Housing: When is it best to invest?

How do high transaction costs, as well as real-estate booms and busts, affect optimal portfolio choices for housing investments?

Buying a house is a big financial decision. Not only is a home a significant investment in itself, but transaction costs can run high.

The decision to invest is affected by current housing market conditions — whether prices are heading up, down or nowhere in a hurry.

Given these considerations, IESE's Carles Vergara, Stefano Corradin, of the European Central Bank, and José L. Fillat, of the Federal Reserve Bank of Boston, developed a model of optimal portfolio choice with housing and nonhousing investments. Their results are published in The Review of Financial Studies.

They find that housing adjustments are infrequent and characterized by both the household's total wealth-to-housing ratio and the expected growth in house prices. This means that, most of the time, buying or selling a house is less attractive than not moving at all.

They also find three main portfolio allocation results:

1. The housing portfolio share immediately after moving to a more valuable house is higher during periods of high expected growth in house prices.
2. The share of wealth invested in risky assets is lower during periods of high expected growth in house prices.
3. The decrease in risky portfolio holdings for households moving to a more valuable house is greater in high-growth periods.

Price predictability & transaction costs

Where are house prices heading? The answer varies with time, but the authors show that it is predictable.

Armed with U.S. housing price data from 1925 to 2012, the authors find that U.S. housing returns are "well captured" if divided into three regimes: low, medium and high growth. Average growth rates, adjusted for inflation, are as follows: low regime, -16 percent; medium regime, 0 percent; and high regime, +9 percent.

According to this framework, the United States is in a regime of medium growth most of the time. Periods of low or high growth are rare and fleeting, they say.

In fact, there have been only two periods since 1925 in which the probability of being in a high-growth regime in the United States tipped over 50 percent: during the post-World War II period, and during the real-estate boom prior to the 2008 global financial crisis.

The authors also measured the future growth of house-price-to-rent ratios to fortify their stance on price predictability.

Next, the authors looked at transaction costs, which could range anywhere from 5 percent to 25 percent of the underlying house price. These high transaction costs make housing consumption "lumpy", that is, families do not move to a different house very often.

Moreover, high transactions costs make potential homebuyers more risk averse, which also affects their optimal portfolio choices of stocks and bonds.

Setting aside moves forced by household changes such as divorce, death, birth and job loss, the authors find that housing moves occur when a consumer's wealth-to-housing ratio reaches an optimal upper or lower boundary, which is dependent on both the market conditions and transaction costs.

"The effects of transaction costs and house price predictability are key elements of both housing and nonhousing portfolio allocation decisions," they state.

In short, when transaction costs are low and house prices are expected to soar, real estate agents are kept busy. But when the reverse is true, these brick-and-mortar investments tend to lose their appeal.

As such, the housing market cycle affects not only housing investments, but also investments in the stock and bond markets.