THE ECONOMIST: Stockmarket data has never been easier to access but at what cost?

Sometimes efficiency is obvious.

On a production line for, say, chocolatey treats, it is a series of whirring, specialised machines busy enrobing a biscuit in caramel, covering it in chocolate, and drying, packing and stacking the product.

For an office worker communicating with colleagues it probably involves email.

In both cases, the process has been made more efficient by technology. Across almost all industries the story, since the industrial revolution, has been one of tech boosting efficiency.

A new paper by Cliff Asness, an illustrious quantitative investor, suggests the stockmarket is an exception—a case that holds appeal today, given the market madness of August and the recent rollercoaster ride in the stock of tech giant Nvidia.

The reason for the stockmarket’s exceptional status is, in part, because market efficiency differs from production-line efficiency. An efficient market is one where “prices reflect all information”, according to Eugene Fama, a Nobel-prizewinning economist (and Mr Asness’s PhD adviser).

In the 1990s electronic trading was the norm, but it was a far cry from the nanosecond version that dominates today. There is also now more information about any stock, which is disseminated more quickly.

Speed, competition and greater information have clearly enhanced efficiency in one way: they have brought down the cost of trading.

The trouble is that speed is not precision.

“It’s hard to imagine new information doesn’t impact stock prices faster than in the past, and that is a kind of ‘efficiency’,’” writes Mr Asness. “But speed doesn’t imply the level of prices before or after the new information was particularly accurate.”

Indeed, he points to evidence that accuracy has fallen.

One source of evidence is so-called value spreads, which compare what investors pay for the priciest stocks to what they pay for the cheapest.

The simplest version just looks at the price of a firm compared with its book value (what it would be worth if it were sold for parts). Using this measure, value spreads were stable between the 1950s and the 1990s, before spiking ahead of the dotcom crash. They have since climbed steadily over the past 20 or so years to near all-time highs. The flaw with this measure is that it can be skewed by a changing market make-up (tech firms tend to have higher price-to-book ratios than banks, for instance).

Another, more sophisticated measure devised by Mr Asness includes all kinds of definitions of value by comparing prices with earnings, forecast future earnings, cashflows and so on, and only compares firms within an industry.

This shows a similar trend. However things are cut, it is much harder to find a reason why investors are willing to pay today’s prices.

Mr Asness suggests three explanations for the puzzle. One is that two decades of low interest rates messed with measures of value in a way that has been hard to capture. This argument is easier to dismiss now, with rates back at positive real levels, than it was in the 2010s.

A second is that index funds, which buy and hold the whole market, have crowded out smart investment.

Imagine the market before such funds was made up of “sharks” (informed investors) and “minnows” (dumb money). If sharks had given up and opted for index funds, indexing might have made it harder for pros to push prices in wise directions.

The third argument is the most compelling: social media produces mobs. Or, as Mr Asness puts it, “instantaneous, gamified, cheap, 24-hour trading … on your smartphone after getting all your biases reinforced by exhortations on social media from randos and grifters with vaguely not-safe-for-work (NSFW) pseudonyms…What could possibly go wrong?” This is an idea that resonates with others.

“For whatever reasons, markets now exhibit far more casino-like behaviour than they did when I was young. The casino now resides in many homes and daily tempts the occupants,” Warren Buffett has mused.

As Mr Asness admits, he is talking his own book: value spreads inform his investments.

Given these have become more inexplicable over time, his returns have been uneven in recent years. A world in which the investing masses are making big errors should in theory be a good one for sharks.

However, there is one big challenge: whether they can they hold their nerve in the face of obvious chaos. Nvidia, which is one of most valuable and most closely watched firms in the world, shed 10 per cent of its value—$US280b—on Tuesday. The cause? A softish manufacturing data release.