Viewpoint: Mind the yield gap on property

By Monika Ward, deputy head of research and strategy at Axa Real Estate

The difference between the yields on government bonds and on property is sometimes used to determine if property looks cheap or expensive on a relative scale.

Prime property in Europe’s core markets is attractive to a particular sort of investor: international, risk-
averse and with high capital allocations. They now dominate many parts of the market and their main targets remain London, then Paris and then the top five German cities.

Given the strong demand from both foreign and domestic investors, it could be assumed that the gap between bond yields and property yields should be closing – but the opposite is happening.

Comparing a composite of prime office yields of the core markets in the UK, France and Germany with the 10-year gross-redemption yield on German government bonds (used as the best representation of a risk- free rate in Europe), prime property in these core markets is at the ‘cheap’ end of the historic range.

An alternative view is that bonds look expensive. A comparison between bond and European equities’ yields reinforces the expensiveness argument for bonds and severely weakens the case for the cheapness of property. So it is important to understand whether current levels of bond yields are sustainable.

Developed economies’ bond yields have been falling for the past 30 years. The explanation for this secular trend includes: regulatory pressures on, and liability-matching strategies by, pension funds and insurance companies; re-pricing of equities, making them relatively more expensive; reduced risk appetite from many investors; lower inflation; and investors’ willingness to accept lower real returns.

Bond pricing’s cyclical aspect exaggerates some of these secular shifts. With recent rises in global debt, risk aversion is high – a return of capital in real terms has become a much greater priority than earning a return on it. As investor fear dominates, short-term demand for safe assets has driven up the prices for this shrunken market.

At some point, risk appetite will return. We believe that for Europe, and the Eurozone in particular, this is several years away, as the ECB’s recent quantitative easing indicates.

But when risk appetite returns, we envisage the gross redemption yield on 10-year bunds (the proxy for a risk-free rate) rising to around 2%, assuming that in addition to the current yield of c1%, the inflation premium rises by 0.5% and the inflation risk premium by 0.5%.

A 2% bond yield is on the high side of our expectations, but we prefer to be conservative, particularly as investors today require a liquidity premium, which will partially disappear over time. (For bonds, we consider the liquidity premium to be negative, which has the effect of depressing the yield.)

A shift in bond yields is not the only factor that changes the yield gap. Property yields currently reflect very limited rental growth expectations; some investors do not consider it at all. That is shown by the lack of significant differentiation between yields in various parts of core markets despite, for instance, prime London West End office rents having grown significantly in recent years and the City’s starting to recover.

When risk-aversion reduces and rental growth returns more widely to core, prime markets, yields will face downward pressure. Property values will rise and the yield gap will close, more so than the measure would suggest due to rising rental values. We should then expect the two to converge – as in 2006-07.

Until then, the yield gap can be considered a form of risk premium. Strictly speaking, a risk premium represents the difference between the returns of risk-free and risky assets (here, government bonds and property), not the difference in yields. But if property income growth is low, the property yield effectively becomes the return on the property – except, if that yield reduces.

In the past, UK real estate’s long-term average risk premium has been  200-300bps. The risk premium changes over a cycle and is currently low: we have assumed around 100bps for prime property, especially considering investors’ risk-aversion.

Assuming a 2-2.5% long-term, fair-value yield for bonds, and adding a 100bps risk premium, we get a 3-3.5% prime property yield. That is about 100bps lower than the current prime office yield of our composite index. But when rental growth returns, the property yield could fall to match bonds at 2%.

So even with conservative assumptions, it is possible to justify the acquisition of prime assets as low as at the 3% level in the core office markets.