Standing up economic, market data via alternative sources
The rise of what was known as the ‘alternative comedy’ scene in the 1980s prompted a common retort among its critics. “More like alternative to comedy,” they would sneer.
Three decades later, so-called ‘alternative data’ are becoming more prominent on the investment scene. Perhaps you have heard of hedge funds buying satellite photos of car parks for more insight into retail trends.
So the question remains: is this just useful information, or is it really an alternative to data? So great for a story but not much else.
With any new dataset, of course we need to be particularly careful of random noise masquerading as meaningful signals. But with that caveat, we do find that alternative data can complement our analysis based on more traditional sources, whether by confirming or challenging consensus views.
At the start of the pandemic, for instance, we turned to academic research to test prevailing narratives. Then we harnessed mobility data from Google and Apple to help us look through the official rhetoric on lockdown restrictions and see what was really happening on the ground. Similarly, Emiel has created his own set of bubble indicators that can give us a different perspective on equity valuations.
Jobs for the buoys
Looking ahead, another area in which alternative data can inform our understanding is in employment statistics. Tim, our head of economics, expects 9-9.5 million US jobs to return by the year’s end. Although the pattern for labour-market growth is highly uncertain, we would not be shocked if this week’s payrolls exceed one million.
As evidence of how quickly employment can rebound, Australia is way ahead of even its central bank’s most optimistic scenario. But equally we know any individual set of numbers could surprise to the downside too.
These have revealed some encouraging developments. There’s been a big pickup in job posts for lower-skilled workers and in the virus-hit sector of leisure and hospitality, for example. Revenues of small businesses are also close to their highest level since July 2020.
Our long equity position already reflects what we have been seeing, but in fixed income these trends suggest we should be aware of the risks around the upcoming payrolls number beating expectations. Investors have already demonstrated their sensitivity to the prospect of rate hikes.
Our core forecasting models will continue to rely on the long timeseries of official data, but big-data approaches like these can help to ensure we don’t miss the boat on wider trends due to economies re-opening.