Deadly blows to the Castellum case

Today I will look into on a favorite among Swedish dividend investors and promoted as a strong buy among bloggers - Castellum. A compilation shows that 26 popular blogs own Castellum (link). Data in this post will be in Swedish, sorry for that.

The Castellum Investment Case

In summary, the investment case looks something as follows:

• Castellum is the largest public commercial real estate company in Sweden and is a very stable foundation in a long term portfolio. 
• Castellum have increased dividend every year since listed on the stock exchange 1997. The company have an objective of increasing income from property management (förvaltningsresultat) with 10% every year. Although this is a stretch it is very likely that Castellum will continue as before - Increasing profits and dividends every year. 
• With a current dividend yield of 3,76% it is a fair price for a wonderful company

The history sure looks promising::

Today I will deliver three, in my opinion, deadly blows to the investment case

1. Historically unsustainable investment pace and inflated property valuations
2. Inflated income from property management driven by low interest rates
3. The stock is trading at historically high levels on this inflated valuation

1. Unsustainable investment pace and inflated property valuations

The table above shows the annual net investments for Castellum. Subtracting "income from property management" we get a proxy of annual "over investments". This is because:

• Castellum is using 52% of income from property management for dividends
• Castellum has a target 50% debt to assets

Example: 

Castellum had a profit from property management of some 1,5 B SEK in 2015. A bit more than half of this result is used for dividend and the other half 0,75B SEK is used for investments in new properties. Their target of a debt to asset of 50%, provide an opportunity to borrow 0,75 BSEK, leaving the sustainable investment level at 1,5 BSEK. 

This sustainable investment level is for a balance sheet in equilibrium, with no spare debt capacity or need to repay debt. Goeff Gannon has a great post on debt capacity and impact on returns which is a must read. 

Totally they have overinvested 5 billion SEK the last 10 years. The debt to asset ratio has increased from 45-49%, which is close to the target level of 50%. The reason the debt level have not increased more can be seen in the following table:

10 Years Overview

Dividing the "operating profit per square meter" with the "valuation per square meter" we get the return requirements implied in the real estate valuations. Compared to 2006 we have had a non-fundamental positive valuation effect of assets of 4,72 BSEK, driven by lower return requirements. 

It should be noted that the return requirements 2006 was not particularly high with 7,146%, we had higher return requirements 2012 of 7,231% and if we go back to 2003 the return requirements on properties was 7,99%.

Conclusion:

Castellum has been growing at a rapid rate the last 10 years. This growth have however been inflated with one-offs namely

1. Increased debt level
2. Record high valuation on properties, due to decreasing return requirements from the market

All things the same as today, the future growth will be slower than the past 10 years. However, the big risk is of course that the pendulum swings in the other direction

Example: 

What happens if the return requirements on properties (valuation) normalizes on the 2012 level?  

In order to keep the target 50% debt to asset ratio, Castellum need to amortize more than two times the 2015 income from property management. If they where to keep the current dividend level we would have 4 years with no investments in assets and growth.

Market turmoil generally move the property valuation in the wrong direction, and higher interest rates would over time certainly normalize property valuations. That takes us to the second blow to the Castellum case, the 2015 years inflated income from property management driven by low interest rates

2. Inflated income from property management due to low interest rates

In the annual report from 2015 we can read the following:

"Den genomsnittliga räntebindningstiden uppgick per 31 december 2015 till 2,5 år (2,8).................Kassaflödesmässig resultateffekt nästkommande 12 månader vid en ränteförändring om +/– 1% uppgår till – 78 respektive – 30 Mkr. På grund av räntegolv i kreditavtal har Castellum inte möjlighet att dra full nytta av negativa räntor, därav uppstår ett negativt utfall även vid en sänkning av räntan med 1%-enhet."

We can conclude that any interest rate changes have negative effect on the income from property management. By looking at the interest maturity structure below we can make this calculation ourself:

Calculation:

• 1 % higher interest rate next 12 months gives:
• Interest hike of 1% on 9,46 BSEK of debt with a average binding period of 0,2 months equals some 77 MSEK income effect

What would happen to results if interest rate normalize to historical average?

• Normalized income from property management would be ~20% lower relative to the 2015 result
• • Castellum have a average interest rate of 4,74% since 1997 and had record low interest rate 2015 of 3% 
  • Interest bearing debt of 20,4 BSEK 
  • If interest rates normalize to historical average the annual interest costs would (over time) increase from 602 MSEK to 947 MSEK. 
  • That implies a decrease of 22% on 2015 years income from property management

Does rental price adjustments acts as a buffer?

Some bloggers have argued that an interest hike have relatively small effect for Castellum, since price adjustments in rent levels are linked to interest rates. Here is what Castellum says on rent price adjustments in the 2015 annual report:

"Hyresavtalen innehåller normalt en bashyra, d v s överenskommen hyra vid avtalets tecknande, samt en indexklausul som innebär en årlig för- ändring av hyran motsvarande en viss andel av inflationen under föregående år, alternativt en procentuell minimiuppräkning"

I find no evidence that they have a direct link between interest rates and rental pricing. Both the statement above and the calculation of interest rate sensitivity indicate that higher interest rates directly hit the results. 

Conclusion:

Current income from property management is inflated. Normalized levels are significantly lower due to the record low interest rates. Actually, changes in interest rates can only have a negative effect for Castellum. 

Furthermore, I do not see any cyclical factors affecting the result not being on the high end of historical averages. For instance, vacancy rate 2015 is at record low of 9,7% 2015 relative to a historical average of 11,1%. 

As we conclude that the 2015 income from property management is inflated, we move on to the final and mortal blow to the Castellum case. A historical high valuation on this inflated income.

3. Historical high valuation

Summing up

Castellum is a good commercial real estate company. They have stated an ability to buy properties with high vacancy rates at lower prices and with their size and related customer relationships/channels quickly fill these properties. A rare competitive advantage in the real estate sector. 

However, at current price levels I do not view the expected value of the stock as attractive over a 5-7 year period. We have correlating factors, providing quite substantial downside risk in mean reversion, namely:

• Reversion of current record high property valuations would imply a need for Castellum to use income from property management to amortize debt 
• Income from property management normalizing at a lower level, driven by higher interest rates
• A stock valuation reverting to historical levels

Usually the pendulum swings both ways and the real estate industry have feedback loops built into it which are quite interesting. See illustrative example below:

Confirmation Bias

I´ll encourage you to comment on any mistakes you see in above reasoning for mutual learning. Castellum is a favorite among the Swedish investing community and I expect few shareholders reading this to take any action. I leave you with some nice pictures of confirmation bias, something we all must fight as investors:

BR / RQ

Full disclosure: I do not have any long or short position in Castellum