top of page
  • Writer's pictureChristian Schitton

Connecting Markets and Portfolios with Predictive Analytics Tools


In the last articles and videos we talked a lot about the behaviour of commercial real estate markets, how to cope with uncertainties and how to predict potential changes in those markets, what to do regarding mid- to longterm market developments and in which way hidden market clusters can counteract to portfolio diversification strategies.

However, for a company to have a real practical and strategical use of that tool and to move away from a pure market report (however advanced it might be) the idea behind this kind of solution is to combine the external market data and their respective analytical results with internal portfolio data. This way we are able to create interactive, responsive and automated system which is able to deliver fast and efficient results, of a very complexed analytics, in a flexible way.

Automisation is something which will make the life easier for decision makers, shareholders and supervisory boards because it will allow them to have data driven overview of markets, portfolios and risks and will enable them to act fast. Beside the decision makers, it will make day to day handling completely different for risk managers who will not have to do almost any manual handling anymore and will be able to focus on other, more important tasks.

But automisation by itself is not enough as it fails to include the deep insights needed for understanding what is happening "below the surfice", uncover breaking of the market patterns and offer the competitive advantage for the company. In this the system has to move a way further from the traditional approaches which can be quite limiting in the current global economy requirements.

First, we have a look at a more traditional risk evaluation approach and then we take a deep insight into advanced predictive analytics methods in order to see the decisive step forward in the aim to building up an effective basis for decision making under uncertainty.

To keep things simple, we focus on one single investment project.

The Investment Asset

As an example, the purchase of an office building in a local market with the following rough project key parameters is under consideration:

Based on this framework, a potential purchase price would stick around MEURO 315. Based on the intended leverage of 70 %, the equity capital to be invested amounts to EURO 94.5 with a requested minimum equity yield of 9 % p.a.

A glimpse into the latest development of this specific real estate market exposes the following picture:

Conventional Evaluation Approach

Given the investment framework and the latest market scenario, the focus is put on the expected equity yield which could be achieved with this investment.

In a more optimistic approach the rent price is expected to further increase to EURO 26.5 per sqm and month and the investment yield to further decrease to 2.75 % p.a.

Hence, the expected equity yield amounts to 12.1 % p.a. for the overall investment term which would be above the minimum requirement. Here is an excerpt of the underlying Cash Flow calculation:

In a more pessimistic scenario, the rent price decreases to EURO 24.5 per sqm and month and the investment yield remains at 3 % p.a. which results in an expected equity yield amounting to a mere 3.2 % p.a. This is not thrilling but it is still a positive result.

Based on latest market developments and given the expectations for this real estate market, the optimistic scenario is given preference and the investment is put on a ‘go’. After all, some key parameters still could be tuned, e.g. increasing the debt leverage, in order to get a more stable equity yield outlook.

Of course, the scenarios shown here are by far not exhaustive. A more deepening sensitivity analysis would be necessary in order to get a clearer picture about the perspectives of this investment project.

Nevertheless, conventional approaches like this reveal the shortcomings in giving a decision maker the right information basis in terms of market development, in terms of investors’ requirements, in terms of risk considerations and especially in terms of investment opportunities.

To name a few of those shortcomings:

  • First of all, there is no reflection of market dynamics in the analysis. Sensitivity analysis definitely does not offer this smooth fabric connecting market behaviour with project/ portfolio data and perspectives.

  • The focus on typical real estate market parameters is too narrow for a comprehensive overview.

  • Cluster risk behaviour is not taken into account though it has an immense impact on risk exposures.

  • All of the input parameters of a project/ portfolio are driven simultaneously by market parameters based on their associations and conditional dependencies. Changing one, two or three parameters here or there in order to examine different scenarios will not do the job.

  • The offsprings of those analysis are not quantified by ‘how sure it is to get such a scenario’. In other words, what is more probable e.g. to have a market with increasing rents and a decreasing investment yield or a market with decreasing rents and a stable investment yield? This is definitely not reflected in the traditional approach.

Hence, tools are needed to take all this into account and to improve the information basis of a decision maker.

Advanced Predictive Analytics Tools

1. In general

Actually, advanced analytics tools, especially advanced predictive analytics tools come in various characteristics and different nuances, of course always depending on the task at hand.

What makes them such a great technique is the fact that those methods are highly adaptable, that they are incorporating very quickly upcoming changes in market behaviours and that they can be customised based on the pending problem or based on a specific situation.

In other words, they provide a very effective tool kit giving a decision maker the best possible support to understand risk exposures, to detect market opportunities before others do, hence to help a decision maker to take a market position according to her risk strategy/ risk appetite in due time.

2. the market level — external data

In case of our planned office building investment, this adaptive approach could mean to first zoom into the local market via network dynamics applications.

And indeed, this approach discloses that this real estate market is not only comparably small but heavily associated to another bigger market hub having a certain influence on the behaviour of our observed market.

Given this, we are facing a regional market cluster. And the consequences of this cluster behaviour are to be implemented in the further analytical procedure which also means that we cannot look at our specific market in an isolated way anymore.

Here is the graph model showing the interconnections of those real estate markets and their respective real estate market parameters as well as other important input parameters:

This new information allows us to provide a much better simulation of potential market developments. In an additional step, the rising uncertainty in the market (e.g. due to the insecurity about post-COVID circumstances) is implemented in the analytical approach by means of statistical modelling.

Below are some of the simulated results of selected market features:

2. The market level — external data internal data

As from here, the really exciting part starts: The results of potential market developments weighted by their probabilities of materialising are waived in the fabric of the risk/ performance metrics of a project, a portfolio or the whole entity.

In other words, external data are combined with internal data!

In our example, we are interested how the investment in the office asset will work out in terms of

  • equity yield and

  • current liquidity

*we can focus on any other risk metric or combinatio of porfolio risk metrics

Here is the result for the equity yield:

The blue area represents a positive equity yield. The red area indicates an equity loss. The black line shows the required minimum equity yield of 9.0 % p.a. while the black dotted line displays the initially calculated equity yield of 12.1 % p.a. as derived in the optimistic scenario in the conventional approach as shown above.

It turns out that this optimistic scenario “overshoots” the possibilities. With this investment project, there is only a mere 9 % chance of achieving this result or even doing better.

And there is a 78 % probability not to even reach the required minimum equity yield.

Chances of making a loss with the equity invested is up to 11 %. Though, the potential to suffer a loss bigger than 5 % is just 2.5 %.

From this point of view, the potential investment seems not so bright anymore. A final ‘go’ will depend on the risk appetite and the final minimum yield requirements of the investor.

Now, let’s see the results for the current liquidity (i.e. operational Cash Flow including financial expenses but excluding any proceeds from the sale of the asset) provided by the intended asset during the investment term:

Actually, the results for the current liquidity match the initial evaluation as done with the optimistic approach as shown above with an expected value of approximately MEURO 6.

Though the range of potential outcomes (see the graph) is much broader and reflects the current insecurity in the market. And with this comes a risk of a liquidity shortfall amounting to 8.5%.

4. An overall evaluation

Overall, the equity risk seems more emphasised than the liquidity risk for this project. Reason for this is partly embedded in the market dynamics which was incorporated in this analysis. Not to forget that there is an average lease term of 3 years which buffers any market development to a certain extent.

Partly, it is to be found in the initial investment parameter of the asset. While the investment yield is at the level of the market yield, the average rent price of the asset is slightly below market and the current vacancy of the project is slightly above market.

In the conventional evaluation, we mused about the possibility to increase the debt leverage in order to get a more stable outlook for the equity yield. Indeed, within this traditional framework an increase of the leverage from 70 % to 80 % would move the equity yield perspective from 12.1 % p.a. to 16 % p.a.

Unfortunately, advanced predictive analytics methods are not so gracious. The simple tuning of investment parameters cannot escape the market rules. Have a look at the graphs below showing the equity yield and the current liquidity situation for the same project but with an 80 % leverage (instead of 70 %):

The chances of achieving the newly calculated equity yield of 16 % p.a. is still minimal with 8 %. The situation for the required minimum equity yield improved — a bit — but the risk of experiencing a loss also increased.

However, the more crucial topic here is the liquidity situation. The risk of experiencing a liquidity gap rose from a manageable 8.5 % to 50 %! The higher financial obligations in the wake of the higher leverage are taking their toll.

As can be seen, the slight spinning on one parameter in order to improve the performance metrics of the asset had a very (dynamic) negative impact on some other risk metrics. By the way, the performance metrics did not really improve from a probabilistic point of view.


Although we just touched the surface here, it should be obvious by now that advanced analytical methods support a decision maker in a much more effective way.

External market data is automatically connected to internal project and/ or portfolio data which makes portfolio management process very lean and manual handling free. People can focus on other important tasks and avoid loosing weeks on sending incomplete list back and forward, only to get very limited analytics.

Provided this, market patterns, cluster behaviour and a changing market conduct is smoothly incorporated in the risk/ performance metrics of the respective management information system.

The grid of the risk/ performance metrics can be adjusted in a very flexible way depending on the problem at hand. Of course, in order to being able to initialise the best solution for a given issue a lot of deep business expertise is necessary. This makes sure that the most effective solution is developed for any given problem.

What makes advanced analytical tools so effective is the fact that results are coming in quantified (in terms of probability), objectified and transparent way and can then be compared with the risk and investment policies as well as the pre-agreed risk appetite of an entity, as general guidelines.

This shortens the way of decision making decisively, makes it less risky for the decision maker and supports executives in the most flexible way, to reveal hidden risks and to detect market opportunities before the competition does.

In short, it supports a decision maker to take a suitable market position in due time, while keeping the costs of implementing a management information limited.


bottom of page