Disaggregating the investment drought

You can click twice to see this Excel file on investment.

That is from Spencer England of Spencer England’s Equity Review.  He writes to me:

I think a major reason for the difference between net and gross investment is the growing share of Info Tech (IT) in business fixed investment. IT now accounts for over half of real business investment. But IT has a much shorter life span than traditional business equipment or investment in structures.

It is like going up a down escalator –you have to run harder just to stay even.

Here is a previous post on the investment drought, with another good picture.  Karl Smith offers some related remarks, though I do not focus on transportation as he does, rather given his framing I would put more emphasis on our inability to replace industrialization with something comparably important.


I'm not too sure what the argument is here -- presumably if IT is turning over rapidly, you gain something every time you turn it over. That's sort of counter stagnation, isn't it?

At the margin, the return is not so high as to boost net investment in a satisfactory fashion.

So what do we want here -- results, or "net investment"? Presumably the turnover occurs when the marginal return is high enough to justify it.


Warren Buffett described investment in the textile industry like watching a parade. Everyone stands on their tiptoes but noone gets a better view. It may be better than not standing on your tiptoes, but not better than if you didn't have to stand on your tiptoes.

+1. This strikes me a bit like saying "GDP is not growing in a satisfactory fashion because of all those imports" -- trying to track prosperity through an accounting identity like GDP obscures the value of imports to American consumers.

+1 also to Bob below.

There's a market niche for supplying IT stuff that would have a longer useful life.

I will be more blunt in my comment this time -- what, exactly, is wrong with shorter-lifespan investments?

Today we don't need much investment in railroads or big hydro or other projects with a 50-year payback. The most cogent argument against new nuke plants is the high probability of economically viable alternative energy sources during their lifespan, turning them into white elephants. We could address our congestion issues with ribbons of concrete designed to last for decades -- or we could do so more cheaply and effectively with congestion management technologies which will get changed and improved every few years. Yup, IT beats concrete.

This is not stagnation. This is progress.

Bob, I agree with you.

Yesterday the World Bank issued its latest big report on China with this recommendation:

While providing relatively fewer “tangible”
public goods and services directly, the government
will need to provide more intangible
public goods and services like systems,
rules, and policies, which increase production
efficiency, promote competition, facilitate
specialization, enhance the efficiency
of resource allocation, protect the environment,
and reduce risks and uncertainties.
(ps. xv-xvi Executive Summary).

I fail to see how forced obsolescence is the equivalent to increasing flexibility of capital. When you replace your network infrastructure not because it no longer supports the traffic you're pushing, or because it doesn't support some of the protocols you might add over the next 5 years, but because Cisco has determined that equipment is end of life/end of support, you aren't getting much or any marginal benefit from the investment.

I only have my own anecdotal experience of this type of investment, but that experience tells me than IT managers are VERY risk averse and will heavily overinvest on features that they will never be able to use, and that forced obsolescence is definitely a factor in IT investment. It's not the whole story, but it tells an appreciable fraction of it. I would be interested in seeing empirical work on just how much of a fraction it accounts for.

Methinks he doth disaggregate too much. Or just misaggregate? Is the shift caused by buying modern vs. traditional equipment and software? Couldn't that shift be discerned in any period?

The big long-term secular shift is just the increase in equipment and software relative to structures:


Resulting, not surprisingly, in capital consumption constituting a larger portion of gross investment -- in pretty much every sector (*especially* defense, since we've stopped building bases), but notably here, in the business sector:


Does anybody know of any work on the implications, regarding investment data, of moving to a services-based economy?

I have long wondered whether investment in services industries (rent an office and hire some people and endure operating losses for a while) is recorded as investment in the data everybody looks at. So we might be looking at the data thinking we have declining rates of investment, whereas in fact investment is shifting to a form that does not show up in the data. But this is just a guess / thought. Can anybody set me straight?

All these "insights" are, of course, plausible.

But, "investments" gross and net, for poduction objectives are determined by motivations, most of which appear to be utilitarian.

Are we not failing to consider a prime factor in those motivations - the anticipated return on such "investing?"

Some is necessary to continue current activities at current levels (upgrades for competitive reasons included); some to expand or improve production; but all with the same base anticipation and risk concerns.

Consider the effect on anticipations (expectations?) indicated by the fact that economy-wide return on invested capital has decline over 25%; from about 6.2% (probably higher for production assets) in 1965 to around 1.3% (possibly 1.6% for production assets) in 2010 (and currently??). Thus motivations to invest may be mitigated.

The current rate of return on net invested capital is probably matched by the rate of price-inflation, which, if coupled with acceleration in obsolescence, may exceed net returns, creating a "mood" that dampens motivations in managerial capitalism with its "performance" objectives.

The result is to retain surpluses in a greater liquid (including account receivable) proportion; including cash.

Why invest for production at risk for marginal rates of return and no improvement in "performance?"

Comments for this post are closed