Europe In Triple-Dip Recession, Goldman's Internal Model Finds

If today's European Commission slashing of Euro GDP forecasts did not leave a warm and fuzzy feeling in Europeans that the greatest depression ever is proceeding just as planned, if not quite as "forecast" as the following chart shows, confirming yet again that when it comes to predicting the future nobody can hold a candle to the Fed, the IMF or Europe...

...  then here is Goldman with the loudest warning yet, courtesy of its internal RETINA model, that Europe is now effectively in a triple-dip recession, with Q3 GDP for the Euro area at -0.2%.

From Goldman's Huw Pill

RETINA retreats further into Q3 contraction  


Bottom line: We are less than a fortnight away from Eurostat's publication of its flash estimate of Q3 GDP growth in the Euro area. In today's Daily, we look through the lens of our contemporaneous tracker of real-time inflation and activity. Since our previous update in mid-October, RETINA's median estimate of Q3 GDP growth has moved deeper into negative territory, driven largely by a disappointing print for area-wide industrial production in August. The downside risks to our +0.1%qoq judgemental forecast for Q3 GDP now look skewed to such an extent that our point estimate no longer falls within a 50% confidence interval around RETINA's median reading.


RETINA sees negative GDP growth in Q3


As Chart 1 shows, from mid-September to mid-October, RETINA's median estimate for third-quarter GDP growth (the red line in Chart 1) fell from around +0.3%qoq to just short of -0.2%qoq. Following a disappointing contraction in area-wide August IP on 14 October, RETINA's median estimate fell a further 10bp -- yet deeper into negative territory. Having stabilised at around -0.3%qoq in the past fortnight, RETINA's median estimate is now some 40bp weaker than our current judgemental forecast for Q3 GDP growth (+0.1%qoq, the black dotted line in Chart 1). This is yet more pessimistic than the latest available poll among other private-sector economists (collated on 8 September), which envisaged Q3 growth of around +0.35%qoq.



RETINA's latest leg lower is down (solely) to Euro area IP


As Chart 2 shows, the latest move lower in RETINA's median growth tracker (from around -0.2%qoq to -0.3%qoq) was driven almost exclusively by the 1.8%mom contraction in Euro area IP in August. Conditional on this out-turn, subsequent releases of national business surveys (ranging from the Italian ISTAT, the Belgian business survey and the French and German PMIs), as well as a +0.5%qoq sequential expansion in Spanish Q3 GDP, left our RETINA growth tracker largely unmoved.



The mechanical nature of the RETINA framework implies that it may underestimate the potentially significant 'calendar effect' in the German IP data (changes in the timing of holidays is likely to have shifted production out of August into July, as reflected in the month-to-month volatility of outturns). Some caution is required in interpreting the downward shift implied by these data, at least until we see the September print later this week. That said, the broadly confirmatory signal offered by business surveys (e.g. with the German IFO index continue to decline) suggest that idiosyncracies in the data should not be overstated.


RETINA suggests that degree of downside risk to our forecast has returned


As Chart 3 shows, RETINA's growth tracker implied an escalation of downside risks to our former (+0.4%qoq) judgemental forecast through most of September. The latest indications are that the intensity of that downside skew has returned through the course of October -- even as it pertains to our much weaker current forecast for +0.1%qoq growth in Q3. The Bayesian underpinning beneath RETINA's growth tracker allows us to quantify this skew. Chart 3 shows that the model-implied probability that Q3 growth beats our judgemental forecast has fallen to 25% -- down from around 35% at the time we made our forecast change. Furthermore, as Chart 1 also underscores, the downside risks to our judgemental forecast for Q3 now look skewed to such an extent that our point estimate no longer falls within a 50% confidence interval around RETINA's median reading.



So far, we have discussed RETINA's median tracking estimate. We can also track RETINA's modal estimate of GDP growth - that is, the single estimate of sequential growth that the models deems more likely. As Chart 4 shows, this modal estimate has been broadly unchanged over the past month.


* * *

There is some good news:

It is still too early to be emphatic, but RETINA seems to object less to a small expansion in Q4 GDP than it does to a positive print in Q3.

And that is, what in the New Abnormal, passes for good economic news.