High Frequency Economic Data takes centre stage as we try to understand the economic recovery

There’s a problem with official statistics from the ONS – it’s too slow. Important economic data is published monthly but the data lags by about six weeks. The most recent current GDP data, published on the 12^th August, covers the period up to the end of June. In usual years that’s not a problem – after all it only changes fractionally each month. Or rather it did – we’re in uncharted territory, time has accelerated and six weeks ago may as well have been last year for all intents and purposes.

Of course, demand to get data fast isn’t anything new. Anyone who read Flash Boys knows the story of the construction of a $300 million fibre optic cable from Chicago to New Jersey just to shave a few milliseconds in the transmission time of trading data. In that instance it was about getting the edge in high frequency trading.

High frequency trading might seem a world away from GDP data but there is a link: It’s the financial markets that want the information. If you can be first to accurately predict the winners and losers in the economy, you can place better bets on the market – and make more money.

Please step up High Frequency Economic Data. Otherwise known as High Frequency Indicators of Economic Activity (IEA). When the economy is changing daily, you need realtime indicators to tell you how people are actually behaving. How much are they travelling, eating out, going to work, consuming. The fact is we need to spend and earn our way out of this crisis, and so simple indicators are powerful.

So how can we get high frequency economic data? There are a number of online sources we can scour – flight activity, job listings, traffic congestion, energy usage are a few. The difficulty is interpreting conflicting data – yesterday The Times reported on data from Jefferies which said the economy is recovering strongly because (in part) traffic congestion is close to pre-Covid-19 levels in some areas. On the flip-side, the Citymapper Mobility Index tells us that London is only moving at 43% of usual.

The most interesting (and promising) source of high frequency economic data is satellite images combined with deep learning. If you can get it right (a big if), this is the modern-day equivalent of running a fibre optic line from Chicago to New Jersey.

A few years ago I attended a talk by Bird.i which collates satellite imagery. They presented a case study where they photographed the car parks of out-of-town shopping centres and counted the cars to see how many people were shopping. This can be extended to all sorts of areas – traffic, recreation, construction, mining. The sky’s the limit – literally.

The question is whether this data will ever become public. The company that discovers the golden goose is likely to keep it a closely guarded secret.

Perhaps they already are.