I have written about the two different models before and have also compared their accuracy over time for the data that we have available (the Nowcast methodology is relatively new and the models have only existed for a few years).
In my analysis, I have shown that the Atlanta model so far has a better track record than the New York Fed model, with an average absolute forecast error of about 0.6% (compared to a forecast error of about 0.9% for the NY model).
Currently, most private forecasters are also predicting a quarterly growth rate somewhere on between 3 and 4%, meaning that the NY model is probably much too bearish right now.
It thus wouldn’t surprise me at all if the quarterly GDP figure came in somewhere in between 3.5 and 4%. In fact, I think the likelihood of this happening is pretty high.
Also keep in mind that the quarterly GDP figures for the first two quarters of the year are 2.2 and 4.2%, respectively, meaning that the likelihood that the annual GDP growth rate exceeds 3% is increasingly high (depending on the actual outcome of the third quarter). A more than 3% growth rate would be the highest figure since the financial crisis of 2008, which is quite remarkable given that we have been living in the era of low real interest rates and weak productivity growth, sometimes known as secular stagnation.
So what happened? Well, actually "Trumponomics" has some to do with it. The trillion dollar tax cut, basically a big handout to large corporations and high-income people, was always expected to provide a short-run stimulus to the economy , unless the entire effect would be offset with a tighter monetary policy by the Fed. While Fed officials were able to increase interest rates somewhat faster than previously expected as a result of a booming economy, they did allow for some additional stimulus as some slack in the labor market remained.
Note though that the long-term estimate of potential growth has not been revised upwards, Furthermore, Fed officials now seem all to eager to push up interest rates to neutral or even beyond, at least temporarily, to restrain potential inflationary pressures of a full employment economy.
There is a very good chance then that next year’s growth rate will be substantially lower as Trump’s economic stimulus will wear off while the Fed will continue its tightening cycle. He big risk here, of course, is that they will overdo it and push the economy into recession because they underestimate how low the new neutral rate really is. As Rüdiger Dornbusch once said: Expansions don’t die of old age, every single one of them was murdered by the Fed.
New York Fed GDP Nowcast
UPDATE:
I published this blog post this morning. I just saw that the estimate of the Atlanta GDP Nowcast model dropped to 3.6% from 3.9% earlier today, which increases the probability that the actual GDP number released tomorrow will somewhere in the mid-3% instead of being at 4% or above. That being said, with last quarter's growth rate of more than 4%, we can still expect annual GDP growth to be quite high, in all likelihood just exceeding 3%, depending on the actual performance in Q3 and Q4.