Linden Labs, the operator of Second Life, has decided to increase the amount of virtual real estate in the game, causing a sudden drop in virtual land prices. This is interesting because in real life when cities expand the new land is at the edges, which is typically far from the more valuable centre, and probably has less of an impact on land prices overall.

In a virtual world, however, I guess that location is much less important. I don’t know exactly how people move around in Second Life, but in principle transportation in a virtual world from one location to another can be instantaneous and costless. In that case the distinction between ‘centre’ and ‘edge’ is blurry, and increasing the supply of land anywhere could affect prices everywhere, as appears to have happened in Second Life.

I guess there is a lesson here for real-world real estate investors. Improvements in transportation technology make cities more ‘connected’ and, everything else equal, should increase the correlation of real estate prices between different locations. My hypothesis is that with good transportation, extra land at the edges can have a significant effect on prices at the centre. It would be interesting to try to test this with some data.

(HT: UsableMarkets)

I suspect not many of my blog readers are from New Zealand. Anyway, I was playing around with some New Zealand real estate market data and I made this interesting graph. This graph shows the ratio of the real monthly average selling price in Auckland (the biggest city) versus the real monthly average price in all other areas. It’s interesting because it shows that Auckland prices are currently about as cheap relative to the rest of the country as they ever have been over the past 15 years.

Of course, if we calculate the relative price across time, real prices are now at their all-time high. So relative to the past, housing has never been more expensive. But relative to other areas, Auckland looks surprisingly cheap at the moment compared to the recent past. People will no doubt attribute this to the increase in the number of smaller lower-cost apartments in Auckland. However, you can’t measure quality based on floor area alone. They may be smaller, but an apartment in the city is … in the city, whereas a larger house in the countryside is not.

Coming from New Zealand, I have a strong interest in the state of the real estate market there. Like many other countries recently, real house prices have increased dramatically over the past five years or so. Here’s a graph of the monthly median real selling price across all New Zealand and in Auckland (the biggest city) since 1992. I inflated past values to current dollars so these are real values measured in today’s money (all data are from the REINZ website):

Prices have increased by about 2.3 times since 1992, but most of the growth occurred since 2002. However, in the past few months there have been some signs of weakness. As you can see from the graph above prices have decreased slightly since their peak this year. In the latest data for October 2007, median prices are down 2.5% in Auckland and 1% nationwide compared to their peak. However, many small downward fluctuations have occurred in the past few years without disturbing the upwards trend. It’ll take a few more months to see if the trend has changed or not.

This chart shows a 12 month moving average of the monthly change in real prices. Roughly three periods can be distinguished, as in the median price graph above. Before 1998, prices grew at about 0.5% per month. From 1998 to 2002, real prices basically didn’t change. Then after 2002 prices have been growing at a bit less than 1% per month. Again the latest data indicates that growth rates are down, but it’s still too soon to tell if this will be long-lasting.

Next up is a graph of the number of sales that were completed each month. There’s a lot of seasonal fluctuation, but the sales figures for the last two months are quite low compared to similar months in previous years. There’s also been a bit increase in turnover of houses in the past few years, which has coincided with the period of rapidly rising prices.

A lot of the price increases occurred during times of relatively low interest rates. Here’s a plot of floating mortgage rates together with the price graph from above (interest rates from the RBNZ website):

Interest rates have been rising recently, as the Reserve Bank has been battling inflation pressures in part due to people feeling wealthier from the higher value of their houses, and consequently taking out further loans and spending. The big question is how much effect this is going to have on the housing market. Here’s a scatter plot of the monthly percentage change in the median real house price for all New Zealand versus the floating mortgage rate in the same month:

There’s a very slight negative relationship, but it’s not statistically significant. It doesn’t seem like higher interest rates are associated with lower price increases. Or there’s some more complicated relationship that doesn’t show up in a simple correlation.  There does seem to be a stronger relationship between number of sales and interest rates though, and this is statistically significant:

It’s just a hypothesis but it seems like higher interest rates lead people to stay in their homes and not sell them, rather than selling them at lower prices. New Zealand also doesn’t seem to have the low-quality credit problems that are plaguing the housing market in the US. So it seems more likely that any ‘correction’ in prices will take the form of constant nominal prices, while real prices fall.