We are born with a short position in housing

Felix Salmon has a lovely metaphor that helps me articulate something I've been wanting to say about the risk of buying a house:

"In that sense buying a house isn’t an investment, so much as it’s a way of permanently covering your built-in short position when it comes to the shelter market."

If you have a long position in TD stock, you receive annual dividends. If you have a short position in TD stock, you have to pay annual dividends (to the person from whom you borrowed the stock to sell it short). If you are neither long nor short TD stock, you neither receive nor pay annual dividends.

The rent on a house is like the dividend on a stock. If I do not own a house, and rent one to live in, I pay rent, which is like having a short position. If I own a house and live in it, I neither pay rent nor receive rent, which is like being neither long nor short. If I own two houses, live in one and rent out the other, I receive rent, which is like having a long position.

So we are born with a short position in housing. We need shelter, and must pay rent to live somewhere. When we buy a house to live in, we are covering that short position.

That metaphor drastically changes the way we think about the risks of investing in housing. Having a long position is risky. Having a short position is risky too. Covering a short position eliminates that risk.

Therefore, not owning a house, and having to rent one, is risky. Buying one house eliminates that risk. Buying a second house to rent out creates risk again.

Like all good metaphors it contains a kernel of truth; but like all imperfect metaphors it is not the whole truth.

The kernel of truth is that future rents are uncertain, and when we switch from renting to owning we eliminate that risk; but when we buy a second house and rent it out, we again face that same risk. Though now we fear falling rather than rising rents. The owner-occupier doesn't care about how much rent his right hand pays his left hand.

Why is that metaphor not the whole truth? There are many reasons.

1. My house may burn down, fall down, or in some other way stop providing me with the shelter I need. Some of those risks I can insure against; others I cannot.

2. The house may stay the same, but the type of shelter I need may change. I may need  a bigger house, a smaller house, or one in a different place. I face the risk that my old house may fall in price when I sell it, relative to the price of the new house I buy. True, but if my old and new house prices are positively correlated, I have at least partly hedged my risks, which is better than renting.

3. If I need to borrow money to buy the house, the future interest rate may be uncertain, and the risk that my mortgage payments will rise needs to be compared to the risk that my rents would rise.

4. If my house will last another 100 years, but I will only live another 40 years, I will have covered 60 years more than my short position. The future reverse-mortgage value of my house is uncertain. But if I make a bequest to my children this is not a problem. My children will have a short position in housing too. They will need somewhere to live.

5. Other reasons I can't think of right now.

Of course, none of this means it is always prudent to buy a house that is more expensive than the one you would otherwise rent. But there is a risk/return trade-off, so it could be prudent to pay a little over the odds (an insurance premium) to eliminate the risks of renting. (Not forgetting of course the risks cited above, plus the extra transactions costs of selling and buying a house compared to switching rental accommodation).

And we are also born with a short position in food, clothing, cars, etc. as well. So you reduce risk by covering those short positions by buying shares in companies that produce food, clothing, cars, etc. as well. Especially the sort of food, clothes, and cars you like.

And since we are born with a long position in labour, we would want to sell  labour short, if we could, to reduce risk. And use the funds to cover our short positions in houses, food, clothes, and cars. But we can't sell labour short. So instead we would need to use leverage: using mortgages to buy a house and buying shares on margin, thus facing the risks above.

Comments

[Lord]

Two other considerations:
You may not be able to rent a place that you can otherwise buy (often the case if rents don't cover ownership costs).
Whether rents can be expected to rise or fall (increasing or declining population and local growth or a company town).

[b0rk]

How does government regulation of rent alter this dynamic? Aren't the 'short' and 'long' (renter; landlord) positions asymmetrically risky, since there is a legislated cap to the increase in rent that tenants can be asked to pay?

[anon]

We're not short housing when we rent.

If we were, we'd make a capital gain with a decline in house prices.

And we'd have income from the investment that we made with the money we got from selling the house that we were short. That income would help pay the rent.

[Nick Rowe]

Lord: population changes etc. would be underlying reasons for changes in rents, and hence uncertainty over future rents. I'm not sure how to handle (think about) cases where you might not be able to rent (only buy), or might not be able to buy (only rent) the particular type of house you want to live in. Seems like risk could go both ways there. If you see your unique ideal home up for sale, you buy it; if you see it up for rent, you rent it. But the asymmetry is that if you buy it, you get to keep it as long as you want. If you rent it, you may not be able to keep renting it in future (12 month lease). (Perhaps that's the way of thinking about your point.)

b0rk: If you already are a tenant in a rent-controlled apartment, you are safe(r). But if you aren't, then rent controls may increase the risk of renting. Because you may not be able to find one, or one you want. (Rent controls create excess demand).

anon: if a share price falls, you make a capital gain when you cover your short. If house prices fall, you make a capital gain when you cover your short, in the sense that you pay less than you otherwise would have done.

And we do have income from the interest earned on the money we didn't spend on buying a house.

OK, the analogy isn't exact. Not owning a house is like having a short position in a stock, where you blew the money you got from shorting the stock. (Or your parents shorted the stock for you, spent the money, and left you with the short position).

[moz]

Most people don't need a whole house for shelter. But they're hard to buy in smaller units, so it becomes "do I buy more than I need and hoard or rent the excess, or do I rent excess from someone else? For what I pay for my shelter I could cover the mortgage on a garden shed+land, if I was lucky enough to find a block of land that small. But it would be small enough that I could take it with me if I wanted to move :)

Buying a house with a group of friends is legally interesting but it can work quite well. To date, friends who have done that have come out behind my position of investing the excess. But the government subsidises their choice much more than mine, so I suspect the low-risk option is also low-return once you exclude my involuntary contribution.

[Patrick]

"But the government subsidizes their choice much more than mine"

How so? Are you talking about the US? In Canada interest mortgage interest is not tax deductible...

[Andrew F]

In a couple ways. The government just introduced a first-time home purchase tax credit, and rental properties are taxed more heavily than owner-occupier residential (at least in Toronto).

[Nick Rowe]

I would say the most important way the Canadian government subsidises owner-occupied houses is that if you and your neighbour rent each others' houses, you each pay income tax on the rent you pay each other, but you don't pay income tax on the rent you pay yourself.

Maybe the capital gains tax exemption on owner-occupied housing too, given that we live in a world of 2% inflation, so houses can be expected to appreciate at around 2% in nominal terms.

[happyjuggler0]

2a) If I buy a house with massive leverage (e.g. 20% down payment), pay a higher monthly mortgage rate than I would renting (routine), my home price goes down (hard to imagine, eh?), and due to economic circumstances I lose my job and get a lower paying one, and then...I have to sell my house and start renting again, I:

Suffer a capital loss on my my massively leveraged "neutral" position in housing, and

Suffer a large opportunity cost on the difference between rent and mortgage+insurance+property taxes+maintenance that I've paid since being "housing neutral" that I could have invested in, like, an actual investment.

Buying a house is not the equivalent of being on the sidelines.

I love the blog by the way, both before you showed up and since then. I also enjoy your comments at Sumner's blog too.

[Phillip Huggan]

There is an oil sheikdom, possibly Saudi Arabia, where they start out each citizen with $1 or $2 hundred thousand.
Assuming $500/month room or bachelor suite rental, is $6000/yr. If a wealth manager can make %6/yr, it would take a one time grant of $100000 to pay out a cheapie rent annuity. Could make it $25000 Universally, and 2x for HS, 3x College and 4x for a university degree. Permit adults easy re-entry. Maybe the $$ estimates are off, but the idea is you can directly remove the risk without derivatives (always a good rule of thumb).

#2: Flex housing, houses that can be turned into multiple units as a family moves out and adults age, is a solution. Assuming tenants will always have money to rent from landlords is the downfall of your hedge. Could move to safer bonds. Could invest in Walmart, Videogame makers and TV networks, food retailers (cheap cocooning corps), Staffing companies, Gold, India...buy a hotdog cart outside Value Village, don't increase finance sector churn unnecessarily.

[Nick Rowe]

happyjuggler: Thanks!

I've been trying to get my head around your 2a, and fit it into my schema.

Suppose I lose my job, and need to switch to living in a house exactly half the size, price, and rent, of my current house (a 50% downsize). Assume that prices and rents of 1/2-sized houses are always exactly 1/2 of full-sized houses, so they are perfectly correlated.

I think I would say that I suddenly switch from having an innate short position in 1.0 houses, to having a short position in 0.5 houses. So if I own 1.0 houses, I suddenly find myself in a net 0.5 long position. Which is risky. And I expect if that 0.5 long position is leveraged, it's even more risky.

I'm not 100% sure I've got that right. It's a new way of thinking for me.

Phillip: Yes, your Saudi example is cleanest. Typically, we have to borrow against our long position in labour in order to cover our short position in housing (we get a mortgage to buy a house and use our wages to pay the mortgage).