- Growth Accelerated Sequentially … but the internals were squishy, core expenditure types slowed, ancillary expenditure types improved (for some of the wrong reasons), and distortions were both prevalent and impactful.
- Moreover, slowing consumer spending growth is likely to weigh on broader GDP growth going forward. In short, we believe the depths of the GDP cycle are fast approaching and that investors would do well to proactively prepare for that outcome.
- In GIP Model terms, all of this equates to a move back down into #Quad3 for the next couple of quarters. We missed calling the narrow move into #Quad2 in 3Q16, which is why we missed getting out of the way on the long side of rates and also missed taking advantage of the tremendous resurgence in tech stocks. Heading back into #Quad3 however, implies that now is likely a time to scale up one’s exposure to long duration, as well as our favorite equity income plays like Utilities and long lease duration, non-cyclical REITS.
Contextualizing 3Q16 GDP
Growth Accelerated Sequentially … but the internals were squishy, core expenditure types slowed, ancillary expenditure types improved (for some of the wrong reasons), and distortions were both prevalent and impactful.
There was a good deal going on in this morning’s advance GDP estimate for 3Q16 and a visual tour of the underlying data is below, but most of what matters can be distilled into a few discrete points:
- Consumption = Slowing: Both Consumption growth and Consumptions contribution to headline Growth got cut in half relative to 2Q. With some of the toughest employment and income comps of the cycle in 4Q (employment growth, aggregate hours growth and aggregate income growth all face harder comps in 4Q) consumption growth is likely to continue to decelerate while providing diminished support to headline growth.
- Investment/Inventories: Good, But... : Inventories saw a notable +1.77 pt reversal in contribution to GDP. With inventories contributing negatively to GDP for 5 consecutive quarters – the longest such streak in 60 years – the gain wasn’t particularly surprising but it’s also unlikely to prove durable. Inventories have grown at a positive spread to sales for most of the last few years and inventory levels across the supply chain remain elevated. A late stage expansion with recessionary capex spending and decelerating household spending growth is not a factor catalyst for sustained inventory investment.
- Net Exports = Distorted: The Trade Balance registered its largest contribution since 4Q13 at 0.83%. Much of the strength was illusory (& well telegraphed by media/research outlets) as agricultural exports saw a step function rise on the back of the ramp in soybean exports. Soybean exports were up +259% QoQ and +215% YoY and singularly helped improve total nominal export growth by ~200bps in 3Q. This distortion will reverse in the coming quarter(s). Net exports may have shown some modest improvement but the 3Q estimate overstates the underlying reality.
So, yes, better sequentially but with some not inconsequential ceveats. Looking forward, consumption growth should remain 2nd derivative negative and the components that drove the bulk of the upside in the current quarter are unlikely to persists (inventory investment) or will reverse (Ag exports) in the coming quarter.
It's been our basic slower-and-lower-for-longer contention that we're at +1.5% +/- economy. Inclusive of 3Q16, TTM GDP growth has averaged +1.5% on a QoQ basis and +1.58% on a YoY basis and with a negative slope - a trend which aptly characterizes the underlying macro reality ... economic cycles never slow straight to 0%.
-Christian Drake, Senior Analyst
GIP Model Update: Betting Against the Math
Normally I’m not smart enough to argue with math, but there are exceptions to every rule and modeling GDP for the upcoming quarter (i.e. 4Q16E) is most certainly one of them. Specifically, the math states that 68% of the time when the base effects recede/steepen, YoY real GDP growth should accelerate/decelerate in the forecast period.
That’s the easy part; calculating the magnitude by which said accelerations and decelerations occur is where the real money is made in forecast accuracy terms. To that tune, our forecasts for YoY real GDP growth boast a standard error of 23bps over the trailing five years, which compares to 38bps for the Wall Street economist community otherwise known as Bloomberg consensus.
Going back to the base effect discussion, receding base effects perpetuated a +20bps acceleration in the YoY growth rate of real GDP to +1.5% in 3Q16. This compares to our predictive tracking algorithm, which was calling for a -10bps deceleration to +1.2%, and to Bloomberg consensus, which was calling for a +10bps acceleration to +1.4%. Importantly, GDP base effects continue to recede here in 4Q16E, which would imply a further sequential acceleration in the YoY growth rate of real GDP based on the probability parameters outlined above.
There are a number of key factors why we think this is one of those times where it pays to bet on the 32% of the time the second derivative of GDP in the forecast period carries the same sign (i.e. +/-) as the second derivative of the base effects.
- The obvious implication here is that we expect YoY real GDP growth to decelerate – albeit modestly – in 4Q16E. This puts us at odds with Wall St. consensus, which is expecting a further +30bps acceleration during the period to +1.8%.
- Most importantly, the delta between our interpolated headline GDP estimate (+0.5% QoQ SAAR) and that of Wall St. consensus (+2.3% QoQ SAAR) is a whopping 180bps!
- Looking out an additional quarter, our model is currently forecasting a -0.7% QoQ SAAR figure for 1Q17E; this is -270bps below the current Bloomberg consensus estimate of +2.0%.
The line in the sand has been drawn and below we detail why we’re on the side we’re on.
The first and foremost factor is that should take a bite out of headline GDP in the upcoming quarter(s) is a reversal of pre-election shenanigans. That’s right. I said, it. Specifically, net exports and inventories combined to contribute a whopping 50% of the headline growth rate 3Q16! This is the first aggregate positive contribution of those factors since the GDP cycle peaked in 1Q15 and the highest rate of positive contribution in over three years; you have to go back to 3Q13 to find a higher proportion than today’s 50% figure.
Unless the U.S. has transformed itself into an export economy that fervently builds inventories into preexisting peak inventory-to-sales ratios, it’s reasonable to bet that the trending drag on GDP associated with these two factors is likely to continue – especially with the U.S. dollar up +4.5% on a broad trade-weighted basis since it’s 4/29 YTD low.
Secondly, the slowdown in consumption growth recorded in 3Q16 is likely to continue over the next couple of quarters due to a confluence of highly-credible factors. Prior to outlining those factors, it’s important to contextualize the aforementioned deceleration:
- Headline consumption growth slowed -220bps to +2.1% QoQ SAAR in 3Q16; that represents a -51% decline in the pace of consumer spending growth.
- Consumption growth contributed 147bps – or just half – of the +2.9% QoQ SAAR headline figure. This is down -49% from a contribution of +288bps in 2Q16.
Why is the growth rate of consumer expenditures – which represents ~69% of real GDP – likely to continue slowing over at least the next couple of quarters? The answer is simple: income growth should continue decelerating and the personal savings rate is likely to mean revert higher.
We continue to highlight three near-term headwinds to employment growth that should weigh on income growth on a go-forward basis:
- The first factor is the ongoing recession in corporate profits. Specifically corporate profit growth has historically lead peaks and troughs in the pace of nonfarm payrolls growth.
- Moreover, the lowest non-recessionary rates of nominal GDP growth EVER are weighing on the outlook for corporate profitability. The negative trend in capex growth (down YoY in 20 of the past 21 months) implies future earnings potential is falling, not rising.
- Additionally, the confluence of political uncertainty and a scarcity of quality labor supply are negatively impacting small business hiring plans and that negativity is showing up in the JOLTS data, which itself appears to have initiated its topping process.
Moving along, we continue to view the recent decline in the personal savings rate as wholly unsustainable. The sideways trend in consumer confidence doesn’t imply consumers are feeling better about their wallets, which should impute to a static savings rate.
Furthermore, the deeply negative spread between the expectations component and the present situation component of the Conference Board’s consumer confidence report implies consumers might actually start to build savings (i.e. tone down consumption) into peak recession fears. Importantly, the latest reading here implies a recession may be imminent as we’re already in the area code of prior pre-recession cycle trough spreads.
That all this is occurring amid cycle-peak base effects for real PCE all but ensures a further deceleration the growth rate of consumer spending.
We’ll get the PCE data for SEP on Monday (yes, the advance GDP estimate is consistently released without the government knowing ~23% of the number) and if the softening growth rate of retail sales – particularly in big-ticket items like autos – is telling, total household consumption growth is likely to experience a weak handoff from 3Q16 to 4Q16E.
In GIP Model terms, all of this equates to a move back down into #Quad3 for the next couple of quarters. We missed calling the narrow move into #Quad2 in 3Q16, which is why we missed getting out of the way on the long side of rates and also missed taking advantage of the tremendous resurgence in tech stocks. Heading back into #Quad3 however, implies that now is likely a time to scale up one’s exposure to long duration, as well as our favorite equity income plays like Utilities and long lease duration, non-cyclical REITS.
All told, we missed nailing the 3Q16 GDP print right on the button largely because our predictive tracking algorithm can’t accurately track government shenanigans – which we expect to dissipate after the election. Moreover, slowing consumer spending growth is likely to weigh on broader GDP growth going forward. In short, we believe the depths of the GDP cycle are fast approaching and that investors would do well to proactively prepare for that outcome.
-Darius Dale, Senior Analyst