An interesting article by The Telegraph
The headline findings are reasonably familiar ones. If you had invested $100 in US equities in 1900, with reinvested dividends it would today be worth $22m, against just $191,000 for bonds. Expressed in real terms (after adjusting for inflation), the numbers are in some respects even more remarkable. Equities have shown an 851-fold increase, against just 7.5 for US Treasuries.
If these rates of return had proceeded evenly in a straight line, then investment would be a dull and predictable business. Tens of thousands of pundits, both gainful and more usually un-gainful, would be out of a job. Fortunately for them, equity prices are highly volatile, while even government bonds are subject to inflation, currency and sovereign credit risk. They are always on the move, generally in quite unpredicable ways.
Since the trough of the credit crunch in March 2009, UK equities have recovered 87pc, and emerging markets by a jaw-dropping 150pc. Rarely have rates of return been so good.
But these gains followed dramatic falls in the preceding year and a half, and in neither the US nor the UK have equities yet recovered to their pre-crisis highs. Worse, this is the second bear market for equities in less than a decade. Even with dividends reinvested, you'd be down on your money so far this century.
Bonds and cash, on the other hand, have done well. The bottom line is that equities are shot through with risk; there are plenty of years in which they return less than cash. To compensate for this risk, investors demand a higher rate of return, or an equity risk premium.
The question therefore arises of how long you have to hold equities to be certain that the premium fully compensates for the risk of a sudden loss of value. In his book Stocks for the Long Run , the US investment analyst Jeremy Siegel puts this timeframe at 20 years.
In other words, over any 20-year period you care to take in modern US history, equities will always do better than cash. The outperformance in some 20-year periods will be marginal, and in others very substantial, but his point is that on a 20-year view, there is no risk from holding equities at all.