Google’s share price has needed a little support lately. On Thursday night it got it, from the body that has arguably become the most powerful on Wall Street: the selection committee that decides the membership of the Standard & Poor’s 500 index.

The S&P is the most closely watched index of the US market. So much of US investors’ money is now tied up in index funds that joining the magical club of 500 instantly obliges managers to go out and buy the stock. According to Goldman Sachs, the 23 companies to join the S&P in the 12 months before Google have enjoyed an average pop of 2.9 per cent compared with the index.

Google, typically, did rather better, gaining 7 per cent on Friday’s opening even though it will not officially join the S&P until the end of next week.

Google’s rise in turn was enough to add almost 4 points to the Nasdaq 100, the second most widely followed index, of which it is already a member. That was more than triple the impact of any other stock on the Nasdaq 100.

So thanks to S&P, Google is now suddenly back to the levels it had reached before Wall Street’s absurd overreaction last month to its chief financial officer’s unwisely honest musings about how the “law of big numbers” would begin to reduce its growth rate. The law – that percentage growth rates are harder to sustain as a company gets bigger – is immutable but still less powerful, apparently, than the selection committee at S&P.

Friday’s incident also throws a spotlight on to the indices themselves. Although indices such as the S&P 500 and the Dow Jones Industrial Average are treated as though they are objective benchmarks, they are not.

Membership of both the S&P and the Dow – the most famous US index, though not one that is used much by index funds – is determined by committee. Changes are made at irregular intervals, and often opportunistically in response to a corporate event. For example, Google gets its chance thanks to the departure from the index of Burlington Resources following its takeover by Conoco.

Given the role of human judgment, the indices themselves can be second-guessed. There are large Wall Street research teams that do nothing but try to predict the next new entrant to the big indices.

We also begin to meet methodological objections. One is that by focusing on market valuations, as the S&P 500 does, there is an overweighting for overvalued companies. If the choice comes down to two companies with the same book value, the one with the higher price-to-book ratio will win.

Therefore a market-value-weighted index must be biased against value stocks, so called because they look cheap compared with their fundamentals.

Another problem with indices as a gauge of the overall market is survivorship bias. When companies run into trouble, as happened last year for example to Delphi, the giant auto components company that went into bankruptcy, they drop out of the index to be replaced by better-regarded stocks.

Take these objections on board and the market looks different. The Dow is nearing a six-year high and is not far from at last regaining its all-time high set in January 2000, just before the internet bubble finally blew up. This helps portray US equities as recovering, but not back to their peak and arguably still caught in a bear market.

Other indices tell a different story. The NYSE Composite, covering all the 2,700 stocks quoted on the New York Stock Exchange, has been setting regular new highs of late. The same is true of the Dow Jones Transportation Average and of the Russell 2000, mapping smaller companies.

So are things better than the Dow and the S&P tell us?

Russell indices, originally intended as a tool for fund managers, are engaged in a feisty battle against S&P. They are reconstituted annually and determined on the basis of market value, with no human intervention.

The biggest 1,000 stocks are in the Russell 1000 (where Google has been for many months), and the next 2,000 are in the Russell 2000. Combine the two and you have the Russell 3000, which has a claim to be the most objective gauge of the overall market.

What does this tell us? Unfortunately the Russell 3000 is still 13 per cent below its peak.

Copyright The Financial Times Limited 2018. All rights reserved.

Comments have not been enabled for this article.