Last week, Brett Arends at the Wall Street Journal made the argument that homes were now cheap.

It's important to note that real-estate prices in many areas are far from a historic bargain. And where there is a glut, prices -- obviously -- are likely to stay lower for longer. It is still a buyer's market. If you are buying, drive a hard bargain.

Prices may still fall further. Yet if you are tempted to keep waiting for homes to get a lot cheaper, there are several reasons to think that might not happen.

First, there are too many other bargain hunters out there.

Second, the falling dollar has made these homes even cheaper to foreign buyers. There are plenty of people in Europe for whom Florida is now a bargain.

Third, interest rates are low right now. I hesitate to give my fellow Americans any extra incentive to borrow yet more money, but you can get a 30-year fixed-rate mortgage under 6%. If the economy recovers that won't last. If you are shopping for a home, it is probably worth seeing if you can lock in one of these rates cheaply.

Finally, in an age of weak currencies and rising inflation, "real" or "hard" assets are in demand. That should include land, bricks and mortar. Sure, real estate isn't as cheap as it has been at other times in the past. But are Florida homes any more expensive these days than steel, or copper, or gold? I'm not so sure.

Calculated Risk takes issue with the idea that homes are now cheap, and posted two graphs as evidence, one on price-to-income and one on price-to-rent.

I cannot argue that relative to those two metrics, homes are cheap.  I do take issue, however, with the following argument from Calculated Risk.

So why does Mr. Arends think "homes are now cheap"? Because he is using a measure of affordability based on mortgage interest rates. This is a mistake without further anlaysis.

Imagine this simplified example with a buyer willing to pay $1000 per month. With a 5% mortgage rate, the buyer could afford a $186,282 30 year fixed rate mortgage (principal and interest). But the buyer expects to sell the home in seven years, and he expects mortgage rates to be 7% then. That means the new buyer - who will also be willing to pay $1000 per month - can only afford a mortgage of $150,308.

So how does the affordability index account for this expected $36,000 loss? It doesn't.

It ends up in this simplified example, the current buyer would be willing to pay about $161,000 today because of the lower interest rate if he was planning on selling in seven years at $150,000 - excluding all expenses, transaction costs, tax savings, discount rates, etc. The actual calculation would be extremely complicated.

I am not sure why a 7% mortgage and a seven year holding period were used in this example, though I assume it is because fixed mortgages have averaged 7% over the long-term and average home ownership duration is seven years. 

Running of the Bulls makes no pretense about knowing where mortgage rates are going in seven years, though if we were to guess, we would probably guess "higher."  However, nobody knows. Mortgage rates could be lower.  So making the argument that home prices are not cheap today because of what mortgage rates will be in 2017 is spinning the roulette wheel.  Calculated Risk may be correct.  Or they may not.  If we are in a deflationary environment - and an implosion in credit is deflationary - mortgage rates may be lower than 7% in seven years.  

But let us accept the argument and assume mortgage rates will be 7% in seven years.  Does this mean that the typical home bought today at $186,000 will be worth $150,000 seven years from now?  No.  In fact, the typical home will be worth more than $186,000.

Why?  

Because incomes will rise (assuming normalization) and the home buyer will have more to spend on housing.

Over time, the nominal rate of income rises at 4% to 5% per year.  If a home buyer can afford to spend $1000 a month on housing payments today, in seven years, the home buyer will be able to spend $1316 at a 4% growth rate of income, and $1407 at a 5% growth rate of income.  A home buyer who has $1316 a month to spend on housing payments will be able to buy a typical house for $197,794 financed by a 7% mortgage.  Someone who can spend $1407 a month can afford to buy a typical house for $211,498.  Thus, assuming normalization of 7% mortgage rates and 4%-5% income growth, buying a house valued at $186,000 today means that same house will be worth between $197,000 to $211,000 seven years from now.  The buyer is still better off buying today even if mortgage rates normalize.

Of course, income growth could be lower than the long-term average over the next seven years, rendering our example moot.  However, this is consistent with the argument that nobody knows what state of the economy will be in 2017, including what the mortgage rate will be.  Maybe mortgage rates and income growth will be higher than the long-term average, maybe they will be lower.  Nobody knows.

Last Spring, we at Running of the Bulls stated here, here, here, here and here that though we did not know if home prices were at the bottom, we thought the bottom was close.  We still do not know if the bottom is in, but it is looking increasingly likely that it is.  The Case-Shiller index, after dropping for 33 straight months, has risen 5.3% since April. 

My current thesis is that housing has bottomed and will stay flat for 2-5 years, with prices choppy and moving in fits and starts.  There are bargains on the low end, particularly foreclosures, the mid-market is stabilizing, and the high end still has some ways to fall.  Geography matters as hard hit places such as Florida and California may have bottomed whereas other places such as Manhattan probably have not.

Of course, I could be completely wrong.  Maybe home prices are headed much lower.  However, I think for this to happen, there has to be a cataclysmic shock or a serious policy error.  A cataclysmic shock might be a sovereign default.  A policy error would be the Fed tightening too early.  Barring either, I believe the bottom for home prices is probably in.