Welcome to the first of many! This week's topic is the Swedish property market.
A brief introduction
What you have here is the first of many contributions from Roger Josefsson, Chief Economist at Macrobond. Each week Roger will write an analysis on a topic of his choice and present it with Macrobond from start to finish. You'll get a sense of what data we have, what analytics you can perform, and how easy it is to showcase and share work this way.
For those of you who use Macrobond, you can click on any of the charts to open it in the app and see exactly what data was used, what analysis was applied, and chart styles and formatting. Not to mention Roger's amazing analysis of current market movements around the world. Speaking of Roger's perspective...
The world's strongest economy has a fatal flaw
Swedish real estate market
A few years ago, in the midst of the financial crisis, OECD Secretary-General Angel Gurria called Sweden a "Pippie Longstocking economy", referring to Astrid Lindgren's favorite story of "Pippa Longstocking, the strongest girl in the world".
But if Gurria had delved deeper into the Swedish economy, he would have noticed what many domestic and foreign investors have long been obsessed with; Swedish real estate market.
Bild 1: Rising real estate prices
Note:HOX house price indices are linked to SCB and BIS data. Longer time series can be constructed using the Riksbank's historical property price indices.
Economists like to look at real estate prices in real terms, while controlling price developments in general. This makes the hockey stick shape of property prices in Sweden even more attractive.
Figure 2: A good investment if there ever was one
Note:HOX house price indices are merged with SCB and BIS data and deflated with Statistics Sweden CPI and historical shadow CPI. Even longer time series can be constructed using the Riksbank's historical house price and inflation series.
Lower nominal interest rates
Lower nominal interest rates are of course one of the main causes; Mortgages have gotten cheaper (yes, the illusion of money) since the Swedish financial and banking crisis in the early 1990s. As depreciation periods gradually lengthenedto infinity1. It allowed households to climb the real estate ladder by increasing their debt while keeping their monthly mortgage costs intact (or even reducing them). As a result, the household debt ratio has been rising for a long time.
1 The Swedish FSA reintroduced mandatory depreciation in 2016, seetighten this ruleIn addition, a loan-to-value cap was introduced in 2010 and a number of measures to overcome bank lending were put in place.
Figure 3: The debt ratio of private households is worrying
A classic example of a search market
The real estate market is a classic example of a search market, which implies that the prices of the entire market (and hence household housing wealth) are determined by just a few transactions. However, while prices (and assets) can change quite quickly, debt (nominal) remains intact.
Keep in mind…
Before we go into how to interpret the above developments, it's probably good for you to know that I've had a negative (and wrong, as my better half always reminds me) view of the Swedish property market. However, my views have never been about a specific date for the inevitable collapse in house prices, but rather whether current prices are sustainable over the long term. Here I look up to Keynes, who said it quite eloquently:
"Markets can remain irrational much longer than you, and I can remain solvent."
Finally, we cannot say positively:That's crazy when it comes to a crash". A possible way out of the crisisBito see how house prices rise less than incomes over a period of time to give an idea of equilibrium. Admittedly, such slowly deflating solutions to asset bubbles are rather unusual. Regardless, it is clearly the aim of the Swedish FSA and other relevant authorities to achieve such a result.
How is the real estate market analyzed?
error correction models
Swedish analysts agree that property price estimates are overblown, but can largely be explained by fundamental factors. A number of statistical models have been developed in recent years to support these arguments. Most of these belong to the Error Correction Models (ECM) category. The advantage of such models is, of course, that they show quite clearly the relationship between property prices and other variables (interest rates, etc.) that are closely related to the property market. Among others, the Riksbank has evaluated such models2 on several occasions and has come to the apparent conclusion that “property prices are in line with fundamentals”.
Figure 4: Statistical models show that real estate prices are highly appreciated
note: 'HPR' is the price of houses in levels, 'd_p’ are house prices in logarithmic differences. The Riksbank used an after-tax interest rate, which we replicated by multiplying a range of interest rates by (1-td), where td is the tax deduction ratio used in Sweden in the sample period.
My preliminary update to the Riksbank model suggests that current house prices are somewhat high, albeit at a modest 10%, which should not worry policymakers.
Of course, these models have their merits, but they also suffer from the assumption (probably too strict) that correlated trends in the underlying data are constant. For example, it's hard to imagine how interest rates could go down much further...
In more fundamental models with long-run stable equilibria, house prices are explained by three main variables:
- user cost (tj.Interest charges)
- Budget share of living space
- Residential capital (in relation to the number of inhabitants).
When these three variables are "permanently stable," the real estate market is in equilibrium and real estate prices are at fundamentally correct levels.
Niels Bohr is often credited with saying:
"Predictions are very difficult, especially when it comes to the future."
Given that I'm having a hard time hitting the current quarter GDP growth forecasts, this is criticalforeverFinding stable levels for three composite variables isn't something I'm going to spend days doing.
However, some economists are undeterred by the difficulties in deciding about equilibria. And even if it's theirsEssaysThe Swedish Riksbank and studies by the Fiscal Policy Council3 concluded that Swedish house prices could be up to 40% too high as the country had reached a certain patina. Unfortunately, small changes in the assumptions in these models cause house prices to range from being correctly estimated to being dramatically overestimated.
Back to basics
For reasons we've already figured out, I prefer simpler, more robust measures. In the simplest fundamental model, asset prices are explained in terms of future utility (income) over the life of the asset. That future income is discounted to the present day using the amount I could have earned had I invested my savings elsewhere (the opportunity cost in this case is expressed as an interest rate). Part of this rate is also due to how much total income is expected to increase in the coming years.
In all honesty, it's almost impossible to calculate future earnings or interest with any certainty. For this and many other reasons, economists often use historical data—not just data from a few years or decades, but data spanning many decades (which are also not without challenges). Because over time, and on average, we can better assess what real estate prices should be relative to our income. I am somewhat reassured by the existence of a similar stock market indicator named after another elderly gentleman, Warren Buffett.
Figure 5: Buffettov-Indicator
note: In this chart, we used a composite house price index and a composite measure of per capita income. In addition, we have added a similar key figure for Euroland for reference.
While property prices in Euroland are only slightly (about 10%) above their long-term average, property prices in Sweden are almost 50% higher in the same comparison. Either the interest rates (risk premiums) should have been structurally reduced or our preferences would have changed significantly towards a larger share of the housing budget. If not, the alternative is that income growth will (hopefully) outpace house price growth for the foreseeable future.
The other side of the coin
So is that it? Are Sweden and the Swedish real estate market on the brink?
No, not necessarily.
Despite my pessimistic tendencies, I readily admit that there are valid arguments that this time around in Sweden is different.
Good old supply and demand
Given the strong population growth and despite the rapid increase in supply, demand appears to remain strong. As an indicator of the supply and demand situation, we can look at the number of completed houses in relation to demographic development. This indicator, albeit a simple one, shows that supply is not yet keeping pace with demand.
Figure 6: Still looking for apartments
Given the continued lengthening of the average payback period, low inflation and the aforementioned interest costs, it would be quite alarming to see household savings fall. However, the savings rate is on a clear uptrend and is currently near historically high levels.
Figure 7: The savings rate is at a historically high level...
wealth of private households
High savings and generally positive developments in asset prices contributed to a strong increase in households' net wealth. Even though foreign assets are price-sensitive, Swedish household net worth is significantly higher than in most (all) countries that experienced sharp declines in house prices during the financial crisis.
Figure 8: ...and the fortune is pretty great!
note: Analysts often relate assets and liabilities to disposable income. However, I would like to switch analysts to using GNI (per capita) as it is a much better measure of household well-being. Relating Swedish net wealth to disposable income would increase household net wealth to around 800%.
Financial situation of the government
In addition, the Swedish government's financial position is much stronger than in most comparable countries and (via very strong automatic stabilisers) provides an important shock absorber should the housing market and the economy fall into recession.
Okay, but what about household debt?
The last of the, in my opinion, really strong arguments for Sweden being different contradicts the main claim made by the Riksbank and Governor Ingves; “Worrying” high level of private household debt. To support this view, the Riksbank often displays a chart of household debt ratios, usually in comparison to many other countries with lower debt ratios.
Figure 9:The Riksbank has its own opinion. But is it balanced?
I contend that such a chart is quite misleading. Disposable income is measured after taxes and in Sweden (as in other Nordic countries) most social benefits are financed through taxes4. To correct these shortcomings, Swedish statistics actually produce a measurement of the so-calledadjusteddisposable income5. By swapping these two measures in the denominator and thinking superficially about the concept of debt (don't try that at home), we can actually create quite a wide range of debt ratios, all of which have something to say. It is important to note that no matter how we construct them, they are always lower than the Riksbank's debt ratio (black line) and Sweden is more mid-table in terms of debt ratio. - Find out!
4 The OECD has somebig datain addition.
5 This is a conservative estimate of how much welfare is spent individually.
Disclaimer: We generally do not have views and opinions on the economic and financial situation (at least not the ones that we express publicly as a company). However, we believe that people can and do appreciate different perspectives. What you have just read is the author's perspective. While we think our writers are very smart, Macrobond Financial does not expressly endorse the views expressed here. And as the old saying goes, don't believe everything you read (not without finding the data, doing some analysis, and putting it in a nice graph). We want to make it clear that we are not offering this information as investment advice. However, if you have Macrobond, you can easily check everything listed here and decide for yourself. If you don't have a Macrobond, now you have a good reason to get one.