Houses are assets not goods: What the difference between bulbs and flowers tells us about the housing market


John Lewis and Fergus Cumming

A tulip bulb produces flowers. Those flowers are what people actually enjoy consuming, not the bulb. Whilst that’s blindingly obvious for tulips, the equivalent is also true for housing. The physical dwelling is the asset, but it’s the actual living there (aka “housing services”) that people consume. The two things sound very similar and are often lumped together as “housing”. But in today’s post, we argue they are as different as bulbs and flowers. Sketching out a simplified framework of houses as assets we show how this can radically change how one views the “housing market”. Tomorrow, we use this to develop a toy model and bring it to the data to shed light on house price growth in England and Wales.

You can get flowers by growing them yourself, or buying them off someone else. The fresh flower market works just like any standard textbook demand and supply model. A sudden glut in supply and prices plummet, or if demand spikes, say for Valentine’s Day, prices soar.

Similarly, (in the market sector) people can obtain housing services in one of two ways: purchase the asset and consume the services yourself (be an owner occupier) or just buy the housing services in the spot market (be a renter) from someone who owns the asset (a landlord). Of course, the average rental contract lasts longer than a bunch of flowers, but in both cases you are consuming something without owning the asset that produces it.

In the absence of rent controls, the rental market behaves like a textbook market where the price is determined by demand and supply. If flower prices rise, people might reduce the amount they consume, by buying smaller bunches or purchasing less frequently. If rents go up people might cut back by consuming less space: perhaps two friends would share rather than having their own separate places. Or just by remaining at their parents’ home and consuming some of their housing services.

You can’t buy flowers when they are cheap and store them for months until Valentine’s day. Similarly, you can’t store housing services by, say, renting two flats this year and saving one’s rental services for next year. So the price of rents is determined “on the spot” by the current balance of demand and supply of places to live. Add a load of extra people and/or make them richer and the higher demand pushes up rents. Boost supply and rents fall.

On the other hand standard economic theory views the houses as assets, meaning their price is determined by the (net present discounted) future value of its annual yields. For tulip bulbs, that’s the flowers. For houses that’s the housing services. That applies whether the property is rented (and the rental payments are monetary income for the landlord) or owned outright (and the flows are the market value of the housing services consumed by the owner). This distinction between assets and the services they produce is also in embodied in many countries GDP calculations.

So armed with that basic framework, let’s see what this means for understanding housing markets.

House prices should be sensitive to interest rates…

Economic theory says asset prices should be determined by the value of future income flows. So how much is a bulb that produces £100 of tulips annually worth? If real interest rates on other assets are say 10%, then people would be willing to pay £1,000 for the bulb to get the same return. If they fall to 5%, the value of that same stream of flowers doubles to £2,000. Supply hasn’t changed, the price of tulips hasn’t changed either, but bulb prices have doubled.

As Simon Wren-Lewis, The Economist, and others have noted, that same logic should apply to house prices. In the face of falling global real interest rates, this means rising property prices. That doesn’t tell us whether prices are over or undervalued in any specific place because there are many other factors to be analysed. But the direction is consistent with the rise in house prices seen in many countries over the past two decades.

This channel works purely through equalising rates of return. Abstracting from lending, leverage, collateral effects or other financial frictions, a simple rate-of-return argument alone implies quite strong effects of interest rates on house prices.

Assets can be mispriced and prices can move sharply…

The value of the whole stock of housing is given by the price of the (tiny) fraction that changes hands in any given month (less than 0.5% in 2017). On one level, those prices are set by what the buyers and sellers active in the market are willing to offer and accept. But, more fundamentally, what those customers are willing to pay or accept for a house depends on expectations about the future path of prices and rents, not to mention the (perceived) value of their own or neighbouring houses. As with any asset, if investors aren’t fully rational or liquid, bubbles, panics, fire sales, herding and other forms of mis-pricing can kick in, on both the upside and the downside. As sentiment changes about the value of that future income, that can lead to substantial price swings. Ironically, the Dutch tulip market in the 17th century often cited as a classic example of this.

It’s not always about the supply…

Many commentators attribute rises in house prices to a lack of supply. Indeed, if dwellings get more scarce, an asset price model would predict a rise in the spot price of housing service, i.e. rents.

Other things being equal this would also push up house prices. But house prices can be affected by lots of other things (interest rates, the financial frictions mentioned above), so rental prices should be a “cleaner” measure of relative scarcity: at least in markets where rents are unregulated and free to adjust to market forces. If rental prices don’t move in the same way as house prices then this suggests something else must be at play.

Build more houses and this framework suggests that– other things equal – both rents and prices should fall. But if we observe that prices and rents aren’t moving together then that suggests something other than relative scarcity is driving price movements.

Regional mismatches shouldn’t raise aggregate prices…

Another common refrain is that the houses are “in the wrong place” and so this mismatch of houses and people has the same effect as limited supply in pushing up prices.

But an asset pricing approach contradicts this view: Assume that houses (= tulips) are permanently fixed in one place but people can move. If a shock hits (say jobs move from one region to another + it’s too far to commute), that raises the relative demand for housing services (= flowers) in one region and reduces it by the same amount elsewhere. Rents go up in the region with the inflow and down in the other. But assuming the sensitivity of demand to rents is the same in both places, then they should exactly offset.

The asset price mechanism should move local house prices in lock step with local rents. For prices and rents, the dispersion rises across different places, but the national average of each should be unchanged. This mismatch might bring with it other important issues and challenges (e.g. wealth inequality, labour mobility and others) but it shouldn’t raise aggregate prices.

Higher property taxes needn’t mean higher rents…

Some argue that taxing rental income or buy-to-let transactions will lead some landlords to sell up, reducing the supply of rented housing and thus raising rents. The flaw with such reasoning is that it looks at the rental market in isolation and doesn’t consider the market in the round.

Assume that the total supply of dwellings (= tulip bulbs), doesn’t respond to price movements. The physical stock is fixed in the short run because it’s hard to convert individual flats or houses into business or other uses. Since demand for housing services (= flowers) hasn’t changed, rents shouldn’t change either.

Some landlords will sell up as letting becomes less lucrative. But at the end of each sales chain is either another landlord or someone who was previously renting. If it’s another landlord, aggregate rental supply and demand are both unchanged, and so are rents. If it’s a new owner occupier, the supply of rented property has shrunk by one, but so has the number of renters. The tightness of the rental market and thus rents are unchanged.

The net yield (rents minus tax) from letting property has fallen, so to restore yields to match outside options the asset price must fall. As with any asset, asset prices adjust to the news in yields, not the other way round.

Final thoughts

Like any model, this is a simplification. But viewing houses as assets rather than goods yields some important insights and challenges some received wisdom. What does this approach have to say about developments in UK house prices? That’s the subject of tomorrow’s post…

John Lewis works in the Bank’s Research Hub and Fergus Cumming works in the Bank’s Monetary Policy Outlook Division.

If you want to get in touch, please email us at or leave a comment below.

Comments will only appear once approved by a moderator, and are only published where a full name is supplied. Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.