Crunching the numbers: Is Andy Haldane right?

Danby Bloch white

Houses are better investments than pensions according to Bank of England chief economist Andy Haldane.

But is he right? And how can one decide?

There are several possible approaches. One is to look at “pure” past performance, subject to all the usual provisos.

Another approach is to consider the financial and tax engineering of house ownership and pensions on the assumption that the basic underlying investment performs the same. And that is what we have done – with many thanks to actuary and financial guru John Housden. You might be surprised by the answers.

Houses and pensions are both complicated in different ways – legally and as investments. Houses are more familiar to most people than pensions; pensions are arguably less complicated than they look, despite Haldane’s provocative protestations.

One possible approach would be to use equities as the underlying pension asset and compare them with average house prices. The following basic investment performance figures come from the Barclays Capital Gilt Equity Study and the house costs are from Nationwide. But there are some limitations to this approach:

  • One tends to build up pensions by regular contributions; houses are usually lump sum investments. A practical approach might be to compare the relative performance, say, over a 20-year period. However the start and end dates can make a big difference. Both equities and houses have been through rocky patches. So it is also worth looking at longer periods as well as different start and end dates
  • Both measures leave transaction and management costs out of account. Pension and fund charges can take quite a bite out of returns; but stamp duty land tax as well as transaction and management fees can impact on property returns.
  • The Barclays Capital equity investment metric includes reinvested dividends, which make a considerable difference to the returns.Based on the last 20 years to the end of 2015, with reinvested dividends, the equity portfolio increased to 3.55 times its 1995 level; without the dividends it did not quite double in value – increasing to only 1.91 times its 1995 level.
  • The Nationwide house price index takes no account of rent (or imputed rent, which is the benefit of living in a property for free). One approach might be to reinvest the rent from property into something else – but what? You cannot easily diversify residential property in small units of value. The Nationwide figures also just reflect prices of Nationwide’s mortgaged properties – not the whole market.
  • Diversification is easy with equities and the Barclays Capital uses the FTSE Actuaries All Share Index . Property is harder to diversify because it is much larger and it is easy to get it wrong with property. Some property has done really badly – the North on average increased to 2.88 times its 1995 level compared with 3.87 times for the UK as a whole and 6.29 times for London. And in some localities, people have lost money in absolute terms – think former mining areas.
  • It might be argued pensions are not wholly equity-based investments. And that is fair comment. We have had a bond bull market over the last 30 years.

Over the last 20 years to the end of 2015, houses have been the winners. Houses (excluding rent) are just ahead at 3.87 times their 1995 values, compared with equities (including reinvested dividends) which increased by 3.55 times, while long term bonds have increased by 4.10 times their 1995 values. And if you put equities on a level playing field by forgetting (or spending) all the dividends, they would fall far behind houses at 1.91 times their 1995 values.

But if you go back a further 20 years, the positions are dramatically reversed in terms of the capital value. Back in Q4 1975, the average house price was £11,288 and by Q4 2015 it was £197,044 – just over 17 times the end 1975 value. But equities did even better. In the first 20 years, which includes the 1987 crash, shares did amazingly well. So by 2015 the FTSE Actuaries All Share index without reinvested income rose to 21.8 times its 1975 value. So on the basis of omitting income, shares outshone property over the last 40 year period. That is important because it reflects the position of people who spent all the income from their investments.

The position changes yet again if income is included. By 2015 the FTSE Actuaries All Share rose to a very impressive 105.7 times its 1975 value – reflecting the importance of compounded growth from reinvesting into the market. The actual difference in the average annual return is much less impressive – 12.4 per cent against 8 per cent – such is the power of compounding.

One might get a very approximate idea of the value of the rent by assuming a net annual yield of say 3 per cent after expenses and basic rate tax. Without rental income, the annual return for residential property over 40 years was 7.41 percent a year; if we add 3 percent net annual income to that figure, we get a compound rate of 10.41 percent a year and cumulative figure after 40 years of 52.5 times the original investment – roughly half the return from equities.

So even on past performance, it looks as if equities have generally outperformed equities over the last 40 years. But the position is reversed and property has been the clear winner over the most recent 20 year period to the end of 2015.

Of course this all ignores such major issues as liquidity, risk, tax and especially the impact of gearing or leverage on borrowing. And one of the clearest conclusions you can draw from the data is equities have been much more volatile than residential property.

Haldane has a case with the most recent data for his very broad generalisation, at least with 2015 as his end date. But as we all know past performance is not a reliable guide to the future.

Danby Bloch is chairman of Helm Godfrey and consultant at Platforum