So what's the story as regards future returns?
It's an important issue because if returns are high maybe you don't have to save so much. If returns will be, say, 14% pa it might even make more sense to leave the mortgage where it is and concentrate your free cashflow on saving. Goodness me, at 14% pa it might even be worth quitting your job to go "day trading". Additionally if returns are going to be 14% pa who cares if you pay 2% in fees. Conversely if expected returns are 6.5% pa or lower perhaps fees are critical, we should be saving more and paying off that mortgage before you start any unsubsidized saving is a no-brainer.
That mix of cash, local and global bonds, given yields of 3%, 4% and 6% (hedged) respectively will return 4.3% pre-fees, pre-tax.
Mr Paul's 6.5% assumption was made in the context of people saving via KiwiSaver.
If we are generous and just look at the highest return KiwiSaver group i.e. growth orientated KiwiSaver funds - we see they have an average asset allocation of 7% cash, 14% bonds, 7% property and 72% shares. We can derive what Mr Paul is assuming shares will return because cash and bond yields can be observed today.
That mix of cash, local and global bonds, given yields of 3%, 4% and 6% (hedged) respectively will return 4.3% pre-fees, pre-tax. To get to 6.5% from a 20/80 split means Mr Paul is assuming more than 9% from shares, adjusting for fees and tax, which is not too dissimilar to what global equities have done in the long term. The Global Investment Returns Yearbook shows that the world stockmarket has returned 8.2% pa in US$ since 1900.
Mr Paul appears to have made the mistake of assuming that the past is prologue. It isn't. In fact, perversely the better the stockmarket has done in the past the worse it will do in the future. The certainty of lower long term returns from financial markets going forward has been rehearsed by academics and experts ad nauseam for the last few years. There is a huge body of research which shows that historic returns from financial markets can't be repeated unless markets fall greatly first. Luminaries such as Robert Arnott, Peter Bernstein, John Bogle, Clifford Asness, Warren Buffett, Dimson Marsh and Staunton and the authors of the Barclay's Equity Gilt Study have made that point.
So why should we expect returns from shares be lower in the future?
It's not rocket science as it can be derived from a simple formula. Note we aren't forecasting next year's return or even 10 year returns because that is virtually impossible. A 25 year old starting a savings plan is however interested in ultra long term returns and forecasting those are made much easier thanks to the Gordon Growth Model (GGM).
This model shows that the future long term return on shares is equivalent to the dividend yield that you buy the shares at plus the growth in dividends that will occur in the future. Numerous studies have shown that dividend growth in the long run is pretty constant at inflation plus 1 or GDP growth minus 1. But the dividend yield on the world stock market today at around 2.5% is much lower than it was in the past. For example Robert Arnott and Peter Bernstein, writing in the Financial Analysts Journal, state that the dividend yield from US stocks back in 1926 was 5.1%.
Chart:
Today the dividend yield is about 2.5% so that is one unimpeachable reason why returns on shares will be lower in the future. The other big reason is that in the last 100 years shares became more expensive and the simplest way of understanding that is to see that the dividend yield has dropped from 5.1% to 2.5%. In a landmark paper Arnott and Bernstein reckon that shares becoming more expensive has added 1.8% pa to the historic growth of shares in the past. This is highly unlikely to be repeated so now we have two reasons why future returns will be about 4% pa less than those of the past. There are others but these are the main ones. So throwing those numbers in the GGM i.e. r = d + g, where d is 2.5% and g is 3% or 4% tells us that the long term return from global equities will be 5.5 - 6.5% pa. Let's assume 6%.
Now let's look at future bond returns. Arnott says "as with stocks we prefer to take current yields as a fair estimate of future bond returns". All he is saying is that with the 10 year US bond yielding 2.3% we can be pretty confident that "buy and hold" investors will receive a return of about 2.3% pa. For 30 year government bonds the current yield is 3.05% pa and that in itself is a good forecast of what returns from bonds will be in the long term. But what were 10 year bond yields in 1926? Arnott advises they were 3.7% so historic returns have again benefited from a higher starting yield, a much higher average yield over the period and recently the benefit of bonds becoming more expensive.
In summary Arnott and Bernstein say:
Stock returns in the past have been extraordinary, largely as a result of important non-recurring developments. It is dangerous to shape future expectations based on these lofty historic returns.
Back to Mr Paul's comments - I looked at the asset allocation and fee structure of the average growth orientated KiwiSaver fund. The asset allocation is detailed in the table below together with realistic forecast return assumptions as detailed above. If we apply the asset allocation weightings to those return assumptions and deduct actual NZ tax for a PIE and an average fee of 1.6% p.a. the weighted average, post tax, post fee return is around 3.5% p.a. That is very different to Mr Paul's 6.5% return. To illustrate how different it is $1,000 invested at 6.5% for 30 years produces a capital sum of $560,000.
Reduce the assumed return to 3.5% and you get $180,000. Mr Paul however could yet be proved correct because if stock markets fall dramatically soon and bond yields rise or the NZ$ depreciates sharply the average return over a 40 year horizon will rise.
That scenario however only benefits people who are saving consistently over the period. For someone retiring or buying an annuity today they are stuck with the current forecast share market returns.
Brent Sheather is an Authorised Financial Adviser. A disclosure statement is available upon request. Brent Sheather may have an interest in the companies discussed.
