Click Here for PDF Newsletter

What a difference three years makes. We have moved from a borrow-and-shopuntil-you-drop economy to a period where large banks are reporting multi-billion dollar losses in the credit
markets.

Households have overestimated their purchasing power by using adjustable rate mortgages, having fun and games with the discount rate, with results similar to the S&L’s of the 1980’s. This article examines the underpinnings of the run-up in real estate prices and provides some historical perspective on where prices might be going.

Evidence of a change in the consumption function for housing

Housing prices are all about the purchasing power of households. The Joint
Center for Housing Studies at Harvard analyzes the relationship between
housing prices and income and maintains a time series of data going back to
1981. From 1981 through 2000, the ratio of median home prices to median
household income nationally remained virtually constant at about 3.0. There
was almost no variation in this figure, as the range during the 20-year period
was from 2.9 to 3.0. Certain metropolitan statistical areas (MSAs) were
higher, most notably in California, but as a whole the nation appeared willing
to spend three times household income on housing.

Things started changing in 2001, and by the 2004-2006 period the priceto-income
ratio had skyrocketed to an average of 4.3. In hot markets in California (San Diego, San Francisco-Oakland, and Los Angeles) the ratio reached about 9.0. San Jose was not
far behind at 8.2. These are staggering increases in commitments to housing.
The interest rate advantage of adjustable rate mortgages widens in 2004-2

The interest rate advantage of adjustable rate mortgages widens in 2004-2006

ARMs cost 1.8% less than fixed rate mortgages during the 24 year period from 1984 through 2007. This is about the same as the figure from 2002 through 2005. The startling difference is the percentage of the 30-year rate that this represents. In the four years leading to this period, the ARM interest rate savings was about 17%.From 2002 through 2005 this figure exploded to about 30%. For households with affordability issues, this is a huge difference in interest costs.

Screen Shot 2015-12-29 at 5.59.20 AM

To understand this you need to understand valuation of a household’s purchasing power. This is done in roughly the same manner as financial instruments such as bonds are valued. Financial types are accustomed to making discounted cash flow calculations in estimating the value of assets. The simplest calculations are for perpetual bonds with fixed coupons. The discount rate that reflects term and default risk is divided into the payment amount to arrive at the value of the bond.

$60 coupon / 6.0% = $1,000 market value

Companies with higher default risk would pay more, and the discount rate would be higher.

$90 coupon / 9% = $1,000 market value

The interest rates on short-term investments are usually lower because the lender does not have to be compensated for term risk. In the case of ARMs, the spread between long and short mortgage rates has been 1.8% over 24 years. Imagine valuing the bond with a $60 coupon with an interest rate that is 1.8% less, 4.2% instead of 6%.

If an investor were to use a short-term rate to value a long-term bond, the market value of would be overvalued by 43%!!!!!!!! Think about other errors that could occur of other investments were valued using the wrong discount rate.

  • Payments on a 30-year government bond were capitalized at the interest rate associated with the 3-month T-bill
  • The dividends from a real estate investment trusts were capitalized at money market rates in estimating the value of the shares

The key point here is that the long-term borrowing rate is the market’s estimate of a household’s cost of money over the life of an investment in real estate. This is the interest rate that should be used to compute the cost of housing and the estimated monthly payment.

The capitalized value of a household’s purchasing power is done in the same manner. The amount of cash flow that is available to service debt determines the household’s credit capacity. The household cost of mortgage debt determines how much debt can be serviced.

Screen Shot 2015-12-29 at 6.07.55 AMYou can see where this is leading. With 30% lower interest rates charged on ARM financing from 2003 through 2005, those households who could tolerate risk had 43% more purchasing power per dollar of cash flow than those who chose to use fixed rate financing.

This is an analysis that is done at the micro level, and it is important to see if the idea is confirmed with macroeconomic statistics. How did this increase in purchasing power stack up against the numbers reported by the JCHS?

Screen Shot 2015-12-29 at 6.08.27 AM

Well, it looks like households took the new paradigm of ARM borrowing to heart and bid up the price of real estate to the extent of their collective credit capacities. Assume that by 2006 the effects of innovative financing had percolated through the pricing structure of housing. If purchasing power had been overestimated by 43%, this would have raised the ratio up to 4.3. The actual figures from the JCHS survey are 4.0 for 2004, 4.2 for 2005, and 4.4 for 2006.

Households emulate the savings and loans of the 1980’s

Before the inflation of the 1970’s and 1980’s hit, home lending was dominated by the savings and loan industry. These were referred to as the 3-6-3 guys. Borrow at 3%, lend at 6%, and be on the golf course by 3 PM. It was a great business as long at the cost of capital was stable. As the inflation of the 1970’s took hold, the cost of short-term borrowing rose to the point where the banks were actually losing money on loans that weren’t in default.

The savings and loan industry had made the fundamental error of mismatching the duration of their assets and liabilities. The return on their assets (home loans) was fixed at the time the loans were made. However, the cost of money to fund these loans was variable, having been obtained through short-term time deposits. Profits were great when the cost of short-term borrowing was low, but nasty things happened when inflationary pressures increased interest rates:

  • The market value of assets in their portfolios declined.
  • Before deregulation, time deposits money would leave the banks in favor of higher returns elsewhere. (The term for this is financial disintermediation.)
  • After deregulation, the banks had a cost of money that was higher then their portfolios were returning.
  • Most individuals’ memories of it are from banks making reckless investments like forays into commercial real estate lending. However, the real underlying problem and the one that caused the most damage was the mismatching of assets and liabilities.

If an investment is financed with a liability of a similar duration, an interest rate spread is locked in. You can make a 30-year home loan and issue 30 year bonds to fund it. Ignoring the issue of mortgage prepayments, the key risk would be the default rate on the mortgage portfolio. If interest rates went up, down, or sideways, the bank wouldn’t care. The profit spread is locked in.

S&L’s were hooked on access to cheap time deposits. The profit spreads were good, and risk management took a back seat to maximizing the spread between the cost of funds and the return on the mortgage portfolios. Federal deposit insurance helped in gathering funds by making S&L time deposits a riskless investment. Borrowing short and buying long positions maximizes financial leverage. It also maximizes risk.

Investment in residential housing financed by fixed rate debt was a benign form of the carry trade. In exchange for taking on a stream of fixed payments, homeowners got the following advantages:

  • They locked in their cost of land and building.
  • They got the right to inflationary increases in the value of the property.
  • In most states they got the right to prepay the mortgage without penalty, allowing them to accelerate the retirement of the debt if less expensive funds were available from other sources.
  • They could pay off the mortgage in cheaper dollars whose purchasing power was diminished by the effects of inflation.

In an inflationary environment, this was an investment that was hard to beat. Americans have never been guilty of moderation and the idea of leveraged investment went to excess. Americans engaged in aggressive borrowing, including:

  • Lower down payments (20% was the norm in the 1980’s).
  • Buying the most house instead of economizing on living expenses. (Moderately sized housing has been torn down in favor of McMansions.)
  • Use of interest only loans (“rent-a-house”)
  • Use of adjustable rate mortgages (ARMs)

By using ARM interest rates to calculate their purchasing power for housing, households grossly overestimated their ability to afford housing. When these buyers were computing monthly payments, they were using interest rates associated with adjustable rate loans. What this does is totally underestimate the opportunity cost of housing. More critically, it overstates the real estate purchasing power of households.

It gets worse. Things rarely happen in a vacuum, and housing prices are no exception. The availability of ARMs unleashed this illusion of purchasing power on the national real estate market, and the price of housing was bid up way beyond what would have occurred if households were using fixed rate mortgages. A good portion of the interest rate savings evaporated into higher purchase prices. So that the end of the day we see households who bought at the peak being inadequately compensated for the large financial risk that was
assumed.

Subprime borrowers lead the real estate market down

Every good party has a hangover, and this one is no exception. In 2006 and 2007 the ARM advantage decreased from 43.3% to 15.1%. Again, using the Freddie Mac interest rates:

This means that 2/3 of the expansion of purchasing power that was unleashed on the real estate market by ARMs has been sucked out of the market by the narrowing of the spread between long and short borrowing rates. In 2006 and 2007, 1-year ARM borrowers faced a 32.2% increase in interest costs in comparison to the 2003-2005 rates.

Imagine households stretching to buy overvalued real estate in 2003-2005 with ARM financing. Then imagine their biggest expense going up by 32.2%. Think this is a budget buster? These subprime borrowers are just like the S&L’s who entered into leveraged transactions without controlling their most significant cost.

The nail in the coffin for this period of wild borrowing is the coming defaults in mortgage-backed securities. The ability of Wall Street to sell securities backed by pools of mortgages is based on the lenders’ belief that:

  • Borrowers will have adequate incomes to service the debt, and
  • Adequate due diligence has been done to assure there a low enough loan to value ratio so the loan will be paid off via sale of the home, whether or not by foreclosure.
  • Lower loan to income ratios
  • Lower loan to value ratios
  • Higher rates for jumbo and second home mortgages
  • Higher rates for loans backed by a single wage source
  • Higher rates for adjustable rate mortgages
  • All these factors add up to larger down payments. If homes are not appreciating, this money can only come from savings on current income. For the past several years the US savings rate on current income has been close to zero. Higher down payments and a reduced interest rate advantage of ARMs produce reduced purchasing power for home buyers. This will have a negative effect on prices.