The Atlanta Nowcast estimate for Q1
The New York Nowcast estimate for Q1
"The ideas of economists and political philosophers, both when they are right and when they are wrong, are more powerful than is commonly understood. Indeed the world is ruled by little else. Practical men, who believe themselves to be quite exempt from any intellectual influence, are usually the slaves of some defunct economist."
The General theory of employment, interest and money (1935)
US quarterly GDP for the first quarter with a growth rate of 3.2% came in surprisingly strong yesterday. However, a substantial part of this can be attributed to inventories and net exports, meaning that another big upside surprise for next quarter is rather unlikely. I have kept track of the different Nowcast models on this blog before (see here, for example). So far, the Atlanta GDP Nowcast has turned out to be the most accurate (see table below). Not only does it have the smallest absolute forecast error, but it’s missed also seem to be unbiased, i.e. they average out over time.
More importantly, the Atlanta Fed model revised its Nowcast upwards on several occasions over the last few months whereas the New York Nowcast and all the private forecasters revised their GDP estimate downwards. Clearly, the Atlanta's model forecast got better overt time as this quarter progressed whereas most other GDP estimates got worse. It therefore seem that the GDP Nowcast model has a significant edge, not only compared to the other Nowcast models, but also compared to private sector estimates. This quarter's GDP Nowcast estimate is another big win for the Atlanta model!
The Atlanta Nowcast estimate for Q1
The New York Nowcast estimate for Q1
I have written previously on the US yield curve inversion. An inverted yield curve has historically been one of the best recession predictors. While it is not entirely clear why, one of the reasons seems to be that when the short-term rate is higher than the long-term rate, the Fed usually has tightened monetary policy too much. Low long-term rates obviously indicate a combination of expected lower future short-term rates and low growth prospects in the future. Moreover, given the degree of interest rate inertia and the Fed’s unwillingness to reverse course, the monetary tightening produces a recession. While I’m not entirely sure why this should be different, the Fed has recently proven that it will abandon its monetary tightening until the inflation rate accelerates. Markets are even pricing in this year a Fed rate cut now. However, this also implies that monetary policy might actively become tighter if the Fed does not fulfil market expectations. As the charts below show, the yield curve has not only inverted in the US, but also in other countries: Canada, Japan, and Germany. The Canadian inversion actually looks particularly nasty (see pictures below). However, the US and Canada still have short-term interest rates that are above zero (even though not very far above). This implies that the two Central Banks do actually have some room for rate cuts and monetary easing to respond to the global synchronized economic slowdown that is currently happening. Both the Eurozone and Japan are not so lucky because short-term interest rates are still constrained by the effective lower bound, thus also explaining why both Germany’s and Japan’s yield curve are only slightly inverted as interest rates cannot fall substantially below zero. The ECB, of course, just ended its asset purchase program last December, and is now presiding over yet another economic slowdown in the Eurozone. This does not seem to be a coincidence. Japan is still executing an asset purchase program right now, which has been extremely substantial in size. Further easing therefore seems to be unlikely. While a rewind of the ECB’s QE might be warranted, it rather seems unlikely to happen. Of course, Central Banks can theoretically never run out of financial assets to buy. So the current constraints are rather political and institutional in nature. Precisely for these reasons, we should be worried about the current economic slowdown because Central Banks have limited capacity for further easing and fiscal policy makers do not seem to intend to pick up the slack. These yield curve inversions therefore seem to be a harbinger of pretty bad news.
Yield curve inversions for the US and Canada:
Both countries feature an inverted yield curve compared to just a few months ago. However, short-term rates are still above zero, meaning that both Central Banks have potentially some room to ease monetary policy.
Yield curve inversion for Germany and Japan
Both countries have a slightly inverted yield curve despite short-term interest rates already being below zero in nominal terms, which is insane.
Turkey has a completely inverted yield curve. This could indicate two things. First, inflation rates, which are currently running at 20% or so, are expected to come down again in the very long-run. But it also probably implies the prospect of slower economic growth in the immediate future. After growing reasonably strongly in recent years, Turkey's economy seems to be in recession territory right now.
In my previous two blogposts (here and here), I have written about the fact the fact that the global economy is currently slowing down quite rapidly, especially in the Eurozone. However, GDP predictions for Q1 for the US were looking quite miserable as well. However, this might have been maybe a too hasty judgement since the most recent data has shown some improvement. I have written before on the different GDP Nowcast models. Historically, the GDP Nowcast model used by the Atlanta Fed has been the most accurate with an average absolute forecast error of under 0.6%. Just in the beginning of March the model forecasted a Q1 GDP figure of less than 0.5%. Now one month later, this estimate has increased to 2.1% and is therefore much more bullish than both the NY Fed Nowcast (at 1.3%) or the Blue Chip consensus (see figure below).
There is still a good chance that the first quarter growth rays might come in at only 1%, which would be well-below the Fed’s estimate of the economy’s long-run potential (~2%), but the probability of this happening seems to be lower now than just a few weeks ago.
According to the Atlanta model, the increase in its GDP estimate comes mostly from a higher contribution of net exports as well as inventories. While both items were supposedly subtracting from GDP growth more than a month ago, their contributions have now turned positive in recent weeks, therefore leading to the revised GDP estimate. As previously stated, the Atlanta model has so far outperformed its NY counterpart in terms of forecast accuracy. The upward revision in its GDP estimate has been quite dramatic and as far as I recall there are very few quarters where the forecast has changed to such an amount in one direction within just a few weeks. Furthermore, the Blue Chip consensus estimate as well as the NY Fed Nowcast have remained fairly stable/slightly trended downwards over that time period. We will find out in a couple of weeks from now whether the dramatic improvement in the Atlanta Fed’s model was warranted or whether it will turn out to be an outlier. Stay tuned!
My name is Julius Probst.