Last week’s runaway winner in the bad ideas department was the Pension Benefit Guaranty Corporation. This quasi-official body, which takes over bust American pension funds, is switching the bulk of its assets from bonds into equities.

Quite a lot of investors are making that switch these days, notably the Sovereign Wealth Funds. It raises afresh a fundamental question: which, in the long run, is better value?

The PBGC is rather a special case, since it has a $14bn (£7bn) deficit and no means of closing it besides gambling. That is a pity, since it does not raise its pensions in line with inflation and is thus ideally suited to Treasury bonds.

Worse, equity bear markets often come with the kind of business downturns that drive companies to the wall and their pension funds into the PBGC – hence that $14bn deficit in the first place.

Still, needs must when the devil drives. The SWFs are under no such pressure. Nevertheless, they are working on the same implicit assumption – that stocks will always outperform in the long run, as they have in the past.

The timing could be better. According to the latest Barclays Equity Gilt Study, stocks produced a lower real return than Treasuries over the past decade in the US and the UK. The last consecutive decade in which that happened was 1967-77 in the US and 1927-37 in the UK – neither very encouraging precedents.

But in logic, it needs to be asked why stocks should outperform.

The equity risk premium, often cited as justification, is a slippery concept. It says investors will require a higher return ex ante, to compensate for higher risk. But if that return is correctly judged, the result should be neutral – the risk-adjusted risk premium should be zero.

That has certainly not been true in the past but the market may simply have miscalculated. If so, investors are capable of learning better.

The Barclays study, as it happens, has relevant thoughts on long-term prospects for each asset class. Without offering a clear preference, it argues that the game is changing radically for both.

The starting point is the familiar notion that the world’s resources are being consumed at an unsustainable rate.

The result, the study says, will be resource-based inflation, probably combined with lower growth and greater volatility in the economic cycle.

The present commodities boom is no doubt largely cyclical, to be followed by some kind of bust. But the underlying trend might be a different matter.

The China effect is crucial. In contrast to the shock increase in the world’s labour supply, which for some years proved disinflationary, the acceleration in per capita incomes points the other way. It produces a multiplier effect on consumption of raw materials, to say nothing of strains on the environment. The result – too much money chasing too few resources.

No doubt, there will be some trade-off with lower growth. For China and India to reach US levels of oil consumption per head would be literally impossible, since it would involve at least a trebling of global production. And if their copper consumption were to equal that of developed Asian economies, known reserves would be gone in a decade.

In both cases, reserves can certainly be extended but at ever-increasing expense. Over the past decade, it seems, new reserves of copper ore discovered were equal to those of the previous decade. But the amount of recoverable copper halved.

The result, Barclays argues, will be a hoovering up of investment by resource sectors generally, including water, waste disposal and alternative energy. In such a world, diversified portfolios may no longer make sense. In the 1970s, resource-based portfolios were alone in producing real returns. In future, industries such as autos and airlines will be gradually squeezed out by competition for funds from the resource sectors.

Meanwhile, equities in general will suffer from the higher risk premium due to economic instability, and bonds from higher inflation. Debt leverage will go into reverse, so the price of housing and other assets will fall. The banks will find basic lending much less profitable and will have to diversify – if memories of today’s credit disasters allow them to.

On the plus side, there may be fewer asset bubbles. In the recent past, as each bubble has burst the central banks have cut interest rates, thus ensuring the next one. Endemic inflation would remove that option.

As a version of the future, this will not win universal acceptance. I am not sure I accept it myself. But it does remind us of a universal truth: that while the past may be our best guide to the future, it is still a pretty bad one. Putting your faith in equities is all very well but you need to construct a case. Appealing to the record is no longer enough.