The toughest questions in investment are not those that challenge specific views, but those that challenge deep-seated assumptions.

Markets exist to accommodate a range of participants with divergent views or economic interests, so it is hardly a surprise that almost any position can be justified somehow. Those who judge the position right are rewarded, while those who do not are penalised.

But that is not how the most grievous or most costly mistakes are made. Those arise when it is investors' entire belief systems that turn out to be misplaced. Long Term Capital Management, the hedge fund that had to be bailed out in 1998, is a good case in point. Clever to a man, the principals lost their business because their faith in historical relationships that had worked so well for many years turned out in practice to break down during a period of extreme market stress.

The same, on an even bigger scale, goes for central bankers who bought in, naively, to Alan Greenspan's convenient view that bubbles in financial markets could not be identified in advance and even if they could, would prove more costly to pre-empt than to clear up after they had burst. As he confessed to Congress in October 2008, bankers' behaviour during the crisis had revealed "a flaw in the model?.?.?.?that defines how the world works".

The cost to the world of the chairman of the Federal Reserve's faulty assumption now runs into billions.

Are we approaching the point with government bonds where the assumptions that have carried this once-derided instrument triumphantly through three decades of consistently good returns need to be discarded? If we are not there already we are surely not far away.

A recent report by Andrew Smithers posed the question bluntly: Bonds - Government and Corporate, Nominal and Real - Why Should Anyone Hold Them?. His argument is that at current levels government bonds are "seriously overpriced" and therefore high risk. Returns are likely to be negative in the short term as the twin props of quantitative easing and bank window-dressing are withdrawn. Yields on inflation-linked bonds meanwhile have been driven so low by investors seeking insurance against a resurgence of inflation that, in his view, they are set to do badly whether or not inflation picks up.

David Fuller's view Regarding Alan Greenspan's testimony before Congress in October 2008, concerning: "…a flaw in the model…that defines how the world works", one could respond with the rhetorical question: What else is new? I propose a new mantra for students of the markets:

Financial models do not age well.

This is because they are based on back-testing, and circumstances eventually change.

It is understandable that bonds did well in the lengthy disinflationary period commencing two years after Paul Volcker's appointment as Chairman of the US Federal Reserve in August 1979. He was nominated by President Jimmy Carter as someone capable of tackling the USA's runaway inflation. Volcker achieved this with sky-high interest rates which peaked for 10-Year Treasury yields just below 16% in August 1981. We do not have that much back history for UK 10-Year Gilts but they followed a similar trajectory.

Over the last two years the Federal Reserve led by Ben Bernanke, the BoE and most other central banks have deployed Quantitative Easing in their efforts to prevent a Japan-style deflation from taking hold. Only if their efforts have failed are we likely to see a further decline in long-dated government bond yields which remain historically low in OECD countries. However if the CB's efforts to avoid a destructive deflation - defined as falling output, prices and profits - have succeeded, it is likely to be at the cost of future inflation.

In the aftermath of 2008's dramatic credit cycle meltdown, it is very difficult to get policy just right, not least because no central bank can control all of the influential factors. For instance, what effect will Asian-led economic growth have on the prices of commodities? If CBs succeed in avoiding a deflationary spiral, it will almost certainly be because they have erred on the side of future inflation.

A great deal of money has flowed into bonds over the last 30 years. In the US, bond funds also experienced net inflows of $36.6bn in 2008, versus equity fund outflows of $93bn, according to Barron's. A question for investors today: Is this now a contrary indicator and might equities outperform bonds over the medium to longer term? I think so, particularly if companies increase their dividends to attract retiring Baby Boomer investors hungry for yield. (See also last Wednesday's question: Do the two bear markets commencing in 2000 and 2007 make it more or less likely that we will have a third bearish event in the not too distant future?)