Exploring the potential positive impact of predictive rent models for improving investment decisions in the Nigerian and South African property markets

Since forecasting rent behaviour is critical to investment decision-making, local and foreign real estate investors want to have simple measurable indicators to track and understand market behaviour. Accordingly, modelling rent behaviour and the future trajectory of real estate price movement is one of the necessary steps to improving the African property value chain. Predictive rent models aim to improve the property market information available to decision-makers, thus helping to stimulate market growth. A recent study at the University of Pretoria evaluates the potential of such models to inform investment decisions across African property markets.

Undertaken at the University of Pretoria, South Africa, the study used a logit regression to model the relationship between macroeconomic data and listed commercial property data in South Africa and Nigeria. The study identified leading economic indicators; evaluating the performance of predictive models for early detection of turning points in the listed property market. The data confirmed that listed property data responds significantly to changes in some macroeconomic data points. In addressing the data gap in African property markets, signals from the macroeconomy provide enough predictive capacity for measuring the future trajectory of the commercial real estate sector. The study therefore suggests that macroeconomic data can also be modelled successfully to inform better decision-making for investors in the African housing market.

This study contributes to the growing body of knowledge on real estate rental forecasting from the African perspective. The data used was obtained from the Iress Expert Database, Stat SA, the Central Bank of Nigeria database (CBN), the National Bureau of Statistics and World Bank Data.[1]

The logit model is recommended for its capacity to capture future rental directions based on the general economic movements and likely turning points in an economy. The model is particularly useful for property analysts because of its adaptability for macroeconomic and indirect/listed real estate relationships. Additionally, the reliable economic indicators identified in this study for the Nigerian market included lending rate, Treasury bill rate, and consumer price index/inflation. For the South African market, the reliable economic indicators identified were coincident indicators and exchange rates. In Nigeria, the logit model’s accuracy in predicting the listed property market performs 22.2 percent better than a simple test of averages. South Africa’s data, while more available than Nigeria, only improves the information about the market by 9.4 percent. However, the accuracy of the South African logit model is higher than that of the Nigerian logit model by 5 percent. This suggests that while the Nigerian listed real estate market predictions might be more sensitive to macroeconomic indicators, the South African market is more accurately predicted by the logit model.

The strength and similarity of the model’s capacity in both countries showed that each market signal is correctly predicted by turning points in the economy as much as 75 percent (Nigeria) and 80 percent (South Africa) of the time. The logit model is a simple regression model that correctly captures the relationship between scarce property data and the more available macroeconomic data. This improves on how much inference about emerging African markets can be made when only a few players are in the listed commercial market.

The study was done as accurately as possible given the available data. This led to some recommendations including the following:

1. African property markets must encourage data gathering and the listing of real estate operations. A growth or influx of commercial property investment and finance will stimulate data just as much as data stimulates investment. Instruments that allow small or medium scale property companies to raise capital through the stock market influence how much information about that property market can be gathered, and analysed.
2. Further study is required to better understand predictive probabilistic models. There is a need to evaluate the accuracy these models add to real estate market analysis and reporting. These kinds of analysis rely on data that is significantly reliable for predicting markets. Macroeconomic data and models are empirical sources that improve on expert opinion.
3. With the limited numbers of commercial property companies listed in a lot of African stock markets, investment relies largely on expert opinions and surveys. These expert opinions or surveys might be limited in terms of evaluating robust data to capture the true behaviour of the market over a long timeframe. Therefore, further research should be conducted on how econometric models fit into business reporting for residential and commercial real estate companies. This would make it possible to evaluate data on real estate performance, based on consistent data sources such as listed real estate data.
4. Open data policies should be adopted in more African countries to improve access to data. A bigger array of data sources would provide a more valid basis for quantitative evaluations. The data timeline helps to identify trends, patterns and turning points which can significantly affect investment decisions.
5. Directional forecasting should be explored in-depth, as it is more likely to serve towards advising investors than point forecast. When the market is evaluated for its qualitative value over time instead of by the actual amount that is paid for rent today, there is a bigger chance of preparing for economic shocks that might affect the property market on a macro scale.
6. Property indexes should be designed based on their capacity to explain the relationship between the economy and real estate data. It’s not enough to understand how the physical, geographical or political landscape affects a property’s value. The impact of the economy on property value is important as it indicates the overall wellbeing of a market. Macroeconomic indicators and models provide additional data sources for a more robust assessment of the property market vis-à-vis rent behaviour and investment.

There are limitations to this approach to analysing the property market in Africa, including access to data such as the ABSA Housing Index. Similar databanks such as those found on the MoneyWeb site and other platforms are only accessible at a fee. This research, however, has succeeded in adapting the logistic regression as a macroeconomic modelling approach for real estate forecasting in South Africa and Nigeria. The study ultimately shows the possibility of achieving greater real estate forecast accuracy, using macroeconomic data. There is a lot of room for further study into this statistical modelling approach and its applicability to real-time data and investment decision-making. However, the study serves to demonstrate the huge possibilities inherent in the use of macroeconomic data and modelling for understanding emerging property markets across Africa.

To read the full study, click here.

[1] The South African economic data comprised time series for fifteen years between Quarter 1 (Q1), 2003 and Quarter 4 (Q4), 2018. The Nigerian data comprised time series for ten years between Quarter 1 (Q1), 2008 and Quarter 4 (Q4), 2018.