With seemingly little break so far this year in buoyant capital markets, investors are focusing hard on what may be growing downside risk in many asset classes. Yet commercial property often escapes this scrutiny as it is seen as a safe haven that delivers steady returns. The risk assessment that does take place still tends to use the blunt and broad tools of geographic and sector forecasting.

Using these common measurements alone the margin for error narrows dramatically as valuations peak and yields fall. For example, in Paris’ Central Business District (“CBD”), overall net initial yields – that is, the rental income divided by the capital outlay for the asset – have now fallen below 3 per cent, compared with around 4.5 to 5 per cent six years ago.

More sophisticated risk examination is clearly desirable. We use two guiding principles: that rental income is the principal driver of total returns and not capital appreciation; and that we should better quantify risk and volatility associated with real estate investments, just as we do for equities and bonds.

Towards better risk calculation: rental income is crucial

Real estate is a cyclical asset class, meaning that risk is dynamic and rises as property valuations climb. Yet even in a bull market, the main driver among investors tends to be the asset’s estimated capital appreciation, despite the fact that capital returns are notoriously difficult to predict. Time and again, investors pour money into property at the top of the cycle, only to get burnt when a correction occurs and property’s illiquid nature hits home.

Surprisingly little attention is paid to the rental income component of total return. This is despite the fact that rental income is by far the largest component of total returns in every major market bar Taiwan and Hong Kong. In the Eurozone as a whole, rental income accounts for more than 90 per cent of total returns.

Source: Fidelity International, MSCI Multinational Digest, December 2016. Datasets vary in length depending on availability of data, with starting points ranging from 1985 (UK, Australia) to 2009 (Indonesia).

The lesson here is that investing in commercial property is a bet on future rental income and income growth. It follows that to understand and quantify risk in real estate, investors must pay far more attention to the factors that might affect rental income flows.

It isn’t all location, location, location

And calculating rent profiles and risk is not uniform by area. Of course location is important, but no two buildings are alike and far closer examination of individual opportunities is essential.

Even prime commercial locations in the world’s most desirable capitals exhibit a startling range of risk and return between individual properties. A survey of 74 office buildings in the Stockholm CBD showed an astonishing 4,300 basis points spread in total returns, which computes the rental income and capital appreciation of a property. Central London is no different: the office sector is a highly opportunistic, volatile market that often underperforms the broader real estate sector.

And the performance of the Frankfurt office market in 2011 showed a spread in asset returns of 3,300 basis points – a range more commonly associated with junk bonds rather than bricks and mortar.

Source: Fidelity International, IPD Germany, 2012.

Towards better risk calculation: Focus on tenants and lease structures

When total returns are driven by rental income, it is really important to put tenants under the microscope. What factors might affect their earnings? Is the industry they operate in facing headwinds? Supermarkets, for example, used to be considered bulletproof real estate investments until discount and online retailers came along.

In 2012, Fidelity was looking to invest in logistics warehouses. We saw that assets in the logistics space in Europe were trading at attractive yields of about 7 per cent.

But not all tenants offer the same peace of mind to landlords.

After surveying a number of logistics companies, we decided we were particularly interested in assets where DHL was the tenant. The German government indirectly owns a 20 per cent stake in DHL, implying a very low risk of default. Furthermore DHL’s bonds were yielding 3 per cent. So you could earn 4 percentage points more by buying a warehouse and renting it to DHL than by buying the company’s bonds. Of course we were taking on some real estate risk, but the low risk associated with the tenant and the much higher income return were enough to compensate.

With leases, it is important to dynamically quantify the future as well as current risk of a tenant defaulting on its rent. In the absence of a credit rating and quantified default risk, this has not been possible for real estate investors to determine. However, this can now be measured.

Fidelity’s FIRM Index (Fidelity Income Risk Monitor) calculates future failure risk from a quarterly sample of 5.5 million businesses across Europe. The index tracks the probability of a tenant failing within a 12-month period, and the results can be classified by tenant, tenant industry and property sector. Finally, real estate investors can analyse risk adjusted cash flows from a lease in the same way a fixed income investor might do with a bond.

The lease structure of a property can also have a big impact on total returns but is often overlooked, and rarely priced properly. Different kinds of leases will affect the cash flow from a property over time. Is the landlord able to review rents periodically? Does the landlord have the ability to switch tenants? Can the tenant break the lease to move elsewhere? And when a lease expires, how quickly can a property be re-let?

A lease with indexation built into it, allowing for rents to rise with inflation, can be a very attractive proposition for a pension fund looking for long-term returns. But in a low-inflation environment, landlords might need extra protection, perhaps in the form of periodic rent reviews.

It is our view that investors can achieve more genuine diversification via a combination of different lease structures than by investing in different locations, where the risks in two different investments may turn out to be the same once all factors are considered.

Conclusion

For institutions seeking good risk-adjusted returns, the message in this short paper should be clear: income returns in property matter more than capital returns and are significantly more reliable. However, an over-reliance on the covenant strength of a tenant today or the security of a long lease can be detrimental. The pace of change facing businesses has never been faster. To illustrate the point, less than half of the companies in the FTSE 100 in 1999 remain today.

Investors need to demand a more rigorous approach to measuring risk than the blunt scenario-planning that has served for much of the real estate industry. Real estate investing should be no different from any other asset class. By identifying and measuring the risks associated with rental income streams, investors can gain a much better idea of expected volatility, and therefore a far better assessment of the underlying value of each asset and its place in the portfolio as a whole.

How we quantify risk

Fund managers are like fighter pilots. They need to train and ready themselves for a full range of future threats. But in real estate, the flight simulator has often been a crude instrument. The worst-, medium- and best-case scenarios that are used as rule of thumb within the industry are like a joystick that can only go up or down or stay neutral. There is little preparation for the full range of outcomes over the life cycle of real estate investments.

Over the past 10 years, we have refined a dynamic approach to real estate risk management by working at the level of individual properties and quantifying the risk associated with their income flows. These flows are the biggest component of total returns and more predictable than capital appreciation. Put another way, we have applied fixed income technology to real estate in determining risk adjusted cash flows.

We assess a lot of factors to compute the risk-adjusted cash flow of individual properties: tenant failure risk (from a database of 5.5 million businesses), industry risk, rental prospects, lease structure and the risk of interruptions in income flows, in addition to the more traditional risk factors such as location, the physical state of the asset, liquidity and its supporting infrastructure. This last point should not be overlooked: part of the reason investment banks in London moved wholesale to Docklands was that it was difficult to upgrade the dealing rooms of Central London offices with the right cabling and technology.

Our modelling is dynamic. It looks to produce some 4,000 probable outcomes based on the full spectrum of risks that might affect income returns over the life cycle of an asset. The model then predicts which outcomes are most likely for any given asset, providing a more robust analysis of how individual asset risk impacts the portfolio.