Morgan Stanley Defends Retail Sales' Seasonal Adjustments From "Crazy Zero Hedge Analysis"; BAC Upgrades Netflix

Tyler Durden's picture

You heard our side of the story. It is only fair you hear the other side too.

From Morgan Stanley's chief economist, David Greenlaw:

There is some crazy analysis on Zero Hedge that seems to be getting some traction because a few clients have asked about it (see their chart below). 

 

 

Basically, ZH claims that a bias in the retail sales seasonal adjustment factor distorted the July results to the high side.  However, they are looking at the dollar adjustment for the level of sales in July alone.  Since the main focus in the retail sales report is on the percent change in seasonally adjusted monthly retail sales, the relevant comparison is the percent change in the seasonal adjustment factor between June and July.  Moreover, calendar effects matter.  For example, the Fourth of July holiday was midweek this year and there were only 4 shopping weekends in July 2012 -- one less than in July of the past few years.  The last time the July calendar was identical to 2012 was in 2007.  So, we show below a comparison of the monthly percent change in the seasonal adjustment factor for 2012, 2011 and 2007.  The bottom line is that seasonal adjustment factor for July 2012 added about 2 percentage points to the monthly change in retail sales -- essentially the same as in July 2007. 

 

 

Note:  a positive % ch for the SF subtracts from the NSA change and a negative % ch in the SF adds to the NSA change

* * *

And now you know the Morgan Stanley version of things.

What can we say? In X-12-ARIMA we trust:

A Basic Seasonal Adjustment Glossary

ARIMA models     These are a versatile family of models for modeling and forecasting time series data. Seasonal ARIMA models have a special form for efficiently modeling many kinds of seasonal time series and are heavily used in seasonal adjustment. ARIMA is an acronym for AutoRegressive Integrated Moving Average.

Flow series    Time series measuring consecutive changes over a unit of time, such as monthly sales or monthly cash flows, or monthly changes in any stock series.

Irregular component    This is the residual time series that results from the removal of estimated seasonal and other systematic calendar-related components of an observed time series, along with the removal of an estimated trend-cycle component.

Moving holiday effects    These are systematic changes in the values of a time series that are associated with the timing of moving holidays, i.e. holidays whose dates vary from year to year, such as Easter, Passover, Ramadan, Chinese New Year and U.S. Labor Day. Estimates of one or a combination of such effects define the moving holiday component of time series.

RegARIMA models (Also regression+ARIMA models.)    In the seasonal adjustment context, a hybrid model in which some features of the time series, such as moving holiday, trading day and outlier effects, are modeled with linear regression variables while the remaining features (those of the regression residuals, including trend, cycle and seasonal components) are modeled with a seasonal ARIMA model.

Seasonal adjustment    The estimation of the seasonal component and, when applicable, also trading day and moving holiday effects, followed by their removal from the time series. The goal is usually to produce series whose movements are easier to analyze over consecutive time intervals and to compare to the movements of other series in order to detect co-movements.

Seasonal component    A time series whose values quantify (usually in percents or in the units of data measurement, e.g. dollars) variations in the level of the observed series that recur with the same direction and a similar magnitude at time intervals of length one year. (Length is measured in the calendar units of the observed series--usually quarters or months, sometimes semesters, weeks, or other units.)

SEATS (Signal Extraction in ARIMA Time Series)    This approach is used within TRAMO-SEATS and also within the X-13A-S seasonal adjustment packages. It can simultaneously estimate the different components of a time series.

Stock series    For the unit of time of the series, months for example, these are time series like end-of-month inventories that arise as the cumulative sum of inflows and outflows (i.e. monthly net flows) starting from some initial value in the past.

Time series    A sequence of measurements of an economic (or other) variable made at approximately equally spaced times. It is important that the definition of the variable and the method used to measure it be consistent over time.

Trading day effects     As practical concerns, these are systematic effects in monthly times series related to changes in the day-of-week composition of each month and, in some cases, also to changes in the length of February. For flow series (monthly accumulations of daily activity e.g. monthly sales), the increases or decreases from average day-of-week activity associated with the days that occur five times in the month in a given year are important. (If they are days of high sales volumes, the monthly value will be inflated, etc.) For flow series, the length of February can have an impact. (More days than average should produce more sales than average for February.) For stock series, such as end-of-month inventories, the extent to which inventories tend to rise or fall on the day of measurement (e.g. the last day of the month) can have an impact that is different from year to year. Attempts to measure analogous effects in quarterly series are seldom successful. A series of estimated trading day effects defines a trading day component for the time series.

TRAMO (Time Series Regression with ARIMA noise)     This approach is used within TRAMO-SEATS seasonal adjustment sofware. It is an approach to estimate and prior correct time series before seasonal adjustment.

TRAMO-SEATS     Seasonal adjustment software developed by the Bank of Spain. It uses models to estimate the different time series components.

Trend-cycle component     This is an estimate of the local level of the time series that is expected to include the effects of moderately short- and well as long-term movements associated with the "business cycle". It is often obtained by applying a customized smoothing procedure (a data-dependent "trend filter") to the seasonally adjusted series to suppress its oscillatory movements over short time intervals, i.e. its higher frequency movements.

X-11     Seasonal adjustment software originally developed by United States Census Bureau. It is based on an iterative application of linear filters.

X-11-ARIMA     A Seasonal adjustment software developed by Statistics Canada. It incorporates ARIMA modelling to improve estimation of the different time series components.

X-12-ARIMA     A Seasonal adjustment software developed by the United States Census Bureau. It incorporates regression techniques and also ARIMA modelling to improve estimation of the different time series components.

X-13-ARIMA-SEATS (X-13A-S)     Seasonal adjustment software under development at the U. S. Census Bureau in collaboration with the Bank of Spain that integrates an enhanced version of X-12-ARIMA with an enhanced version of SEATS to provide both X-11 method seasonal adjustments and ARIMA model-based seasonal adjustments and diagnostic.

* * *

In other news, Bank of America just upgraded Netflix from Neutral to Buy with the following logic:

More potential upside than downside; maintaining $72 PO

 

We are maintaining our $72 price objective based on a sum-of-the-parts (SOP) analysis. Our SOP assigns a 20x multiple to the domestic streaming segment, a 6x multiple to the domestic DVD segment and gives no credit for the international streaming segment. As shown in Exhibit 2, we see a theoretical upside and downside case of $30 and $117 by using the lowest and highest multiples and contribution margins we believe are reasonable, respectively.

 

Bears have valid concerns…

 

Bears are primarily concerned about: 1) the likelihood that content prices will be increasing due to Netflix’s exclusive and original content emphasis, which could pressure the streaming bottom line; 2) international will likely suck up consolidated global profitability for years and this is too long to wait in a rapidly changing industry; 3) increasing competition both domestically and internationally; 4) unproven  longterm domestic streaming segment profitability; and 5) rapidly declining domestic DVD profit as the most profitable customers churn off the service.

 

…but Bull arguments more convincing to us

 

The bulls are mainly positive about: 1) continued international expansion being a drain on future profits but the right long-term strategic business decision; 2) Netflix becoming better at buying content and eliminating wasted spend as it analyzes audience viewership trends and historical data – likely increasing bottom line streaming profitability; 3) domestic earnings are becoming significant and Netflix can benefit from being the first mover in numerous international countries; and 4) at point of or beyond incremental domestic streaming segment profitability.

In other words, even the bulls admit the company will not make money, but at least it is the best thing they can hope to achieve. After all, who needs cash anyway. Just copy the AMZN model of ever lower gross profits resulting in an ever higher stock price as the former bookseller morphs into a bank-space station-precious metal miner-cloud gamer-retailers' retailer-[everything else you have never imagined].