For full box (Hebrew), press here
Since 2008, prices in the housing market have increased rapidly, accompanied by a rapid increase in the volume of mortgages. The price increases, low interest rates on mortgages, and low returns in other investment channels put the question of whether investment in a dwelling is preferable to investment in the capital market on the public agenda. The box discusses the different characteristics of investments in housing and in the capital market, and compares the gross and net average returns generated by investments in housing, equities, and government bonds in the past three decades (1988–2017).[1]
The main findings:
· In the past decade, investments in housing generated higher returns than investments in the capital market. However, looking at the performances of investments in the various channels from a long-term perspective (1988–2017), we find that investments in equities generate higher average returns than parallel investments in housing, and significantly higher than investments in government bonds. This result remains in place even after taking into account the costs involved in investing in the different channels, including taxation.
· It is important to emphasize that investment in a dwelling differs in many characteristics from investments in the capital market, including the extent of liquidity, diversification ability, level of leverage, transaction and maintenance costs, taxation terms, the correlation with the returns on other assets, and more. These characteristics also have an effect on the investment decision.
Gross (including investment costs) and net (excluding investment costs) average annual return from a 10-year investment
|
Investment channel |
Gross real return |
Net real return |
Difference in percentage points |
|
Tel Aviv 125 |
7.59% |
6.06% |
1.53 |
|
General government bonds index |
3.48% |
2.66% |
0.83 |
|
Dwelling |
5.76% |
4.87% |
0.88 |
We see that even when weighting the effects of the investment costs, the ranking of the channels by level of return remains the same. If we examine the difference between gross and net returns in each channel, it turns out that the investment costs eroded the investment in equities much more, while investment in government bonds and housing were eroded less.
The net return from an investment in housing shows the greatest sensitivity to the duration of the investment. The shorter the investment, the greater the impact of the transaction costs and purchase tax, and the lower the net return. With longer investments, the impact of the transaction costs declines, while the tax benefit on rental income becomes more significant. The net return on investments in the capital market is affected less by the investment duration. Transaction costs are lower and current income is fully taxed. As such, investments for short durations create a larger gap between net returns from investments in equities and net returns from investments in housing.
In summation, between 1988 and 2017—the longest time frame examined—investments in equities for various durations generated average (gross and net) returns that were higher than parallel investments in housing, and significantly higher than investments in government bonds. However, an examination using different time frames may lead to different results. For instance, if we examine the returns from investments that began in early 2007, just before the global financial crisis, we will find that investment in housing carried the highest return among the alternatives.
It is important to emphasize that the various investment tracks do not only differ from each other in their returns, but also in other material characteristics, including tradability and liquidity, financing and leverage, diversification, taxation, and psychological aspects.
[1] The results of the comparison are sensitive to the selection of the start and end dates of the investments. In order to overcome this and estimate the typical returns in each channel, we calculated the average returns generated by investments for a given time period for many dates throughout the period. The returns were calculated assuming that the investments in all channels were not leverages and that interest, dividend and rental proceeds were reinvested.