If I’ve heard Keith reference PTJ’s quote about the last third of the move being “the toughest to risk manage” in a meeting once, I’ve heard it a thousand times. In bear markets, those who mismanage “the last third of the move” are short sellers that cover too early (in fear of the policy response) and bulls that cite “valuation” as a reason to purchase securities – quite often far too early. Both are buy orders, FYI.
Did the shorts cover too early?
No, well at least not from the perspective of our short-term trading signals. Last Wednesday morning, we published an immediate-term risk range of 1 for the S&P 500, which was an explicit signal for investors to cover shorts well into last Thursday’s low (1810 to be exact).
That’s not to suggest that we’ve nailed it by any stretch of the imagination, but rather to reiterate the omnipotence of actively managing risk in bear markets. We’ve been vocal in stressing that bear-market bounces are typically more sharp than their bull-market counterparts. The +645bps rally in the SPX from Thursday’s intra-day low to today’s closing price is allegedly the sharpest three-day rally since the thralls of our last bear market in late-2008.
Indeed, high short interest stocks are the 2nd best-performing style factor on a WoW basis, just behind high beta stocks. We consider that to be ample enough evidence of aggressive covering of crowded momentum shorts. I’m sure there will be plenty of research notes from prime brokers highlighting this very point tomorrow morning – if there haven’t been already. Obvious is as obvious does.
In light of the aforementioned strength in the U.S. equity market, has anything materially changed, quantitatively speaking?
No, certainly not from the perspective of our Tactical Asset Class Rotation Model (TACRM), which shows that:
- Realized volatility continues to accelerate on a trending basis across asset classes.
- TACRM continues to generate a [bearish] “DECREASE Exposure” signal for U.S. Equities at the primary asset class level.
- Across the 47 sector and style factor exposures TACRM tracks within U.S. Equites, only two are signaling bullishly from the perspective of TACRM’s multi-duration, volatility-adjusted view of volume-weighted average price momentum.
***CLICK HERE to download the latest refresh of our TACRM presentation.***
A much more difficult question to answer at the current juncture is, “Has anything materially changed, fundamentally speaking?”
On a trending basis, the answer to that question is a resounding “no” – especially with respect to: 1) the corporate profit cycle, 2) the C&I credit cycle, and 3) the monetary policy cycle. As we’ve detailed extensively in recent notes, the confluence of the likely progression of each will be the determining factor for ultimate downside in risk asset prices:
- The Domestic #CreditCycle Is About to Get A Lot Worse (1/26): “As the ongoing corporate profit recession deepens, we highlight the obvious risk of companies allocating earnings to buybacks and dividends, rather than preserving cash and paying down debt. If our explicitly negative view on the domestic economic cycle continues to be corroborated by the data, then the growth rate of such payouts will likely have to slow dramatically – if not decline outright. That is an obvious headwind to the broader equity market in terms of removing key cogs from the post-crisis secular bull case.”
- Ex-Energy? (2/3): “And even if the U.S. economy avoids recording a technical recession, we could just have an 2000-02-style corporate deleveraging cycle that drags down equity market cap alongside it. After all, it's EPS that matters most to stocks, not GDP. Recall that the 2001 downturn was the shallowest recession in U.S. history; that didn't preclude the stock market (SPX) from getting cut in half.”
- Sentiment Update: Three Things I Learned Today (and Three Weeks From Now) (1/20): “While ZIRP and LSAP have proven to be powerful tools in perpetuating income-inequality generating asset price inflation throughout this economic and corporate profit expansion, the Federal Reserve has yet to demonstrate the effectiveness of monetary easing during concomitant recessions in economic activity and corporate profit growth.”
On a shorter-term basis, astute investors have been asking us the right questions about sequentially improving high-frequency economic data. In the 1Q16 to-date, we’ve seen three releases that lend some pause to our bearish outlook for the domestic economy – if only for a brief moment:
- First, the ISM Manufacturing PMI ticked up ever-so-slightly to 48.2 in JAN from a reading of 48.0 in DEC. The New Orders index rose back above 50 for the first time since OCT; its 51.5 reading for the month of JAN is the highest since AUG.
- Then, the growth rate of Retail Sales – specifically the omnipotent “Control Group” therein – accelerated meaningfully to +3.1% YoY in JAN from +2.2% in DEC. Per the JAN release, the growth rate of Retail Sales – which accounts for a third of consumer spending and a fourth of GDP – is now accelerating on a trending basis.
- Lastly, the growth rate of Industrial Production accelerated to -0.7% YoY in JAN from a downwardly revised -1.9% in DEC. Sequential strength was recorded across all key product categories as Capacity Utilization arrested what had been five consecutive months of MoM contraction and YoY deceleration.
While it would be easy to poke holes in each of the aforementioned releases (for example, the Employment index of the JAN ISM report crashed to a new cycle low of 45.9), it’s not clear to me that that would be a valuable use of your time; nor would it be a valuable endeavor on which to expend analytical credibility. The risk of being perceived as arguing with the data for too long for a research outfit that doesn’t have banking, trading or asset management to support its revenues is arguably just as punitive as the risk of a money manager fighting the market for too long.
A better use of our collective time is attempting to determine the sustainability of the aforementioned positive deltas in economic data:
- With respect to a sustained recovery in domestic manufacturing activity, we flag sequentially declining U.S. dollar comps throughout the balance of the year as positive and consider the high likelihood that we tip back into #Quad4 – which has historically been positive for the USD – in the upcoming quarter to be a potential negative shock.
- With respect to a sustained recovery in consumer spending, we still believe the sharp acceleration in base effects through the third quarter of this year will continue to be a material negative catalyst for reported growth rates of household consumption. Specifically, base effects for Real PCE growth in the four-quarters ended 3Q16 are far and away the toughest since the four-quarters ending in 3Q08. Recall that Real PCE growth decelerated sharply from +2.7% YoY in AUG ’07 to -1.2% in SEP ’08.
- With respect to a sustained recovery in the industrial sector, we highlight generally receding base effects throughout the balance of the year as a positive catalyst. This is juxtaposed with the negative cocktail of rising credit costs, peak inventories and GDP-negative capital allocation decisions from major U.S. corporations (e.g. AAPL, IBM buybacks).
All that having been discussed, let us turn our attention to the Atlanta Fed’s recent revision of their “GDPNowcast” for Q1 to 2.6% per the latest release. That forecast is up from a trough of +0.6% as recently as mid-January and has the attention of many concerned short-sellers and confident bulls alike.
How important is the aforementioned estimate within the context of the Atlanta Fed’s ~4.5 year-old track record at forecasting U.S. GDP growth? Not very.
The following chart shows us the Atlanta Fed’s peak (green line) and trough (red line) estimates 47 days into any given quarter as we are currently. These figures are juxtaposed with the actual recorded QoQ SAAR growth rate of U.S. GDP (blue line) and the maximum tracking error of the Atlanta Fed’s GDPNowcast (black line). The average maximum intra-quarter-to-date tracking error is a startling 248bps over the duration of this study. That is a whopping 113% of average growth rate of GDP over that same time frame (i.e. 2.2%)!
PLEASE NOTE: We are not showing this analysis with the intention of undermining the Atlanta Fed. Their GDPNowcast estimates are a good barometer of what consensus thinks about near-term growth prospects and their work in charting the “Shadow” Fed Funds Rate has been equally additive to the Macro discussion. We simply highlight the elevated risk of their current forecast for Q1 2016 being revised materially lower (or higher) as the quarter progresses. It’s worth noting that we’ve hardly received much in the way of January data – let alone enough data points to have statistical validity in any regression model.
As we highlighted in our 9/2 Early Look titled, “Do You QoQ?”:
“In the context of modeling the economy, the more we learn about sequential momentum, the less we are able to know about the underlying growth rate of the economy. Recall that headline GDP growth accelerated +660bps to +7.8% in the 2nd quarter of 2000 and that it accelerated +470bps to +2% in the 2nd quarter of 2008. If you were prescient in forecasting these second-derivative deltas, you could’ve bought all the stocks you wanted en route to peak-to-trough declines on the order of -49.1% and -56.8%, respectively (S&P 500)… Going back to the aforementioned head-fakes, it’s clear that those read-throughs on sequential momentum failed to signal pending material changes to the underlying growth rate of the economy.”
In the aforementioned note, we also highlighted the poor track records of both Bloomberg Consensus and the Fed in forecasting quarterly and annual GDP growth rates in the post-crisis era:
- “Over the past five years, Bloomberg consensus forecasts for headline GDP just one quarter out have demonstrated a quarterly average [maximum] tracking error of 145bps. This means that at some point within 3-6 months of any given quarter-end, Wall St. economists’ estimates for QoQ SAAR real GDP growth were off by an average of 145bps. That’s flat-out terrible in the context of actual reported QoQ SAAR growth rates averaging just 2.1% over this period."
- “Over the past five years, the FOMC’s intra-year U.S. GDP forecasts have demonstrated an annual average [maximum] tracking error of 100bps. Worse, the maximum deviation of their intra-year forecasts from the actual reported annual real GDP growth rate was an upside deviation in every single year, meaning that the Fed’s growth forecasts are consistently far too optimistic.”
As the following charts demonstrate, the aforementioned updated quarterly and annual average maximum tracking errors are now 164bps and 91bps, respectively (data through EOY ‘15). These updated figures continue to reflect the fact that investors should remain highly skeptical of even the nearest-term growth forecasts from economists and/or policymakers. At best they’re playing a game of Marco Polo; at worst, they are feeding potentially hazardous information to market participants.
All told, as we discussed at length in our 1/29 note tilted, “What Recession?!”, predicting headline (i.e. QoQ SAAR) GDP is a fool’s errand. Don’t get caught up in the current strength of the Atlanta Fed’s GDPNowcast or what may actually turn out to be a decent Q1 GDP report on a headline basis. We strongly consider both of those catalysts to be head fakes in the context of the underlying sine curve of U.S. economic and capital markets activity.
Best of luck out there,