GDP data suggest it is less a no brainer, but not yet alarming

Updated at bottom for the inflation side

Wall Street’s obsession with “current” quarter GDP estimate is probably mainly a waste of time.  Between late February and yesterday, for example, the Atlanta Fed’s GDPNow estimate fell about 200 bps, but the bond and stock markets basically went sideways.  The 10-year Treasury yield was down marginally and the S&P500 was up slightly.

Screen Shot 2017-04-28 at 10.22.38 AM

In fairness, the obsession to which I refer is probably less pervasive now than it was earlier in the cycle, even among the talking heads (such as myself), but it is still there to some extent.  You do see a lot of references to the Atlanta Fed as the quarter progresses.

From a social welfare perspective, there is probably also still too much emphasis on GDP.  However, that too seems to be fading. People increasingly recognize that GDP is not a good measure of aggregate happiness, and not just because of bean-counting minutia, like quality adjustment or path dependency in chain weighting.

More fundamentally, just for example, people are prone to jealousy, which makes adding up utility, as real GDP is meant to do, tricky! And even if people weren’t prone to jealousy (or less darkly, were cheering for their fellow man), aggregating utility is tough. In any event, nobody argues that the welfare aspect needs to be tracked quarterly.

Screen Shot 2017-04-28 at 10.11.30 AM

But the GDP release allows us to update some aggregates that may be helpful in assessing the cyclical position of the economy, developing imbalances, and the attendant risk of recession.  I like to follow the share of cyclical demand in GDP and the private sector’s financial balance implied by the (all-government) fiscal deficit and current account.  For a brief explanation of the reasons, you may see here and here.

We now have data to Q1 for the first concept and an ability to get a pretty good bead on the second, based on the movement in net exports to Q1.

On the basis of both these metrics, the risk of recession appears to have risen quite a bit from where it was earlier in the cycle I (and many others) had earlier argued that you cannot fall off the floor and that concept of a “stall speed” is not that relevant when the plane has not even really left the ground.  But such arguments are now a bit less forceful.

Screen Shot 2017-04-28 at 11.03.52 AM

The share of cyclical demand in GDP has risen, which suggests that pent-up demands are less present than they were, particularly away from housing. And the private sector financial balance is smaller, reinforcing the same point, although from a financial rather than real-economy perspective.  (Accounting relates the two, but not quite into identities.)

On the other hand, this is a case of not-as-good, rather than alarming. From the perspective of these metrics, at least, the situation does not appear yet to have become particularly dangerous. It just seems less the no-brainer than it was earlier in the cycle.

We can’t take a view on the medium-term outlook based just on these crude metrics, obviously. The approach of full employment and the Fed’s (slowly) evolving priorities are a bit of a caution, although also  not yet alarming, in large part because inflation is still too low.

But from the GDP report itself, the main points I would take away are largely shown in the charts immediately above.  And their signal evolves slowly, so it is not like we learned so much even about these in today’s release.


Afternoon Update: Quick comment on the inflation side

JPM estimates that the core PCE deflator for March will have been down 0.09% (false precision) on the month. The JPM estimate is usually pretty close, and it is good enough for me.

Today, we got the price data for Q1, although not the monthly breakdown for March in particular. But if we assume the monthly data to February will not be revised (neutral, but not realistic), then to get the already known quarterly result, we would have to pencil in a 0.07% decline March for the core deflator and roughly twice that decline for the market-price-only (MPO) component.  Just as a demonstration of mastery of Excel I do that below.

The precise monthly detail don’t matter much. Quarterly is fine, because we ain’t that smart anyway.  But my charts are monthly and using monthly avoids talking about base effects in the quarterly averages. So here is how one picture would look if we did that bit of interpolation.

Screen Shot 2017-04-28 at 1.56.02 PM

I see from a Bloomberg journalist on Twitter that MEN now hold the lowest expectation of long-term inflation on record. Not sure why men and women wold be different.  But maybe the men are extrapolating, after what has almost been a decade now of the Fed chronically missing to the downside. I doubt inflation expectations drive policy to the extent the Fed occasionally implies (when not dissing inconvenient inconvenient readings).  It is marginally corroborating is all.

Incidentally, ignore the GDP deflator — ALWAYS.  Never mind why.  Need to know basis only. And you don’t need to.