After software engineering and financial engineering comes linguistic engineering. Google this week raised its market capitalisation by $25bn by shuffling around some executive jobs and changing its name to Alphabet. Who knew that swapping your tiles in a game of corporate Scrabble was worth so much?

Everyone reads what they want into the new letters. For Larry Page, Google’s restless co-founder, Alphabet means jettisoning the cares of running a corporation and becoming a full-time inventor and venture capitalist, while Sundar Pichai takes the leadership of Google. For employees, it brings the hope of more valuable share options. For Wall Street, it spells clarity.

Only governance renegades would invent a structure with one board for Google and Alphabet, the founding triumvirate — Eric Schmidt, Sergey Brin and Mr Page — stacked above Mr Pichai, and Ruth Porat as chief financial officer of both. “Google is not a conventional company,” wrote Mr Brin and Mr Page in their 2004 founders’ letter, and by heavens they meant it.

Still, being conventional is not the best way to build an innovative business or to make profits. Warren Buffett runs a unique combination of industrial conglomerate and investment fund at Berkshire Hathaway, and it has worked well for him. He made his largest ever acquisition this week, buying Precision Castparts for $32bn.

Berkshire and Alphabet are different kinds of businesses. Mr Buffett values cash flow and mature brands; Mr Page prefers to create things. One of the purposes of this week’s reshuffle is to prove to investors that not as much as they fear is being spent on experimental start-up projects, such as Project Loon’s high-altitude balloons providing internet access to remote areas.

Mr Page’s naming of Mr Buffett as a role model in providing “long-term, patient capital” to an array of businesses is not idle. He thinks that a multi-business group with a guiding intelligence at the centre can beat the single-sector company favoured by investors. The “conglomerate discount” applied by Wall Street can be defeated.

In principle, that is an odd thing for Mr Page to believe. Google’s technology, after all, uses online auctions and markets — the wisdom of the crowd, not human curation. Why should conglomerates such as Alphabet, with their entrenched interests and fiefdoms, be better than capital markets at allocating capital efficiently? Does he trust in inside knowledge only when the insider is himself?

But he is right. Conglomerates can outperform when they exploit their advantages and remain disciplined rather than falling prey to empire-building. Their ability to build a cadre of skilled managers and to pick the right investment projects is strongest in research-intensive industries that invest in intellectual property, which is Alphabet’s territory.

Neil Bhattacharya, a professor at Southern Methodist University in Texas, found in a study that multi-business companies ran operations more efficiently than single-sector ones. They had particular advantages in areas such as software and life sciences because managers could judge more accurately than stock markets which projects were likely to succeed.

This is counterintuitive, given US investors’ liking for simplicity, and view of conglomerates as inefficient. Public conglomerates in the US are valued at discounts of 10 to 15 per cent to single-sector companies, according to Boston Consulting Group — though the discount is lower in Europe, and Asian conglomerates often trade at a premium.

The suspicion originates in the 1970s and 1980s, the era of companies such as ITT and RJR Nabisco. Michael Jensen, a Harvard professor, later criticised the “billions in unproductive capital expenditures and organisational inefficiencies” at conglomerates, praising the trend toward “smaller, more focused, more efficient” enterprises.

Big corporations remain prey to temptation. Boston Consulting Group found that the conglomerate discount is partly due to conservatism. They tend to invest heavily in their original businesses, which may be stagnant or in decline, while undervaluing newer divisions with more potential. Microsoft, for example, suffered from trying to reinforce its Windows franchise.

Yet even investors who are suspicious of quoted conglomerates delegate capital allocation and management oversight in private markets to informed insiders. Venture capital and private equity funds are both forms of conglomerate — they invest capital in a broad portfolio of businesses on behalf of outsiders who believe that such funds possess superior expertise.

Why, though, should investors seeking exposure to new companies buy shares in Alphabet, which then channels Google’s surplus cash into its own venture and growth funds, Project Loon, self-driving cars and life sciences? They could instead invest money directly in a venture capital fund. Why take the longer and less-direct road?

It depends on trust. Investors could also have bought shares in Precision Castparts last week for less than Berkshire Hathaway paid this week, but they do not complain because they trust Mr Buffett. Alphabet’s shareholders must believe in Mr Page and Mr Brin’s ability to use their intelligence and avoid the traditional pitfalls.

To judge by the shares this week, they prefer a conglomerate called Alphabet to a company that had not made plain what it was. Strange as it seems, it is a rational choice.