• Realty Times
• BBC News
• REIClub

I decided to use data from a large, closely watched market known as Manhattan. The Manhattan real estate market is an excellent market to study, and has data readily available. I used the data from Corcoran’s annual report which included data going back to 1974. I also used mortgage rate data going back to 1972, which is available from the Federal Home Mortgage Corporation. CPI Inflation data was also obtained, to adjust for inflation of the dollar, from the US Dept. of Labor Bureau of Labor Statistics. In all, I obtatined 30 years of data for the investigation.

Plotted in black is the average sale price of a Manhattan apartment over the last 30 years. Inflation adjusted data is in red. The 30-year mortgage rate is shown in blue. The market crash of the late 80s is clearly visible. Interestingly enough, according to the inflation adjusted data, the market has yet to recover fully from that decline. The decline of real estate prices in Manhattan after 9/11 is also visible in the data.

It is not really clear in this graph if there is a clear relationship between interest rates and prices for real estate in Manhattan. So, let's see if we can make a graph that will give a clearer view of the data.

In order to get a better idea whether or not we have seen a relationship between mortgage interest rates and real estate prices in Manhattan, let's look at the rates of change per year, and plot them against each other. With this plot, we expect to see a large rise in prices when mortgage interest rates drop, and a drop in prices when rates rise.

There's a lot of information on this graph, so let's take a closer look at a single data point. The point in red is for the change over the years 1999-2000. The 30 Mortagage rate rose from 6.8% to 8.2%, or a 1.4% rise, while the average price of a Manhattan apartment rose from $368,000 to $438,000, or a $70,000 rise (in 1982 USD). This places the red point in the upper right quadrant, with positive interest rate growth, and positive real estate price growth.

So, according to our expectation, we should see all the points in the upper left quadrant, where prices are rising and rates are dropping, or in the lower right, where prices are dropping and rates are increasing. This we don't see.

It's interesting to note that only four of the points are in the lower right quadrant. These four points are years where interest rates were rising, and real estate prices were falling. Compare this to the quadrant above, where rising interst rates were concurrent with rising real estate prices. There are eight points in this quadrant, almost twice as many as in the quadrant below. This data shows that rising interest rates do not always correspond with a drop in real estate prices. Just don't tell the Fed...

Well, this graph is great, but I'm a little disappointed. I was hoping to see a real relationship between prices and rates, but I don't see much here. I was hoping the points would fall on or near a line, and this would give me some insight into how the market reacts to interest rate changes, but it's not here. Why don't we dig a little deeper... To get a little deeper into this problem, I'd like to look at the price sensitivity to interest rates. One could argue that, when rates are low, prices tend to respond to a rate change more than when rates are high. So, let's test this hypothesis with the data.

Below I've plotted the price sensitivity to interest rates against the value of the interest rate.

Here's how I obtained these points. Let's take a closer look at the point in red , which is again for the year 1999-2000. To get the sensitivity of price to interest rate changes, I took the change in the price from year to year and divided by the change in interest rate for the corresponding year. So, for our point, the change in real estate prices was a $70,000 rise, divided by a 1.4% rise in rates. This gives a sensitivity of $48,814 price increase per % interest rate increase. In other words, it's how much prices moved that year per one point interest rate move. That year prices went up, and interest rates went up, so this produces a positive sensitivity to interest rates. At the beginning of 2000, interest rates were at 8.2%, so this is the horizontal value for the red point.

I should also point out that the two very high points and the two very low points correspond to years when interest rates changed very little, so when you divide by a small number, you get a rather large number.

So, what do we see in this graph? We said that prices could be more sensitive to interest rate changes when rates were low. This we don't see. This is what we set out to look at on this page.

In fact, if we believe that rising interest rates correspond with lowering prices, and lowering interest rates with rising prices, all the points in the above graph should be negative! This we also don't see.

Again I am rather disappointed that we don't see some sort of sensitivity to interest rates. One would think that the market becomes more sensitive to interest rate changes at lower rate values. For example, if rates go from 4% to 5%, that's a 25% change, but if rates go from 10% to 11%, that's only a 10% change. But, we don't see this in the data.

So, what can we conclude?

The 30 years of Manhattan data provided an interesting data set to study real estate prices. We looked at the relationship between 30-year mortgage rates and the average price of a Manhattan apartment over this 30 year period in the hopes of showing a relationship between the two quantities. The data did not support the usual claim that rising interest rates causes a decrease in market prices, and dropping interest rates causes an increase in market prices.

Why might this be? Well, the psychology of the markets is not the same as the psychology of people. While the assumption about interest rates affecting real estate prices might make sense for an individual, there are other factors involved. Other factors that influence the real estate market could be the health of the economy, supply and demand issues, local variability, taxes and maintence costs, transaction costs, savings rates, etc.

Please feel free to email me with any questions or comments on the above article, I look forward to hearing from you.