What does it mean that ROA has declined since 1965?

Deloitte has been getting attention to its 2009 Shift Index report. The index is hoping to become the 21st century version of the Composite Index of Leading Indicators. Although they do not put it in these terms, the indices are attempting to measure the knowledge economy. There are three main indices made up of a total of 25 indicators. The indices attempt to measure:

• Foundation: computing capacity and networking bandwidth
• Flow: internet activity, knowledge flows and movement of people
• Impact: labor, corporate and brand performance

There is a lot of interest here. But in the marketing of this report, the factoid that I have seen reported from a lot of corners is that since 1965, long-term corporate performance has declined dramatically. The calculation is based on corporate return on assets (ROA) which they calculated as net income over total assets. Deloitte says they chose this asset performance as a proxy for corporate performance for two reasons:

1. ROA is a measure of firm profitability without distortions associated with capital structure
2. ROA “takes into account asset investments, whereas other measure, like return on sales, do not.”

The data using these figures shows an overall ROA decline from 4.7% in 1965 to .5% in 2007. The averages mask the fact that companies in the top quartile faced a slight decline from 12.9% to 11% while the bottom quartile went from 1.2% ROA in 1965 to -14.7% in 2008. Given that corporate tax rates also declined in this period, the study asks whether the story is even worse.

While this is all interesting, I cannot help but question whether ROA is the right measure of corporate performance given what we in the intangible capital community know about the distortion of financial statements by the failure to report intangibles investment.

Corporate assets as reported in the Compustat data used by Deloitte include only tangible assets. Yet we know from macroeconomic data that corporate spending on intangibles far exceeds investments in tangibles (for 2007, $1.6 trillion in intangibles, $1.2 trillion in tangibles).

This intangibles investment is what has created the kind of things the Shift Index is trying to measure: computing, networking and human knowledge capability.

So if you really want to understand the asset performance of today’s corporation, you really need to think about this expanded view of assets. But trying to reason this out, I am frankly left puzzled. If we had the data to adjust the ROA numbers for intangibles, here is what we would see:

• If investments in intangibles were not expensed (as they currently are), income would go up
• But investments have to be amortized so the amortization would lower income. This would probably not offset the increase from the adjustment based on removing investment expenses (per the previous point) because these investments have risen over the past decades. So I assume incomes might increase slightly.
• If investment in intangibles were capitalized, then assets would go up.

In other words in the ROA calculation of NI/Assets, the numerator would probably go up slightly but the denominator would go up by even more. This means that the ratio would be even lower than reported by Deloitte. Does this mean that all the technological gains of the past decades have just lead to greater competition and lower returns? Are we missing something here?

One of the authors of the study, John Hagel, says in an interview with ReadWrite that scalable efficiency was the model of the past (hence the ROA) and that scalable learning is the goal of the future. He does not offer a specific metric, presumably the shift index is part of the answer–but the index is focused on the macro more than the firm level.

I would love to get some discussion going on this issue. What does this mean? Is this a relevant discussion? If so, why? If not, how should we be measuring corporate performance?