Datapocrypha

From the Office of National Statistics and Bank of England to private professional bodies and business surveys, data abound. Sizing up these data is a core aspect of understanding recent economic and market trends. But wading through it all is a daunting task, further complicated by the fact not all data carry an equal market impact. Hence, as an investor, time is precious and knowing where to allocate that time is a critical skill. Proper filtering not only saves you from a time-sucking chore, but it may also improve investment decisions—win-win! So in the spirit of helping you win back some valuable time, here are some widely covered datasets we believe aren’t as crucial as many think.

First a quick note: None of what follows is to argue you should ignore these factors in forming an assessment of the economy. Rather, we would suggest making note of them, but only allocating them significant time if there is something very unusual—a random spike or fall out of step with the recent trend. Also, if you are analysing individual securities or sectors, some of these may be more (or even less!) important.

Producer Price Inflation: Input prices (for materials and fuels) and output prices (for goods leaving the factory gate) supposedly track how businesses pass costs from one stage of production to the next, giving an early glimpse into corporate profit margins and consumer price inflation. But inflation is always and everywhere a monetary phenomenon—prices increasing broadly across the economy. Raw material and intermediate wholesale prices are likely to manifest in higher consumer prices only in areas where producers have pricing power. Therefore, so-called cost-push inflation is a misnomer. How much money chases goods and services drives inflation, not what upstream vendors charge each other. For this reason, if inflation is what you are trying to monitor, we believe you are better served watching money supply measures, like the BoE’s M4 excluding “intermediate other financial corporations”—broad money supplied to the private non-financial sector—or perhaps M4 lending (ex. IOFCs). Now, there may be some influence on profit margins from producers’ costs rising, but using PPI as a proxy isn’t going to yield much insight.

House Price Indexes: We get the obsession with keeping track of real estate values—especially for homeowners, landlords or those looking to buy—but except for specific housing-related equities, the various house price indexes (Halifax, Rightmove, RICS, Nationwide and the ONS’s) are mostly superfluous for UK shares and their investors. Housing, housing finance and the assets backing them simply aren’t a major economic driver. Private residential construction, repair and maintenance account for only 2% of GDP. Some say overzealous housing finance drove UK lenders to ruin (and the UK into recession) in 2008, but in our view, the collapse of Northern Rock and UK recession were symptoms of problems that began in America and ultimately drove a global liquidity crunch and financial crisis. As for the supposed “wealth effects” some claim changing home prices cause, there really isn’t a meaningful relationship. We believe consumer spending is most connected to disposable income, not home equity or prices.

Private Retail Sales Gauges: The British Retail Consortium (BRC) and Barclaycard both publish independent reports on Brits’ spending habits with data gathered from their members or processed transactions, respectively. These are more timely than the ONS’s official report, but they aren’t necessarily more revealing. Whilst the BRC reports their monthly retail sales data a week or so ahead of the ONS, it is a limited dataset representing only about 60% of the sector and often diverges from the ONS’s reports as a result. For example, in the Brexit referendum’s wake, BRC reported August 2016 retail sales fell -0.3% y/y, but the ONS reported retail sales up 6.3% y/y. (Sometimes BRC overestimates, too.) Barclaycard is also limited, capturing less than half of all credit and debit transactions in the UK. Moreover, both report only year-over-year changes, which can obscure developments in the month-over-month results.

Construction Purchasing Managers’ Index: PMIs are valuable, but we’d suggest not spending too much time with the construction one. Construction represents only a small slice of the UK’s economy at 6%. Sure, the construction PMI matters for builders, but as a whole, the manufacturing (14% of GDP) and services (79%) PMIs provide a much broader look and, hence, are more useful indicators in this regard.

Consumer Confidence Surveys: Polling people about how they feel may seem fruitful, but we believe it provides little forward-looking insight. People often say one thing and do another. Actual retail sales often contradict the surveys. Consumer surveys hint at overall sentiment, but that is a coincident indicator at best.

Claimant Count Change: The ONS reports monthly figures of those claiming unemployment benefits—mainly Jobseeker’s Allowance, but also those newly out of work receiving Universal Credit. Some use it as a timely economic gauge for clues on the economy’s direction, but like other employment measures, it lags. Layoffs are a symptom of a weakness, not a cause or leading indicator, so a new trend in claimant counts won’t tell you anything you didn’t already know.

Productivity: “Living standards can’t improve unless productivity rises” is a popular argument these days, and efficiency gains are a powerful long-term driver. Hence all the concern over Britain’s supposed productivity crisis. But they are impossible to measure with classic productivity statistics, which basically measure output per unit of labour. That isn’t forward-looking—it is a mashup of a backward-looking measure and a late-lagging one. Whilst measured productivity has been historically low this business cycle, profits and profitability have been high. Which do you think equities pay more attention to?

There are, of course, many more. To size them up on your own requires weighing the methodology, history and logically assessing the connection to markets. Ask: Is this the only—or most complete—measure? Ultimately, if the data you are sizing up don’t pass that logic test, we would suggest you may be better served spending the bulk of your time elsewhere.