“It was the best of times, it was the worst of times, it was the age of wisdom, it was the age of foolishness, it was the epoch of belief, it was the epoch of incredulity, it was the season of Light, it was the season of Darkness, it was the spring of hope, it was the winter of despair, we had everything before us, we had nothing before us, we were all going direct to Heaven, we were all going direct the other way – in short, the period was so far like the present period, that some of its noisiest authorities insisted on its being received, for good or for evil, in the superlative degree of comparison only.” – Charles Dickens, A Tale of Two Cities
At many a crisis or juncture the passage above tends to be re-discovered. I only hope that I am ahead of the curve!
In previous articles we have alluded to the forecasting capabilities of repeated patterns in housing markets, and while the official disclaimer on all investments is that the past is not a guide to the future, however it is often all we have to go on!
Continuing from previous work we took a look at the state of the UK housing market in September 2008, just before the financial crisis got into full swing. We now take a look at what happened to property prices over the following two years, where they first fell, and then rose again.
Two distinct differences emerged in how prices changed during and after the financial crisis. In some parts of the UK it was the most expensive locations which fell the most, and in others it was the already cheap locations which fell the most. In this case we consider ‘cheap’ and ‘expensive’ as the ratio between earnings and house prices. This does not take into account credit availability or interest rates.
For example in the chart above we would consider 2018 to be ‘expensive’ as the average price to earnings ratio was ~7.5x, compared with 2002 which was relatively cheap with the ratio being ~4.8x.
We found that in London and Wales the relatively ‘cheap’ areas which had the worst declines in price, and the more expensive areas were relatively insulated with the effect being strongest in London. In contrast, the North West, South West and West Midlands, showed that more ‘expensive’ areas had the worst declines. Yorkshire and the East midlands also showed this same relationship but it was not as strong. The North West, South East and East Anglia showed now specific relationship.
The 33 Boroughs of London in September 2008 had an average price to earnings ratio of 9.23. Over the following 12 months, prices fell by an average of -6.5%, with some declines as large as -16.2% and others showing growth up to +1.9%.
What may come as a surprise is which of the London boroughs were most impacted, and how that relates to their relative expense. Plotting the price earnings ratio in 2008 against the percentage change in price from 2008-2009 reveals an interesting relationship.
The locations which were the most affected by price falls were not the locations which were already relatively expensive. It was the ‘cheaper’ locations which tended to suffer the greatest declines in price.
The fit of the model can be visualised in a different way, by placing the predicted decline in prices and the actual decline in prices alongside each other. There are some noticeable outliers. On the ‘positive’ side; Hammersmith & Fulham, Southwark, and Barnet all performed better than might have been expected. Conversely Kingston, Camden and Richmond all performed worse than expected. Keep this in mind for the later section on forecasting.
The more expensive locations generally weathered the storm more easily than other locations. This may be due to several factors.
Proportion of equity value held: The proportion of a property which is owned outright, commonly known as equity indicates a relative degree of stability to prices. In a downturn someone who owns their property outright is not going to be under any pressure to sell, even if they lose their job or other major source of income. Conversely someone who is highly leveraged and has lost their job, in a falling market, may look to sell as quickly as possible. Contributing further to the decline. The ‘Area Leverage’; a measure of the total outstanding mortgage debt in an area relative to the value of property in that area, is lowest in the most expensive locations.
The chart below shows the % change in prices between 2008-2009 for each of the London Boroughs.
Just as useful as knowing which locations declined in price the most, is understanding which locations increased in price in the following year. Here the price to earnings ratio is less instructive in indicating which locations would increase in price following the downturn.
Mapped on top of each other:
So what might happen this time around? Well there are significantly different circumstances, however if in a very simple model we applied the line of best fit equation from 2008-2009 to the 2019 prices and price earnings ratios we can see what might happen across London.
A similar pattern plays out with the more ‘expensive’ areas of London being less affected than the ‘cheaper’ areas of the city. It is important to note that since 2008/09, the price to earnings ratio of properties in London has increased by 46%, and so to adjust for this we reduce the 2019 price earnings ratios by the same amount, effectively rebasing the expense ratio to its former level.
Using the same ratio as the previous downturn shows that again the generally more expensive areas are less impacted, however very clearly, all areas are impacted and almost all are negatively impacted.
There are some differences between 2020 and 2008. While again it is important to point out that this is a highly simplified model, if correct it would imply that all areas will be negatively affected, but that some will be more negatively impacted than others. The chart below simply shows the difference between the 2008 predicted changes and the 2020 predicted values.
It is immediately apparent that the expense effect in Wales is far more limited than in London, however it does still account for 11% of the change in price. We we won’t analyse these as deeply as with London as the relationship is not as strong. However we will present some of the same set of charts and let you come to your own conclusions.
What was perhaps more interesting about Wales was what happened in 2010, when prices began to recover. The areas which were relatively more expensive in 2008, had actually gained the most by 2010!
It seems that people may have used the decline as an opportunity to ‘upgrade’ their housing and move to areas which pre crisis had been out of their reach.
North East, South West, West Midlands:
Here we found the opposite relationship to what was found in London and Wales. In these parts of the country it was the cheaper locations which were more ‘insulated’ than there expensive counterparts. For every 1 point increase in the relative expense of properties in an area there was a corresponding 1% decrease in price during the subsequent crash.
The majority of locations were within a reasonable range of the predicted outcome based on their relative expense. However some locations such as Middlesborough and Twekesbury (extreme right of the chart, name hidden) surprised with higher and lower changes than expected.
In 2008 Middlesbrough had relatively high levels of unemployment, and it only got worse. On top of that the football team was relegated from the Premiership that same season. So why did prices actually go up from 2008 to 2009? Well that just isn’t clear, but Middlesbrough completely bucked the trend.
Similarly confusing at first is Tewkesbury, where prices fell by 15% from 2008 to 2009. But unemployment actually decreased in the same period. However, looking back at the news at the time, severe floods hit the region. Images of power substations and villages underwater can still be found online showing the extent of the damage. So this actually had less to do with the financial crisis and more to do with a local crisis.
If the same ratios hold constant for North East, South West and West Midlands then the change in values in 2020 could be as shown below:
Yorkshire and East Midlands:
Much like Wales and London, there was a similar relationship in Yorkshire and East Midlands to the relationship in the North East, South West and West Midlands, however that relationship was not particularly strong. Areas of Yorkshire and the East Midlands with relatively expensive properties tended to see greater price declines than areas with cheaper properties.
Again much like wales there was a stronger relationship between the price earnings ratio and the subsequent recovery, however that too was not as strong.
How to make use of this?
A flight to quality: In both London and Wales there was a relative ‘flight to quality’ as prices of more expensive areas declined less, and in the case of Wales, rebounded more after the crisis.
Cheaper can be safer: In the majority of the country the cheaper the properties in an area relative to the income of the people living there, the lower the fall in price. However, in areas where the LTV and unemployment levels are already relatively high, cheaper can also be quite risky.
Rebounds: When rebounds do occur, in every case we found it was experienced the most in the areas with relatively expensive properties. This may be due to buyers who previously would have bought a lower priced property being able to bid-up the higher priced ones which have suffered a decline. It may also be due to investors and ‘normal’ buyers re-assessing the location on the basis of its former characteristics and so returning its previous prices relatively quickly.
REalyse (Treex Ltd) does not provide any form of investment advice or property advice or any other regulated function. Note that any information or opinions, presented or referred to in this article are for information purposes only. Any actions taken by a reader are done entirely at their own discretion, you are responsible for your own investment decisions and hold Treex Ltd harmless from the results of any such decisions. Whilst every effort has been made to ensure the accuracy of the information herein some inaccuracies may remain.
© Treex 2020