measuring risk and uncertainty

Risk management needs to embrace uncertainty and structural change

The problem with risk is that it is not quite what it used to be. And despite decades of risk management a holistic view that does not simply tick boxes and looks at risk throughout the property lifecycle in all its aspects is missing. One of the ways some of the risks have been managed is by measuring and controlling certain risks using data and models that try to predict future outcomes.

In an environment of structural change historic data series become less relevant. For decades measures of past variance – standard deviation, tracking error, downside semi-variance and many more – have been used as a proxy for risk. They were attractive to analysts as they could be easily measured. Elegant theories were constructed on the founding assumptions of efficient markets and the reliability of the past as a guide to the future.

But there was a problem. The application of so-called Modern Portfolio Theory did not always work, particularly in private equity markets and especially real estate markets. The property sector’s response was to focus on portfolio characteristics that, in themselves, did not rely on past data. Factors like a portfolio’s WAULT, its concentration ratios and use of leverage were compared to benchmark equivalents to establish relative measures of risk in multiple dimensions. However, this did not help with markets.

Approaches such as econometric forecasting and fair value analysis still relied on backward-looking data and a trusting belief in the power of mean reversion. The global financial crisis torpedoed all that. Black swans became one of the most frequently used terms in 2008 and it became obvious that the past is, at best, an unreliable guide.

In the decade following the GFC our real estate markets have faced a succession of disjunctive factors that undermine the relevance of past data:

  • the near break-up of the Euro zone, now quiescent,
  • Brexit, the run up and the aftermath, still ongoing,
  • recognition of the threat of climate change with property as victim and cause,
  • epidemics, ash clouds and other threats affecting supply and demand, and
  • the exponential rise of the internet and reliance on technology.

They are factors whose impact will take years to work through and whose consequences are not quantifiable. We have moved from the known knowns of probabilities to Donald Rumsfeld’s world of known unknowns. Factors which we believe will have some effect on asset values but which defy conventional risk management. We have moved from and age of risk to an age of uncertainty.

Uncertainty, if left unattended, is economically corrosive. It causes decision makers to hesitate, procrastinate and to defer. The challenge to risk managers is how to react when shorn of their familiar tools of analysis based on known knowns, a.k.a. history.

There is no simple answer to the challenge, but there are several approaches that could be adapted to cope with market uncertainty. These are summarised in the remainder of this note.

In equity and bond markets the classic response to uncertainty is to hedge. However, this option is not available in private real estate. It takes time to change a portfolio and there is a 7% round trip involved. What can be done? If circumstances change, is unquoted real estate a sitting duck?

Uncertainty implies the potential for rapid changes in circumstances. This means that assets and businesses that can easily rotate or adapt to the new circumstances will do relatively well.

Thus, when managing uncertainty one should first of all take a razor sharp, holistic look at the existing real estate portfolio. Ask how flexible are the physical, legal and financial designs of each asset. For example, consider the following:

  • Prefer properties with low coverage ratios; they are easier to adapt and repurpose.
  • Think land rather than buildings; let occupiers do their own thing and carry the risk such as depreciation or flooding.
  • Only hold assets which are capable of changing to a more valuable alternative.
  • Forget land use as a basis for sectors. The Government is tearing up the rule book on Use Classes. Location is more important and always has been. Maybe degrees of illiquidity would be a better basis for sectoral definition.
  • Consider how the building is constructed. Structures comprising traditional bricks and mortar are more adaptable than high rise pre-stressed concrete.
  • Lease structure risk can trump location risk: vacant properties have more in common with the risk to income and each than the asset type they belong to.

Regarding market expectations, the use of stress testing avoids the need to identify a specific (unknown) factor. For example, a significant rise in bond yields can be assumed without the need to say why; then assets can be revalued to estimate its impact without resorting to intensive use of past data. Alternatively, if a factor is known but its impact is uncertain, then scenarios can be built up without associated probabilities. Also, alternative courses of action could be evaluated using processes drawn from game theory such as mini max regret which selects the option with the least downside.

These ideas do not yet add up to an elegant theory, but they offer practical ways of coping, and perhaps a source of out-performance in uncertain times. Property as a sector improves by having discussions that push the boundaries beyond the way it has always been done. We need to hear your view on how we can improve risk management, and incorporate the uncertainties that cannot be quantified.