I think that equilibrium in the housing market satisfies two conditions:
market clearing in rents
There is a demand for housing units in a particular area, and there is a supply of units. Note that demand and supply are not fixed numbers. They are functions of price, as in freshman econ. The price in this case is the rental rate on housing units. The market clears when the rent is such that there is neither excess demand nor excess supply.
Given the market-clearing rent, the purchase price of a housing unit reflects rent-buy parity. I was tempted to call this rent-buy arbitrage, but you cannot exactly arbitrage in housing, so I went with the term parity. Buyers are willing to pay enough to make the economic cost of owning a house about the same as renting, but not more. Because of rent-buy parity, when rents go up, prices go up, and conversely. Note that the economic cost of owning the house is not the monthly payment. It includes taxes, depreciation, interest foregone on the down payment, minus price appreciation. The price appreciation part is especially important. When prices rise rapidly, owners effectively live rent-free.
Aren’t single-family homes different from apartments? There are differences, on average, but there is enough overlap, as well as back-and-forth movement between owner-occupied and rented properties, that I believe that it is better to think in terms of market clearing in rents rather than treating the rental market and purchase market as separate from one another.
What about rent control? The devil is in the details. But I think that in most cases rent control is not binding at the margin. At the margin, people are moving in and out of an area, and usually the rent can change when people move. So the marginal renter pays something like a market-clearing rent, while people who have been in apartments a long time under rent control get below-market rents (and probably lousy maintenance.)
Rent control can also mess with rent-buy parity. Properties that might otherwise be rented might be converted to condos if that is a way to evade rent control.
Making money in the housing market is either really easy or really hard. Lawrence Person describes some corporate house flippers who found it really hard.
I think it’s mostly luck. If you buy right before something unexpected happens that makes your property more valuable, it’s easy to make money. If the opposite happens, making money can be really hard.
Because the interest rate is pretty much the same everywhere, the ratio of rent to price should differ primarily because of expectations of price appreciation. Where price appreciation is expected to be greater, the rent/price ratio should be lower. Suppose that is not the case, and you have the same rent/price ratio in city A, where expected appreciation is 5 percent and in city B, where expected appreciation is zero. Then in principle you could sell property in city B, buy in city A, and earn an arbitrage profit. In practice, of course, the transaction costs and risk make it all but impossible to arbitrage housing markets in that way.
Note: as this post was going to press, Kevin Erdmann provided an extensive discussion of the drivers of house rents and prices.
Great post. As I have been in this business for 20 years (albeit, in apartments, not houses), I definitely have some thoughts on this subject.
1. "The market clears when the rent is such that there is neither excess demand nor excess supply." Correct. That's Econ 101.
2. "The purchase price of a housing unit reflects rent-buy parity" I would say this is approximately correct, but not always. For many reasons, a business of buying homes for investment purposes is not always a great idea. Most people buy houses to live in them, not rent them. For various reasons (e.g., tax treatment, less wear-and-tear for owner/occupied houses, etc.), the economics of owner/occupied homes is not the same as homes owned for investment purposes. As an analogy, think of the differences between owning your car for your own use, versus renting it out. Renting your car out to 3rd parties is usually a bad idea, economically speaking.
3. "But I think that in most cases rent control is not binding at the margin." Probably not true. It is correct that many investors ignore rent control, but they do so at their peril. Rent control essentially turns an equity investment into an investment where the upside is capped. This dings your returns considerably.
4. "Properties that might otherwise be rented might be converted to condos if that is a way to evade rent control." Wrong. The clever folks who draft rent control ordinances have definitely thought of this. Almost all rent control ordinances have strict and very thorough restrictions on conversions.
5. "I think it’s mostly luck. If you buy right before something unexpected happens that makes your property more valuable, it’s easy to make money." Sort of. But prices of privately owned real estate tend to be "sticky." This presents an opportunity for long-term investors who are patient. There is a small group of investors of privately owned assets (real estate included) who focus on buying during periods of distress. This group almost always succeeds. Admittedly, this is a small group, but not tiny. Returns among members of this group is due more to skill than luck.
6. "the ratio of rent to price should differ primarily because of expectations of price appreciation." This sounds logical, but I have not found it to be the case. In my experience at least, most real estate investors don't forecast future rents very far into the future. As a result, they don't place much emphasis on price appreciation. Not exactly sure why this is the case. Use of leverage is probably part of the story.
As a side note, your emphasis on the "rent to price" ratio is not exactly correct. Most real estate investors value purchases based on a ratio of price to cash flow (unleveraged), not price to rent This may seem like a subtle distinction, but it is not. I once had a conversation with one of my buddies from MBA school, and he told me about his awesome plan to buy up houses in Fresno California, then rent them out. He cited the low "price to rent" ratios. This is a really smart finance guy, but with no in-the-trenches real estate experience. I described to him how operational expenses in Fresno will be about the same as in the San Francisco Bay Area, but rents will be about a third that number. Consequently, the operating margin (think EBITDA margin) will be WAY lower in Fresno. Thus, you want to look at the ratio of price / unlevered cash flow. Most real estate investors know this, and thus this is why the price/rent ratio is MUCH lower in Fresno than in say Palo Alto.
The legal requirements for renting can change in ways that drive the prices of renting and buying apart. For example, my wife's folks live in Philly, and they have a house they used to rent out, but has been sitting empty for a couple years. The last tenants decided they didn't want to pay rent anymore, and it took a long time to get them out of the place, with no help from the law. Now, to make up for the uncertainty of actually getting paid rent they could increase the amount they charge, or they could sell the house and try to get the money now. The former increases average rents, the latter lowers average house prices, especially as every landlord in Philly saw similar incentives over the last few years.
I would also add that renting is not an equally valuable substitute for buying for some people, and that is going to tend to drive prices apart for them. I don't know how big a deal that is, and whether it is mostly a rural thing and not an urban thing, etc. I do note that your list does not include maintenance as a cost, however, which is a huge cost of ownership (or a merely large one if you do it yourself). Conversely, trying to get the landlord to do maintenance, much less improvements, you want done might be a huge cost too, but a different sort. Overall I am a bit skeptical of tying rental prices to ownership prices too tightly, but that might just be a function of growing up where renting isn't common.