which doesn’t actually happen. The means of growth Mill describes in the paragraph quoted is the only kind that appears in the record.
Evidence from Stock Markets
Market-valued capital, reported in national accounts since 1990 or so and assembled at the convenient Piketty-Zucman website, is measured by a common standard in principle. Measurement begins with stock markets. It should. The stock market is the most exact source of economic information that I know. With due reservations about connivance and “stale prices,” meaning outdated prices from earlier days because the stock has not traded since, or anyhow not enough for confidence, we know pretty well what markets think stocks are worth from tick to
tick.
We would know better if markets were perfectly efficient. Proof that they aren’t shows in medium-term autocorrelation or trend. Autocorrelation (in price) is tendency for markets to be up tomorrow if up today, and down if down. Trend is a shorter word for the same. Perfect efficiency ought to show a “random walk” where prices change captures all current news, news captures reality without optimistic or pessimistic bias, and tomorrow’s price direction is as unpredictable as tomorrow's news. The only exception should be long-term uptrend with productivity gain through innovation. In this case it is not surprise in the news that brings growth, but gradual gain in present value as a foreseen better future is less discounted as it
draws nearer.
There is chicanery as well as inefficiency. Insiders, braving the legal risks, may take advantage of outsiders. But it is not clear to me that insiders are likelier to be sellers than buyers. National accounts follow prices of publicly traded shares collectively,
where some chicaneries should offset others.
Allowing for all this, I think national accounts are wise to accept stock prices as the
best measure of underlying assets. Intangibles such as patents or market advantages
Chapter 4 Mill’s Idea 1/11/16 19
HOUSE_OVERSIGHT_011010