This is partly a result of the growth of big, intangible-based companies such as Google and Facebook but it is not just a tech sector phenomenon; intangible investment has been growing in importance in every sector.

This change matters because intangible capital has unusual economic properties. It tends to be scalable: a brand, an algorithm or a database can be scaled across a large business as easily as across a small one. And it is often a sunk cost: if a firm goes bust its operating processes and designs are generally worth much less to a creditor than tangible assets such as office buildings or vans. What is more, it tends to be poorly represented in the accounts of companies and national economies.

This has implications for three things that monetary policymakers care about: how the economy is measured; the transmission mechanism between policy and investment; and how inflation works.

Consider measurement first. When they set monetary policy, central bankers rely on good information about how the economy is performing, in particular on levels of investment. But if national statistical agencies are not including much of the investment that businesses do, central banks will have only an incomplete picture. National statistics offices now include research and software development as investment but they omit investment in design, training, branding and organisational development and reflect investment in data only partially.

Measurement questions will get more important as technology progresses. Take artificial intelligence: the AI used by a bank or an online grocer involves ever-faster computers running specialised software to create and interrogate vast databases. This is a combination of tangible assets (the computers) with intangible ones (software and data). Without accurate measurement of intangibles (and a way of accounting for the rapid price reductions of computing power) economic statistics risk missing the AI revolution entirely.

The second matter is that the transmission mechanism from monetary policy to business investment may not work the way they are used to. The weapon of choice of the central bankers is of course the short-term interest rate but business investment and consumption are determined by long-term interest rates, not short-term ones. In normal times short-term rates affect long-term rates directly. When they do not, say because lending markets are dysfunctional, central banks use “unconventional” monetary policy, such as quantitative easing, which can calm markets, lower risk premiums and thus reduce the rates that companies face.

These tools might be less reliable, however, in an economy where most investment is intangible. Because intangible assets represent sunk costs, banks are less willing to treat them as collateral than assets such as buildings or vehicles, which can be sold if the business fails. This means that intangible-intensive businesses generally rely more on equity finance or retained earnings for their investment than on bank finance. Indeed venture capital is a form of finance that has come into being largely to fund intangible-based businesses with little collateral but a potentially huge upside from the high growth rates that scalable intangible assets can provide. The availability of equity is related to short-term rates, via arbitrage and commercial considerations, but the link is harder to calibrate and perhaps less potent.

The final impact of the intangible economy on monetary policy relates to central bankers’ arch-enemy: inflation. The standard economic model of the inflation process is that prices start rising as the economy gets closer to capacity. This makes sense in an economy where most investment is tangible. Once a car factory approaches capacity and the flow of new vehicles is limited, car buyers will be more willing to pay more. The net effect is that prices will rise.

In an intangible-intensive economy things might work differently. Consider the video games industry. Once a company has written the code for a video game or app it incurs scarcely any more costs whether customers download a hundred, a thousand or a million copies. The “factory” never reaches capacity and the incentives to increase prices are lower. This makes intangible economies less inflation-prone as they expand. Instead, more companies will enter the market, perhaps trying slightly different products, willing to incur set-up costs in the anticipation of spreading them over a large market. Inflation may well be the same but if the variety of products rises, the measurement of inflation gets even more complicated.

If there were less inflationary pressure at capacity, interest-rate setters might be more relaxed about raising rates when they see demand rising. A less potent transmission mechanism, however, might make them eager for bigger changes when supply conditions change and, perhaps, more unconventional policy if that can affect the links from the policy rates to the rates the public face. All the while the difficulty of measuring intangibles makes it harder to diagnose investment, growth and inflation. The job of the central banker looks anything but boring.

Jonathan Haskel and Stian Westlake are co-authors of Capitalism Without Capital: The Rise of the Intangible Economy, Princeton University Press

Book them to speak at your next conference, email [email protected] – + 353 1 2354905.