Predicting crisis - Markets are connected in (un)expected ways
Updated: Mar 3, 2021
The main question market participants are trying to answer these days is how big the COVID-19 impact will be...on local markets and internationally. As no predictive risk tools are in use, there is lots of guessing and assuming but no objective and data driven analytics of overall situation is being provided.
We did one and here are some observations.
Vienna Prime Office in 2020 - stable/ not stable?
Real estate and the year 2020 - a quite interesting combination. So far, there was bit of uncertainty what will happen in the real estate sector, but some of the market participants already see the big recovery in the worldwide commercial real estate sector. Basis for this optimism is that e.g. worldwide property sales grew by 23 % by end of Q3 2020 compared to Q2 2020. Nevertheless sales in 1-9/2020 are down by 48 % compared to the first nine months of 2019.
But let's not thrive so far. Transactions in Austria's office market (which is heavily dominated by Vienna) are down by around 35 %. And exactly this result was taken by some of the market observers to see this as a strong proof for a stable market in those demanding circumstances. After all, in Vienna's prime office segment the rent prices and vacancy rates are still relatively stable. Additionally, investment yield in this prime segment, after some hiccup in the middle of the year, does not show a significant signal of weakening.
On the other hand Erste Bank, as one of the rare examples, reported worries after the Q3 2020 market results that crisis could also hit the real estate sector during the course of 2021. The banking scene and especially the European Central bank are concerned that in 2021 a wave of insolvencies could affect the overall economy. Insolvencies will cause a lot of companies leaving their markets which will have impact on vacancy levels and this could have impact on real estate markets. Prime office will be no exception to this.
Not to forget, as we learned with the financial crisis in 2007-2009 any impact on the real estate market comes in a protracted way. Last time, the crisis affected rent prices and investment yields significantly, though the effect was to be seen 1 to 1.5 years later.
The Silo Approach
One grave misconception in this business is to view local real estate markets in a rather isolated way. Sure, we get regularly informed about how much of the investment volume is driven by foreign investors here in Austria. Recently I read that the ratio of German investors declined from 27 % to 15 % of the overall investment volume. Nice to have, but what does this mean for this local market?
In order to better understand the situation, let's move back in time a bit and see how the prime office market in Vienna has developed so far. In terms of investment volume and investment yield, we face the following situation:
The recent decline in the overall investment volume was not solely due to the Corona crisis. It started already in 2019. Nevertheless, the investment yield kept the free fall making the market more expensive. In Q1 and Q2 of 2020 there was a slight correction in the yield level but the last quarter saw a going back to "normal" again. Yield for prime office in Vienna was again at 3.25 % p.a. by end of September 2020 (I also read that we are at 3.0 % p.a. already).
Conventional market understanding would say that the more investment volume is flowing into a market the higher the chances that yields will decrease. In case the capital flow recedes, price movements will turn and the market might get cheaper. And so did conventional predictive analytics tools when focusing on the isolated market environment. But this was not the case for the Vienna prime office market. In principle yields kept falling irrespective of the capital flow and irrespective of possible worries of a Corona-impacted economic crisis.
In summary, for the last 18 months market features within the Vienna office market did not give strong enough signals in order to anticipate those declines of the investment yield as they happened as from the second quarter of 2019. And still - they happened.
Breaking Up the Silo
In my point of view, there are two issues to be noted when talking about the Vienna real estate market in general and the prime office market in particular
First of all, this market is quite small. As an example, the overall number of significant transactions in the prime office market for e.g. 2019 was the same as it was just for the central business district in Frankfurt. This small "sample" size allows for outliers to influence any market figure much heavier than it would be the case for larger market environments. In other words, when one transaction sees a price "off" the general market trend, exactly this transaction will influence the average market rate more significantly in a small market environment. The market therefore shows a more volatile and sometimes a more irrational behaviour.
Second of all and much more important is the fact that markets cannot be viewed in an isolated way. Investment capital flows among markets (and asset classes). To some extent markets influence each other and are therefore dependent from each other. The development of e.g. the prime office market in Vienna cannot be significantly different to other markets which are similar in their legal, economic and geographical framework. Otherwise, arbitrage opportunities would open up which would eventually be exploited by investors and - in the mid-run - would level the developments of those markets again. There are no market islands or market silos.
So, why are they treated as such in terms of market reports and business analytics?
Take Frankfurt and Vienna as an example. It is to be noticed that the movement of investment yields in both markets are highly associated. Those two yields reflect different local markets but show a very strong correlation. See the chart:
Allowing for this external signal 'Frankfurt investment yield' to be incorporated in the fabric of market behaviour of the prime office market in Vienna seems therefore reasonable. But pushing the agenda a bit further, and incorporating this market feature in the networks dynamics framework of the Vienna market makes all the difference. The anticipation of market movements got much more accurate, neater and sharper.
Breaking up the isolated view on a local market, incorporating networks dynamics and taking advantage of the powerhouse called causal inference paced the way to a sensible predictive analytics application. Market behaviour, which seemed irrational from a local point of view, got much more sense when interconnecting similar markets. The predictive analytics tool started to recognise those significant signals coming from external markets and much better utilised them by means of prediction accuracy.
Real estate markets are not an isolated event anymore. They are no silos where a certain market behaviour is elaborating without being affected by other, external circumstances. International capital flows make them so much more interconnected and dependent from each other. Saying this, not taking into account those interdependencies and not reflecting them properly when analysing risk positions in a local real estate market evolves in unguided risk exposures in exactly those markets.
On the other hand, reflecting those circumstances in a risk analysis gives a competitive advantage, gives the competitive edge in the market and towards competitors.
A new generation of risk management tools are available just to do this!