Thanks for that.
I'm glad to see that you realise the confusion you could have inadvertently caused about this issue and that you do indeed agree with me that the range of outcomes increases with time.
We need to remember that it is the outcome that the investor gets and has to live off that matters, not the average of the range of returns over the period of time of the investment.
If risk were simply volatility of returns then it would be correct to say that this reduces with time, however if risk is viewed as the volatility or uncertainty of accumulated wealth, or what I would call outcome, then suddenly risk increases with time.
Furthermore, if risk is viewed as the impact of a shortfall in wealth, as it would be in saving for the university fees in Poppy's question for example, then not only does risk increase over time, but most measures of risk tend to further understate just how severe a 'bad' shortfall can be and how longer time periods increase the opportunity for ever more severe shortfalls, especially when they are compared to an alternative such as just investing in a 'risk free' asset.
It would be correct to say that the range of the average annual compound growth rate or return reduces with time, but this is very different to the range of outcomes that the investor could experience.
So, by risk I mean the possibility of a worse than expected outcome and, as I will show, the range of outcomes is greater as the time increases so the possibility of a worse outcome increases with time (as does the possibility of a better outcome), so the 'risk' the investor experiences increases with time.
For example over the long term of 54 years for which we have reliable data (1956 to 2009 inclusive) the average return from UK equities is about 12%.
The real (less inflation) return from UK equities is about 6%. Remember, it is the real rate of return that the investor can 'eat'.
However, the range of real returns in any one year over that period is from PLUS 101.3% (in 1975) to MINUS 59.4% (1974 and that was after a real loss of 35.5% in 1973).
To recover from a real loss of 59.4% you need a return of 146.31%, which is 45% more than has ever been achieved in the last 54 years.
Yet, on average from 1974 to now the real return from equities has been 6.7% per annum.
This means that the difference between the 'expected' return on a £100 investment in 1974 and the actual outcome is £14 or 14%.
From 1973 it is even worse. The actual real outcome was that £100 invested in 1973 became £516.68, but the expected outcome, given that the average return over the period was 5.2% per annum real, was £652.50, so in real terms after adjusting for inflation, the investor was £135.81 pounds worse of on their investment of £100 than an adviser would be telling them if the adviser was to say that equity investing outcomes or risk reduces with time.
Clearly, if we picked different start and end years the result would be different, but the point remains that the range of outcomes or results increases with time rather than decreasing. If it is the outcome or result that matters to the investor, which I would suggest it is, then an increased range of outcomes would indicate an increased risk would it not?
I'd like to see a financial adviser try and justify that kind of error in advice to the Financial Services Authority or any other regulator.
For a more detailed and better explanation than I can give here see
"Is stock Investing Really Less Risky In the Long Run?"
By Michael Kitces | 12.16.10 @ 13:59