future than a coin toss, our plus-minus score will gravitate toward zero. If we're better at predicting the future than that, our score will rise over time, and if we're really bad, our score will surely fall.
In January 2011, our recorded plus-minus score stood at +38. Here's how we stand today for all the predictions we've made through our final public prediction at the beginning of April 2011, or for which we are still waiting to discover the outcomes:
|Political Calculations' Predictions Plus-Minus Score Update, 2 May 2011|
|28 September 2010||We create a tool that predicts that the average cost of tuition and required fees at a four-year institution of higher learning (aka "college") will rise to $14,541 in 2009, $15,394 in 2010, $15,869 in 2011, $15,538 in 2012 and $16,212 in 2013. We modify these predictions for 2009, 2010 and 2011, taking the "under" for each of these years. We would anticipate being within 2% of the values for 2012 and 2013.||These predictions will take some time to play out. The confirmation will be provided by the annual Digest of Education Statistics produced by the U.S. Education Department.||+0|
|21 January 2011||After seeing the price-dividend growth rate ratio spike in December 2010, we anticipate that we'll see a trough in stock prices by the end of February 2011.||While the 1.8% dip in stock prices on Friday, 21 January 2011 suggests a correction may be in the offing, we won't know if the S&P 500 will have passed through a trough in February 2011 until March 2011. Update: We missed it by half a month; the trough in stock prices didn't happen until March 2011. Still, when recognizing that there was an over 80% chance of the event occurring by the end of February, that's one bet we'd make again.||-1|
|1 February 2011||We forecast that the S&P 500's average daily closing price for the month of February 2011 would fall somewhere in the range between 1289 and 1321.||The S&P 500 averaged 1321.12 during the month of February 2011. Yes, we were only off by 0.12 points (or 0.0091%). Yes, it hurts, but fortunately, not as much as it might have, which you'll see when we get down to 18 February 2011.||-1|
|3 February 2011||Using the "final" third estimate of GDP from the third quarter of 2010, we project that GDP, when adjusted for inflation, will fall between $13,082.3 billion and $13,616.3 billion constant 2005 U.S. dollars during the fourth quarter of 2010, with a target value of $13,349.3 billion. We even give a 68.2% probability of it being within $13,209.1 billion and $13,489.5 billion.||GDP in 2010-Q4 was finalized at $13,380.7 billion (in constant 2005 U.S. dollars), so we just missed hitting dead center for our target range by 0.23%.||+1|
|18 February 2011||We suggest that a correction is "in the offing" for the S&P 500. We also up our previous forecast range for February 2011 to 1294 to 1332.||Maybe not quite in the offing, but a correction did take place in the S&P within 13 trading days of our call. Still, this is where our strict standards for making predictions prevents us from claiming this as a hit, because we didn't specify the period in which our new prediction would hold -- only pointing to our previous prediction that a correction would take place by the end of February 2011. With that being the case, those portions of our predictions cancel each other out. Fortunately for us, we made a third prediction, as we adjusted our original forecast range for February upward, and the S&P cooperated.||+1|
|25 February 2011||Sometimes, we do predictions just for fun, and if you go back through our plus-minus table archives, you'll see we miss these kinds of predictions far more often than we score a hit. Here, we predicted that the televised broadcast of the 2011 Oscars ceremony was "doomed to flop."||"Flop" doesn't even describe how bad the Oscars were this year, as the broadcast's ratings were down 10% from the previous year, while negative criticism for how bad the broadcast was achieved an all time high. But then, if you listened to us, you could have avoided any suffering. Unlike poor Craig Ferguson.||+1|
|28 February 2011||We predict that the U.S. unemployment rate will likely bottom out no lower than 8.3% during 2011. We predict that if the national average of gasoline prices in the U.S. exceeds $3.50 per gallon for a sustained period of time, the U.S. unemployment rate will stop improving altogether and will worsen two years out.||We hate to say that we're on track so far with these predictions.||+0|
|1 March 2011||Our primary method for predicting the future level of stock prices suggests that stock prices will rise substantially in March 2011, to a range between 1336 and 1384 during the month. However, we don't buy into it, and uncharacteristically, we take the under on that range.||The S&P 500 in March 2011 spent the entire month below our forecast range, as we actually score a hit by missing our target.||+1|
|11 March 2011||With a correction having now beginning, we predict that the S&P will fall to average within a range between 1272 and 1302 during the middle part of March 2011.||The S&P saw its daily closing price average within that general range from 14 March 2011 through 23 March 2011.||+1|
|4 April 2011||Our two year anniversary for making public forecasts for where the S&P 500 would go a month ahead of time! We offer a "split" prediction, with a high range of 1393 to 1429 and a low range from 1171 to 1264, before rejecting both and picking the middle.||The nice thing about our forecasting methods is that we can usually get a pretty good indication of when they'll work and when they won't (which we can tell by quantifying how much noise is in the market - the more noise there is, the less accurate our forecast results). As stocks averaged 1,331.51 during April 2011, we made the right call in rejecting our methods' forecast results and taking the middle. Since this is our last public prediction for where the S&P 500 will go a month into the future, it's a nice way to go out in style.||+1|
By our final tally, we've made a total of 124 predictions since January 2008 and we're still waiting to find out the outcomes of eight. As for being off or on target, we've been off target on 37 occasions and on target for 79, giving us a plus-minus score of +42 overall. Our overall prediction accuracy rate is 68.1%, which is what you get when you divide our 79 hits by the 116 predictions where we've been able to establish the outcome and convert the result to a percentage.
But now, there's the little matter of how we compare to others, which we'll post soon to complete the grand finale for our forecasting series.