Buttonwood
Home on the range
A useful stab at projecting investment returns over the next decade
Jan 26th 2013
WHAT will be the level of returns from financial assets over the next decade? Answer that question correctly and your fortune is made. Andrew Milligan, the head of global strategy at Standard Life, an insurance company, has bravely undertaken this very challenge in a recent research note.
Such exercises are extremely useful, even though the projections are almost certain not to be exactly right. All too often investors (such as pension funds) are blithely optimistic that returns will match their goals. Forcing investors to spell out their assumptions can induce them to take a more realistic view.
.
The biggest mistake investors make is extrapolation. American pension funds often assume future annual returns of 7.5% or 8% because they have achieved those figures over the past 30 years. But the starting-point for such calculations is 1982, when Treasury-bond yields were in double digits and the dividend yield on US equities was 6.2%. Such returns are far harder to achieve when the starting yields on both assets are around 2%.
As Mr Milligan points out, the initial valuation is of little relevance to returns over the short term but matters hugely over periods as long as a decade. An object lesson in its importance was provided in the late 1990s when investors piled into equities at historically high valuations, without reflecting on the profit growth that would be needed to justify the fancy ratings. Returns in the first decade of the 21st century were duly dismal..
When it comes to cash and bonds the conventional approach is to assume that the future return will equal the current yield. But there is a clear disparity between the likely outcome for the two asset classes. With short-term rates close to zero, the current yield is the low point of the likely range for cash returns over the next decade. But with bond yields at 2% or below in America, Germany and Japan, current yields are at the high end of the range for likely bond returns. The risk is of a rise in yields to something like the historical norm of 4-6% and a fall in price that will translate into a negative return. Government bonds look like the least attractive asset to hold as a result (see table).
For equities the calculations are more complex. The simplest approach is to take the dividend yield and add the expected rate of dividend growth. The next step is to make an adjustment for valuation. You can assume either that the valuation will remain unchanged or that it will revert to the historical average. If reversion occurs then expected returns in expensive markets will be lower than the dividend-yield-plus-growth formula would suggest, and returns in cheap markets will be higher.
For the American market, the starting dividend yield of 2% handicaps future returns, although most people would chuck in a further 0.5-1% a year to allow for share buy-backs. Real dividend growth has averaged 1.4% a year since 1900 in America, and Mr Milligan suggests a range of 0-3%.
Excluding the effect of valuations, that gives a range of 3-6% for real returns. In fact, Standard Life posits a wider range of minus 1%-8%, suggesting the potential for a significant change in valuation.
But achieving the top end of this range will be difficult, given the starting valuation of the American market. At 22, the cyclically adjusted price-earnings ratio (as calculated by Robert Shiller of Yale University) is well above the historical average. When GMO, a fund-management group, conducted a similar exercise last year, it calculated that real returns from American equities over the next seven years would be zero if valuations reverted to the mean. European and British equities both have lower starting valuations than the US market and the potential upside is therefore slightly higher.
The surprise asset class of the next decade could yet be property, which became something of a dirty word after 2007. Rental yields are 6% or so in the main markets, an attractive income stream when European corporate bonds yield just 4%. Property will benefit if developed economies recover; and it is usually a hedge against inflation. Standard Life’s projections for real returns from American property are slightly above its equities estimates; European property looks even more appealing.
It would be foolish to be too precise. But assume for the moment that real returns are in the middle of Standard’s range and investors put half their money into equities and split the rest between cash, government bonds, corporate bonds and property. A diversified portfolio of this sort would return anywhere between 2.2% in real terms a year in America to 3.3% in Europe. Respectable enough, but not as much as investors are probably counting on.
The biggest mistake investors make is extrapolation. American pension funds often assume future annual returns of 7.5% or 8% because they have achieved those figures over the past 30 years. But the starting-point for such calculations is 1982, when Treasury-bond yields were in double digits and the dividend yield on US equities was 6.2%. Such returns are far harder to achieve when the starting yields on both assets are around 2%.
As Mr Milligan points out, the initial valuation is of little relevance to returns over the short term but matters hugely over periods as long as a decade. An object lesson in its importance was provided in the late 1990s when investors piled into equities at historically high valuations, without reflecting on the profit growth that would be needed to justify the fancy ratings. Returns in the first decade of the 21st century were duly dismal..
When it comes to cash and bonds the conventional approach is to assume that the future return will equal the current yield. But there is a clear disparity between the likely outcome for the two asset classes. With short-term rates close to zero, the current yield is the low point of the likely range for cash returns over the next decade. But with bond yields at 2% or below in America, Germany and Japan, current yields are at the high end of the range for likely bond returns. The risk is of a rise in yields to something like the historical norm of 4-6% and a fall in price that will translate into a negative return. Government bonds look like the least attractive asset to hold as a result (see table).
For equities the calculations are more complex. The simplest approach is to take the dividend yield and add the expected rate of dividend growth. The next step is to make an adjustment for valuation. You can assume either that the valuation will remain unchanged or that it will revert to the historical average. If reversion occurs then expected returns in expensive markets will be lower than the dividend-yield-plus-growth formula would suggest, and returns in cheap markets will be higher.
For the American market, the starting dividend yield of 2% handicaps future returns, although most people would chuck in a further 0.5-1% a year to allow for share buy-backs. Real dividend growth has averaged 1.4% a year since 1900 in America, and Mr Milligan suggests a range of 0-3%.
Excluding the effect of valuations, that gives a range of 3-6% for real returns. In fact, Standard Life posits a wider range of minus 1%-8%, suggesting the potential for a significant change in valuation.
But achieving the top end of this range will be difficult, given the starting valuation of the American market. At 22, the cyclically adjusted price-earnings ratio (as calculated by Robert Shiller of Yale University) is well above the historical average. When GMO, a fund-management group, conducted a similar exercise last year, it calculated that real returns from American equities over the next seven years would be zero if valuations reverted to the mean. European and British equities both have lower starting valuations than the US market and the potential upside is therefore slightly higher.
The surprise asset class of the next decade could yet be property, which became something of a dirty word after 2007. Rental yields are 6% or so in the main markets, an attractive income stream when European corporate bonds yield just 4%. Property will benefit if developed economies recover; and it is usually a hedge against inflation. Standard Life’s projections for real returns from American property are slightly above its equities estimates; European property looks even more appealing.
It would be foolish to be too precise. But assume for the moment that real returns are in the middle of Standard’s range and investors put half their money into equities and split the rest between cash, government bonds, corporate bonds and property. A diversified portfolio of this sort would return anywhere between 2.2% in real terms a year in America to 3.3% in Europe. Respectable enough, but not as much as investors are probably counting on.
0 comments:
Publicar un comentario