Hacker News new | ask | show | jobs
by apo 2500 days ago
This is all the more surprising given that inflation in Denmark is running ~1%/year.

That means the bank is absorbing a real loss of 1.5%/year on the deal.

Meanwhile, the housing market is roaring:

https://tradingeconomics.com/denmark/housing-index

It would be one thing to see negative rates in a declining property market and/or in a deflationary environment.

However, negative rate mortgages are happening in what looks like a normal economy with low inflation and red-hot housing market.

Somebody is very wrong about one or more of the following:

- the future direction of the housing market

- the future direction of interest rates

- the future direction of inflation

Edit:

Google translate sheds some light. It appears that the Danish equivalent of "points' clouds the picture:

> Total repayments before tax DKK 277,392 (on a DKK 250,000 loan) - of which total interest and contributions DKK 8,264 changes in the exchange rate will affect the size of the amount paid out.

https://translate.google.com/translate?hl=en&sl=auto&tl=en&u...

Low, but not negative.

7 comments

Some background: In Denmark, mortgages for housing are handled by special real-estate lenders (realkreditinstitutter) who issue bonds and handles defaults.

These bonds are considered to be very stable, about the same quality as state bonds, as the lenders will only lend up to 80% of the value of the house (an ordinary bank loan must be used for the remaining fraction) so can usually recover most of the money through a forced sale. Also there's a fee on top of the bond rate to absorb the losses as far as I understand. The fee depends on the security percentage, so if the lender only has a security in the top 60-80% part of the value of the house, the fee is much higher, see the tables here:

https://www.mybanker.dk/sammenlign/bolig/bidragssatser/

The Danish central bank has had negative interest rates for some time, and these real-estate bond rates have also been falling.

When taking out a mortgage, you can either opt for a fixed rate for up to 30 years, or variable-rate bonds where the interest is redetermined periodically, e.g. once a year. The variable-rate bonds have been negative for some years now. The news here is that the rates have been falling even more lately, and that you can now get the negative rate fixed for up to ten years.

In any case, it's the bond investors who are losing money, not the real-estate lenders who apply a fee on top, both as a rate and a fixed fee when applying for the loan.

The rates are so low that I think that if you're in the low-risk category, you might end up with a small negative total rate. Of course, the fixed fees when obtaining the loan may eat that.

>These bonds are considered to be very stable, about the same quality as state bonds

>can usually recover most of the money through a forced sale

This exact line of thinking is what led to the US real estate crash in 2008/2009

It's a gross simplification, but with the volume that is Danish housing vs demand for said housing the risk is very low.

One of many issues leading into the 2008 crash was how Us mortgages were bundled, rated AAA, then resold as an investment tool. However, the ratings were falsely boosted to promote investment and when the underlying junk mortgages fell through and demand fell off a cliff with the rest of the economy there was no way to flip foreclosed housing to make up losses.

In this case, Danish mortgages are not bundled, rated falsely positive, and not sold as an investment tool like in 2008. Current US auto loans may be much more similar than these Danish loans.

> In this case, Danish mortgages are not bundled, rated falsely positive, and not sold as an investment tool like in 2008.

Danish loans are definitely securitized and sold as investment tools. They are a classic interest rate risk hedge.

It’s been a while since I was adjacent to them but the bigger difference in Danish mortgage backed securities were a) no governmental guarantee on them b) less protections for the borrowers than in the US c) no derivatives d) much smaller market & e) much longer history.

So credit risk, which is what ratings agencies nominally judge is likely not a problem.

Interest rate risk is though, but given its more straight forward, it’s what investors are looking for exposure to & there is low leverage it’s unlikely to cause systemic collapse.

low leverage in proportion to the total asset value

But what’s that total asset value if interest rates are 10% (vs -1%)?

Well in principle yes, but in reality no. The Danish system for financing real estate have historically been extremely stable. Olau explains it above
That’s true, except that you can only get a loan in Denmark if you have 20% of the purchase price to put down as collateral. That means sub prime mortgage lending doesn’t exist in Denmark and that combined with liar/NINJA loans was a substantial contribution to the US housing crash, together with securitisation. If it had been impossible to securitise mortgages with less than 20% down in collateral the housing crash would have been a non event in the financial markets.
> the lenders will only lend up to 80% of the value of the house (an ordinary bank loan must be used for the remaining fraction)

From olau's explanation, it seems that there is enough money flowing, that more than 80% can be financed (in some way) even if it is not in a single or collateral-backed loan.

Indeed.

I think the situation currently is that you need 5% in downpayment yourself, then the bank can supply the 15% (at a much higher, individual rate) and the real-estate lender will provide the most secure 80%.

This is not really my area of expertise, and it's unclear whether you're basing this comment upon an actual analysis or it's just a quick remark?

I was trying to explain how the bond rates can go so low.

The real-estate lenders have rigid policies in place as for who can get a loan.

Of course, enough correlated failures, and the system goes down.

But there was a crashing housing bubble in Denmark too in 2008/2009, and to the best of my knowledge none of real-estate lenders went down, although they did increase their fees to cover losses. Banks were crashing, though, until the bailouts started.

I personally think that the overall system would be more healthy if the real-estate lenders and banks in general did not protect their investors against losses, but just passed them on so you avoid this sure-everything-is-fine until it crashes. But again I don't know much about the banking world. Maybe that idea is too complicated.

And yes, the negative interest rate at the Danish central bank does translate to negative interest rates for deposits. Not yet for consumers, but corporate accounts have negative rates.
Negative central bank rates are an attempt to spur consumption via all of the stimulus created since the financial crisis. Banks have been loath to pass on these negative rates to depositors because no one wants to be the first bank to do so, as depositors would flee that bank for another bank that was still eating the negative interest rate. However, the banks legally cannot coordinate actions as that would be evidence of cartel behavior. So they have just eaten the losses thus far.

This appears to be a first step to pass negative rates onto consumers, in this case onto borrowers where the first mover has an advantage.

US banks have proven (to themselves at least) that they can get away with sub-1% savings rates for long term, regardless of what the Fed rate is.

The first one to do it will be a pariah (briefly) but the next 5 will be "just following the trend" - the question is who will be first?

That sounds similar to how Mortgages used to work in the UK in the 60s.

One would get a mortgage from a Building Society for around 75% of the property value, and had to get a Bank Loan for the remainder which they couldn't cover from savings.

What has changed? How do mortgages work today in the UK?
We had 100% mortgages (no deposit) in the run up to the crises.

We even had 100%+ mortgages for a time which included a loan on top of your mortgage to modernise and furnish your property.

Nowadays, 5% deposit mortgages are available but 10% is more typical. We also have a popular government “help to buy” scheme which some think is reckless as it allows people to get into a more expensive new build home (which typically fall in value after first occupier) with a 5% deposit.

A loan for your deposit would not be acceptable as the mortgage company want to see you have skin in the game.

UK housing market is a complete mess and in my mind hugely overvalued. The can has been kicked down the road for over a decade now. If I didn’t have a family who needed secure housing situation I would sell to rent in a heartbeat.

All Danish “realkredit” morgages are done with bonds, so the “bank” is not loosing money, the buyer of the bond is.

For some loans, on some conditions, the morgage is actually negative, and lender is either paid a quartely price, or the debt is lowered accordingly. (This is for the morgage named F5, with a fixed 5year rate, and when the house has very low debt, then the fee charged, is also lower, as it is based on risk)

As far as I know, fees and charges on the loan make it about 2% anyway. Still a pretty cheap loan, I'd say.
Good point - updated answer with a link. Hopefully there's a better one out there.
Your housing index link seems to say that housing prices increased by 4% per year over the last two years (looking at the differences Jan 2017-Jan 2018-Jan 2019 [edit: clicking "10Y" shows essentially the same trend since 2012]). So unless nothing else in the economy increased in price over this period, I would say that whoever calculated a 1% inflation rate did not weight housing costs adequately.
Housing prices normally aren't a part of the inflation calculation.
"Normally" as in Denmark, or "normally" as in "anywhere in the world", or for some other meaning of "normally"? Why wouldn't it be if inflation is meant to be a measure of changes in the expenses of average households?

https://en.wikipedia.org/wiki/Inflation says "The measure of inflation is the inflation rate, the annualized percentage change in a general price index, usually the consumer price index, over time." and links to https://en.wikipedia.org/wiki/Consumer_price_index which says "The index is usually computed monthly, or quarterly in some countries, as a weighted average of sub-indices for different components of consumer expenditure, such as food, housing, shoes, clothing, each of which is in turn a weighted average of sub-sub-indices." and goes on to give an example (apparently ficional) in which housing makes up 41.4% of the index.

Edit: My copy of Samuelson and Nordhaus, Economics, 19th edition, has an example of a consumer price index including housing weighted at 42.4%. At that weighting, unless my math is off, the remaining 57.6% of the stuff in the index would need to get cheaper by 1.2% to get to overall 1% inflation with a 4% increase in housing prices:

    42.4 * 1.04 + 57.6 * (1 - 0.012) = 101.00479999999999
Some things do get cheaper over time (like consumer electronics, sometimes), but others not so much.
- Normally as in almost everywhere in the world (apart from a few places like Sweden I think?)

- Buying a house is not a living expense, it is a capital investment. This is the reason why statisticians either only include rents (e.g. the EU Eurostat) or replicate the housing costs of owner-occupiers with an “owners equivalent rent” (the US BLS does this).

- Housing might be 30% or more of CPI in the US but none of it is house prices for the reason above. Most of it is rents and owners equivalent rent, some of it is furnishing costs, some of it is utility tariffs, some of it is costs of repairs and maintenance.

> Normally as in almost everywhere in the world (apart from a few places like Sweden I think?)

Aren't you contradicting yourself when you go on to say that both the US BLS and Eurostat do include housing in their indices?

> statisticians either only include rents [or] “owners equivalent rent”

OK. So housing is part of the CPI, yes?

If I understand https://www.ecb.europa.eu/stats/ecb_statistics/escb/html/tab... correctly, Eurostat weights "actual rentals paid by tenants" at 6.1%. (edit: ECB, not Eurostat)

> none of it is house prices

This is the first mention of "house prices" in this thread. The index linked in the ancestor comment called itself "housing", not "house prices". Good for you if you insist on making the difference, but then if you make the difference you are not talking about the thing I was talking about, I think.

Also, some of it is house prices if house prices factor into the "owners equivalent rent", or is that made up out of thin air? Also, rising house prices cause rising rents, so indirectly they are represented anyway.

This is without up to date fact checking so take this with a grain of salt, but during the last housing crisis I remember looking this up and came to the conclusion above. I don't actually know if it's universal but it was valid for Sweden at that point in time. I think the key word in your text above is "expenditure", where a house maybe isn't seen as an expenditure but rather an investment..?

When taking about inflation it's also always good to remind one self that the pool of money also changes. And the fact that housing usually is connected to mortgages, which in turn expands the amount of available money. But this paragraph has nothing to do with your comment so I'll just stop there!

Thanks for your comment. I was a bit intrigued, also in light of the sibling discussion, so I tried to look this up without spending too much time on it.

This document "How is inflation measured?" from the Riksbank: https://www.riksbank.se/globalassets/media/rapporter/ekonomi... doesn't give weights in general but does say this (emphasis mine): "the method in which housing costs for owner occupied housing are measured are an important source of non-comparability, partly because they make up a large proportion of the CPI (around 10 per cent in sweden)". Note the "owner occupied" part, so this isn't even just rents! But then I don't understand the rest of the discussion that follows this.

Also Statistics Sweden at https://www.scb.se/en/finding-statistics/statistics-by-subje... lists "Housing, water, electricity, and fuels" as one of the "main groups" in the CPI, though this only suggests but does not prove that housing does have a non-zero weight.

"The inflation rate" generally means CPI-U[1] in the US or some similar "basket of goods" in other areas. So, this is correct, as we're usually looking at urban consumers, it would include rent but not housing.

As another poster mentioned, though, inflation or deflation as general concepts can refer to any price or cost that increases or decreases.

[1]: https://www.bls.gov/news.release/cpi.t01.htm

> low inflation and red-hot housing market.

The more reasonable way to look at it is that it is high inflation and a weak housing market. If the housing market was so hot, why would interest have to drop below zero?

Rather, houses are extremely expensive because money for houses (i.e. mortgages) is extremely cheap.

Why is it strange to you that the housing market is also roaring?

Cheaper debt means people take out larger debt to continue buying houses at market rate. Housing supply continues to deplete and people continue to pay a premium as more expensive houses are the only thing on the market. Lending drying up for the non-capital class will deteriorate the housing market.

I think a lot of headlines and people are conflating the negative interest rates with other unrelated signs.

>That means the bank is absorbing a real loss of 1.5%/year on the deal.

(As I understand it) the Eurobor rate is negative (the rate that the central bank lends to the other banks), so its unlikely that the bank are taking the hit directly.

Euribor is not a rate that central bank lends to other banks, it is a rate commercial banks lend between themselves.

ECB has its own rates (e.g. deposit facility, marginal lending facility for overnight deposits and loans).

As I mentioned somewhere else, these loans, are based 100% on bonds, sold on the open market, so the issuer is not loosing money. The issuer “realkredit selskab” just gets a fee.

Who buys the bonds, could be anyone.. banks, pensionfunds, people who need to store money, in an “insured” way.