What makes traditional statistical indicators unreliable for investors?

Investors have relied on data and statistical indicators for decades, but they are no longer useful

Helen Thomas says it's time to move on.

To successfully navigate financial markets, one needs a solid toolkit. Sadly, the pandemic revealed the flaws in even the longest-running data sets. We must acknowledge that our statistical measures have not been able to keep up with the rapid advancements in technology, and we must modify our strategy accordingly. The time has come to switch from log tables to GPS.

During yet another round of his reality-television show The Apprentice, US President Donald Trump shocked the typically stoic world of statistics by firing the US Commissioner of Labor Statistics. But for a while already, the US Bureau of Labor Statistics (BLS) had been having difficulties. In the past 18 months, it has "inadvertently" released inflation data ahead of schedule, posted the annual employment revisions late, and alerted certain financial institutionsreferred to as "super users" by a BLS employeeto information in the data that others might have overlooked. Even before Trump's comeback, it was a turbulent period.

Change is something the president is at least starting to understand. Due to Sir Ian Diamond's early resignation due to health concerns, the UK is still without a permanent "national statistician." Additionally, since the Labour Force Survey (LFS) was discontinued in October 2023, the UK has been without its primary employment indicator. It may not be ready for replacement until 2027. Unsurprisingly, the monetary policy committee "continues to place less weight" on the official LFS data, leaving the Bank of England to assemble its own version.

It all boils down to the data's collection method. After sending a letter to each household, a phone call or in-person interview is conducted. This caused the response rate to drop as low as 17% during COVID, but it hasn't recovered sufficiently to guarantee the accuracy of the data. Who in the modern world would answer a phone call, much less a letter?

A minor error in judgment can have serious consequences.

It's not just citizens who are unreliable. Due to a mistake in the vehicle excise duty provided by the Department for Transport, the Office for National Statistics (ONS) had to declare that the April UK CPI data was inflated by 0.1 percentage points. In their first year of registration, someone had inflated the number of cars that were subject to this duty in some spreadsheet. Our navigation system may chart the incorrect path due to such minor errors, and fortunes may be won or lost. Because of completely inaccurate information, the yield on short-dated index-linked gilts moved by several basis points. The volatility of financial markets is already high enough without the addition of purportedly trustworthy data.

Statistics are meant to serve as staging posts that help us determine the direction of the future. But what if we are staring at the wrong map? Over the past five years, the world has undergone a radical transformation. A multi-decade "technological revolution" was already underway when COVID struck, sharply altering the course of events. Working from home and interacting with people virtually prompted the deployment of resources into new technologies like artificial intelligence. Every advancement in computing power has allowed us to advance in our evolution.

Monitoring metrics like the manufacturing cycle is less logical in the modern world. Because it has been available since 1945 and has shown a strong correlation with the US business cycle, investors have long referred to the US ISM Manufacturing data as a bellwether for the US economy. Since economics is a social science, it is impossible to replicate experiments in a lab setting; however, an eight-decade survey is as good as it gets. Until a pandemic that only occurs once in a hundred years forces the economy to close and then reopen.

The ISM Manufacturing number fell to a 12-year low in a single year before rising to a 40-year high. A survey isn't a precise indicator of activity; it's just a sentiment indicator. When the fear subsided and vaccines became available, businesses went from looking at disaster with astonishment to an overwhelming sense of relief. No business cycle existed, and the transition from bust to boom was not seamless. It was merely unexpected. Very little was revealed to us by the ISM data.

The effects of that shock are still being felt as the economy comes to terms with the post-Covid equilibrium. As technology spreads intangibly among consumers worldwide, physical borders have been reimposed. Aging nations' indebted governments are finding it difficult to allocate ever-diminishing resources to growth-promoting initiatives. Inflation or unemployment as a basis point no longer accounts for the marginal difference. It is becoming more difficult to identify the future due to the outdated data that we have been using for decades.

Technology should facilitate the rapid, transparent, and equitable updating, gathering, and distribution of information. Data on Google mobility, aggregated credit card spending, job postings on Indeed and LinkedIn, energy consumption, or even satellite imagery of parking lots and warehouses could provide real-time updates. It should be relatively easy for tech companies to publish aggregated activity data if we can order food, taxis, and romantic partners from an app in our palms. We will all be at sea if we stop believing statistics. We can re-calibrate thanks to technology before we veer too far off course.

The CEO and founder of the macroeconomic consulting firm Blonde Money is Helen Thomas.