Morally Bankrupt

Per-Capita Adjusted Retail Sales: Not Such a Bright Picture

Earlier today Calculated Risk posted about an improvement in total retail sales. I was interested in how these numbers would look once adjusted to their per-capita numbers. Obviously the population estimates (from the Census Bureau) might have a little bit of an error margin, but I wanted to see how the spending trends of Americans had, or had not, changed. In essence, I wanted to know if people were really spending more, or if the growth in total sales was just a by-product of population growth.

This doesn’t say much about the retail sector, because it isn’t adjusted to account for store openings and closings, in other words, as stores close and sales keep steady, revenue per store is increasing, but it does give us some clues as to whether there is upcoming expansion in the retail sector as the recovery takes hold. From what I gather from these graphs, I’d say an expansion in the number retail outlets is not something we should be expecting any time soon, which is bad news for REITs that operate shopping centers.

There is no adjustments for price levels, because of auto-correlation. Please feel free to submit questions, comments, or sector-specific requests.

I look forward to, in the future, making all my data sources available so they can be reviewed by anyone who may have any doubts about my methods. Just hold on a bit for that.

Mortgage Rates, Purchasing Power and Home Prices

As you may have seen, the Case-Shiller housing index has been showing small signs of improvement. While some may consider this a good sign, I am a bit skeptical. Why? Well, as we saw earlier today, mortgage rates are at 40 year lows. These low rates allow us to cover a much bigger principal with the same monthly payment. In the last 40 years mortgage rates have been as high as 18% and as low as 4.9%. The difference something like this creates on your purchasing power is tremendous.

Interest rates are low right now. Really, really low. I don’t think it can last forever. American consumers are too indebted, we can’t count on foreigners buying our debt with reckless abandon forever and, if a recovery shows up, there will be inflation, requiring a rate increase. I don’t think we are headed back to 18% mortgage rates, but sub-5% can’t last forever either. As interest rates increase, home-buyers will be faced with lower purchasing power, putting pressure on housing prices. Sellers are free to demand whatever price they want, but if they hope to sell they’ll have to meet sellers at a point in which sellers can afford the payment. Barring a significant increase in real per-capita earnings, if interest rates go up, buyers will either need to buy less house; sellers will have to lower prices; banks will need to lend for longer; or buyers will have to devote more of their income to housing. Something must give. I imagine that if that time comes, the rental market will be increasingly attractive as people hold-off on purchasing to wait for better interest rates. Note that this works in the reverse as well, as interest rates drop, people can afford more debt for their monthly payment, giving way to conditions favourable to appreciation.

Below you will find 3 graphs: one showing the effect of mortgage rates on purchasing power across the whole range of rates we have experienced in the last 40 years. The second shows you the same buying power, but ordered chronologically, based on the interest rates at the time. A plot of the CPI-minus-housing adjusted Case-Shiller is provided alongside it. Finally, interest rates and real Case-Shiller are graphed together from 1988 on. Please note that the mortgage rate is a running average of the last 3 months to smooth out noise. I chose 3 months because it is the same time-frame Case-Shiller uses. I expected to find a lagged inverse relation between mortgage rates and Case-Shiller, but, at a glance, there appears to be no strong relationship between them. I guess only regression will tell.

As always your feedback is welcome, either via email, or formspring. (No comments just yet.)