Please be careful to separate the price of houses as a capital good from the price of monthly rent.
I'm arguing that all else being equal, if the price of houses goes up, the price of housing as a service (ie rent) can go down under certain circumstances.
Specifically also the opposite: interventions that make houses drop in price, eg restrictions on lending like we saw in the wake of the Great Recession, can lead to less construction, and thus rents going up.
The tightening of lending standards in the wake of the Great Recession saw the ratio of rents to house prices go up a lot.
There are really two types of investors we have to regard here. One is ordinary landlords, but landlords buying properties and renting them out isn't a problem; the more of them there are the lower rents get and then the fewer of them they are. It naturally finds equilibrium.
The other type is global institutional investors who are using real estate as a substitute/diversification against government bonds and then not renting it out, because their desire is for a stable investment rather than high risk high reward. Because of artificial constraints in the housing supply, appreciation in housing prices has been beating the bond market without even renting the properties out, and renting them out is actually a risk because bad tenants in jurisdictions with adverse landlord-tenant laws can be a huge problem. So they just buy up whatever is available and take profits as property appreciation rather than rent.
The second type of investor is obviously really bad for housing affordability, but it also only happens when construction is restricted from increasing supply to meet demand, because otherwise the property itself wouldn't appreciate faster than inflation and would make a poor investment in the absence of rental income.
Your second type of investor has exactly the same opportunity cost for not renting out as everyone else, don't they? Conversely, your first type of investor has exactly the same problems with bad renters? So why does anyone rent anything out?
It looks like the problem of the occasional bad tenant can (and is) solved in practice via diversification. You buy a REIT instead of an individual property.
In my earlier nomenclature, I would classify your 'second investor' as a personal union between an investor and a 'weird' tenant. A 'weird' owner-occupier, who's never home.
> The second type of investor is obviously really bad for housing affordability, but it also only happens when construction is restricted from increasing supply to meet demand, because otherwise the property itself wouldn't appreciate faster than inflation and would make a poor investment in the absence of rental income.
Well, a land value tax would presumable also help here. But again, the opportunity costs for renting out vs not renting out are exactly the same.
> Your second type of investor has exactly the same opportunity cost for not renting out as everyone else, don't they? Conversely, your first type of investor has exactly the same problems with bad renters? So why does anyone rent anything out?
People have different levels of risk tolerance, and different resource levels.
For example, there are ordinary landlords that will take out a mortgage on a building, rent it out and then use the rents to pay the mortgage and the salaries of the people who fix leaky pipes and find replacement tenants when someone leaves etc. If they don't charge rent they're out of business because they'd have nothing to use to pay the mortgage and the bank would foreclose, so they have to.
Then there are investors who just want somewhere to park their money. There is no bank to foreclose, they're paying cash, so they're happy as long as the property is appreciating by more than e.g. the interest rate on government bonds. They can drain the pipes and turn off the heat and make money by doing nothing, as long as supply is constrained so the price keeps going up. Meanwhile they're risk-averse, so the extra income from renting it out (much of which is offset by increased maintenance costs) isn't worth the risk of e.g. some law preventing them from evicting a non-paying tenant.
> It looks like the problem of the occasional bad tenant can (and is) solved in practice via diversification. You buy a REIT instead of an individual property.
That's assuming you can diversify the risk. If someone wanted to invest in, say, San Francisco real estate, they could diversify across multiple properties, but then they'd all be affected by some new law there preventing evictions. Whereas they could diverisfy by also investing in properties in Ohio and Arkansas, but those might not be expected to appreciate enough to attract their interest.
> [...], but those might not be expected to appreciate enough to attract their interest.
I don't understand how that is supposed to work. I can see how a restricted housing supply in San Francisco can lead to us reliably anticipating that the prices there in, say, five years, so in 2030, will be high. I'm with you so far.
Also assume that we can reasonably reliably anticipate housing prices in Ohio in 2030.
But if the buyer is smart enough to figure that out, why wouldn't the seller be? For San Francisco's housing prices to _appreciate_ more than those in Ohio, today's seller needs to accept a bigger discount (compared to reliably estimated 2030 prices) for San Francisco real estate than for Ohio real estate. Ie SF housing prices need to be comparatively lower.
Why would the seller be so nice for SF but not for Ohio? Are current SF land owners less interested in money?
By default, current prices (should) anticipate market consensus expectations of future prices.[0]
> People have different levels of risk tolerance, and different resource levels.
Yes, but they all buy and sell in the same market, don't they?
Your hypothetical 'investor' is (A) risk seeking enough to willing invest in single undiversified properties and return seeking enough to go for SF over Ohio (assuming there's more of a return there, I don't know if that's true). And simultaneously (B) risk averse enough to even contemplate having tenants, and uninterested in returns, too, as long as they are barely above government bonds?
What kind of weirdly split personality creature are we dealing with here? (And where can I find them? I suppose I could try and sell them the Brooklyn bridge.)
Are you sure there isn't something else going? Perhaps some other weird law, or perhaps some weird tax dodge (or even status game?) that leads some investors to leave properties empty? What proportion of those properties are even empty? What proportion would we expect from purely profit seeking risk-neutral investors? (That null-hypothesis proportion is most likely less than 0. If even just because of tenants moving in and out.)
> That's assuming you can diversify the risk. If someone wanted to invest in, say, San Francisco real estate, they could diversify across multiple properties, but then they'd all be affected by some new law there preventing evictions.
You can at least buy an SF REIT, so a single bad tenant has less of an influence. And, new laws are unlikely to affect both commercial and residential renters the same way, so you can diversify across that dimension.
Even without tenants, an SF REIT with only empty properties makes more sense for diversification than a single property. After all, tenants and city wide regulations aren't the only thing affecting the prices of a single house.
[0] As a caveat to what I wrote above: if one asset class is considerably more risky than another, we can expect today's buyers to demand / today's sellers to offer a bigger discount on future prices. Ie on average returns on risky assets can in theory be higher. Risk vs reward.
But that explanation doesn't work for the picture of the hypothetical investor you sketched in the rest of your comment.
> For San Francisco's housing prices to _appreciate_ more than those in Ohio, today's seller needs to accept a bigger discount (compared to reliably estimated 2030 prices) for San Francisco real estate than for Ohio real estate. Ie SF housing prices need to be comparatively lower.
> Why would the seller be so nice for SF but not for Ohio? Are current SF land owners less interested in money?
The reason is that there are two classes of buyer; investors and occupants. The expectation in SF is that population will increase and the housing supply will not be allowed to increase, so occupants will have to pay more in the future, creating the expectation of future appreciation. Investors could then come in and bid up the current price in anticipation of this, but they'll only do this until the current price rises to match the risk-adjusted returns from competing investments.
The expectation in Ohio is that the population will be approximately flat and new construction will be less inhibited, so expected housing price appreciation there might be less than competing investments. That discourages external investors, but the current-day prices are still held at their current level from demand by local residents because they still need somewhere to live.
> Are you sure there isn't something else going? Perhaps some other weird law, or perhaps some weird tax dodge (or even status game?) that leads some investors to leave properties empty?
These things could be contributing to it. You could also add bureaucratic incompetence to the list, i.e. an institutional bureaucracy gets an order to invest in housing so it buys a bunch of housing and is too insulated from consequences to recognize that it could make more money by renting it out.
But all of these derive from the expectation that housing will appreciate and thereby make a good investment. Otherwise whatever weird law or unusual risk function is causing them not to rent the properties out would be irrelevant because they would lose the incentive to invest in them to begin with.
> Even without tenants, an SF REIT with only empty properties makes more sense for diversification than a single property. After all, tenants and city wide regulations aren't the only thing affecting the prices of a single house.
Sure, but then you have a REIT buying up properties and leaving them empty, which is no better than doing it one at a time, and moreover can likewise be prevented by not inhibiting construction to remove the expectation of market-competitive returns solely from housing appreciation.
> The expectation in SF is that population will increase and the housing supply will not be allowed to increase, so occupants will have to pay more in the future, creating the expectation of future appreciation.
See, I don't understand this.
So I grant you that rents will increase in SF. That's what you are describing. (Conversely, current rents are lower than rents in five years; and there's no market mechanism to bring that rent increase from 2030 to 2025.)
But I don't get the proposed mechanism to turn that relative discount of 2025-compared-to-2030 rents into a relative discount of 2025-compared-to-2030 house prices.
There's a clear mechanism in the markets to bring assets prices from the future to the present, but there's no clean mechanism to bring rent increases from the future to the present.
That's because you can buy a house now, and hold it for the next five years. But you can't turn a 2025 month of rental services into a 2030 month of rental services.
> The expectation in Ohio is that the population will be approximately flat and new construction will be less inhibited, so expected housing price appreciation there might be less than competing investments. That discourages external investors, but the current-day prices are still held at their current level from demand by local residents because they still need somewhere to live.
That argument seems very confused to me.
So is the return to funds invested in housing in Ohio higher or lower than in San Francisco? (That includes both capital appreciation and rental income (or the 'opportunity profits' [0] of not having to pay rent)).
You seem to be mixing up (imputed) rental income and the cost of houses as a capital good.
Local residents are responsible for the rent in either place. Both of us agree that we expects rents to grow quicker in San Francisco than in Ohio.
> Investors could then come in and bid up the current price in anticipation of this, but they'll only do this until the current price rises to match the risk-adjusted returns from competing investments.
> That discourages external investors, but the current-day prices are still held at their current level from demand by local residents because they still need somewhere to live.
Contrasting these two sentences, looks exactly backwards from what we are trying to explain? You need to explain why current house prices in San Francisco are so _low_, and Ohio house prices are so _high_, both compared to their expected 2030 levels.
[0] I'm not sure if 'opportunity profits' is an established term. I mean the opposite of 'opportunity costs': the imputed implicit profit of avoided costs.
Please be careful to separate the price of houses as a capital good from the price of monthly rent.
I'm arguing that all else being equal, if the price of houses goes up, the price of housing as a service (ie rent) can go down under certain circumstances.
Specifically also the opposite: interventions that make houses drop in price, eg restrictions on lending like we saw in the wake of the Great Recession, can lead to less construction, and thus rents going up.
The tightening of lending standards in the wake of the Great Recession saw the ratio of rents to house prices go up a lot.