### Monday, January 18, 2010

## DeLong on Interest Rates and the Housing Bubble

In a WSJ survey of economists on the housing bubble and Greenspan's possible role, Brad DeLong wrote:

After the dot-com crash, Greenspan steadily brought down the target for the “federal funds rate,” which is the interest rate banks charge each other for overnight loans of reserves. At its lowest point—from June 2003 through June 2004—the federal funds target was 1 percent, the lowest it had been in about 40 years. So let’s suppose that the rate should have instead been 3 percent. How much of a difference would this have on a very durable asset?

Well, if we simplistically plugged this number in to a standard formula for calculating the present value of an income stream, it would have an enormous impact. For example, if an investor were evaluating an asset that would yield $1,000 in income perpetually, then at a discount rate of 1% it would be valued at $100,000. Yet at the “modestly” higher discount rate of 3%, the asset’s price would collapse to $33,333. In this simplistic example, the artificially reduction of two percentage points would lead to a tripling in the asset value.

Now let’s be fair to DeLong, the guy’s got a Ph.D. in economics. He is aware of the simple formula relating asset prices to income streams and discount rates. Presumably what he had in mind was that the Fed pushed down very short term rates (namely overnight rates), and did this (at least according to the Taylor Rule) for three years. So DeLong could plausibly argue that our calculation above was grossly misleading, since no rational investor would have expected Greenspan (and his successors) to suppress rates indefinitely below their “proper” level.

Fair enough, let’s take a much more realistic approach. Let’s look at what happened to 30-year conventional mortgage rates during Greenspan’s easy-money period. In order to quell the dot-com bubble, Greenspan’s Fed had raised the target up to 6.5% by May 2000. But then starting in January 2001, the maestro steadily cut the target down to a low point of 1% by June 2003, where he held it for a full year.

Mortgage rates fell along with the Fed’s slashing of the overnight target. Now we can’t prove that the one event caused the other, but then again the whole point of the Fed cutting its short-term target is to bring down other interest rates as well. Today, if Bernanke suddenly decided to jack up the target to 10%, I’m sure DeLong would go nuts and warn that this would destroy the economy, including the housing market. Look at this link to see how low mortgage rates were by the time of the housing boom.

So what we’ll do is pick mortgage rates from the middle of the high target rate (“tight money”) period, and then from the middle of the low target rate (“easy money”) period. In August 2000, the average 30-year mortgage rate was 8.03%. In contrast, in December 2003, mortgage rates had fallen to 5.88%. Although it’s not a perfect test, I think plugging in these numbers gives a much better idea of the potency of Greenspan’s policy errors, than the calculation DeLong submitted to the Wall Street Journal survey.

Using a simple amortization calculator, we can see that the drop in mortgage rates makes a huge difference. At a rate of 8.03%, someone who can afford to make monthly payments of $1,000 will be able to buy a house worth a little less than $136,000. But if the rates fall to 5.88%, the person with the same willingness to make monthly mortgage payments can now afford a house that costs almost $169,000. So we see that the apparently modest drop of 2.15 percentage points in 30-year mortgage rates would cause perfectly rational homebuyers to spend 24 percent more on a house.

My point with the above calculation isn’t to make the definitive indictment against Greenspan. All I’m trying to show is that DeLong’s quick arithmetic is a complete non sequitur. The change in mortgage rates during Greenspan’s easy-money period certainly would have the power to kick start housing prices, at which point speculative self-fulfilling (yet unsustainable) prophecies could take over.

Finally, what of DeLong’s point that there wasn’t a depression following the dot-com crash that we now have? Well, as I have explained more fully in my latest book, the reason we are currently experiencing a Great Depression II is that the government has been implementing the New Deal II. It’s also true that there were huge financial crashes--and even worse price deflation--prior to the 1930s, yet with no decade-long depression. It was only when the Fed and government began doing the very things (though halfheartedly in his book) that DeLong currently suggests, that the U.S. economy was saddled with an unprecedented slump.

Although he presumably wasn’t given much space to work with, DeLong doesn’t justify the relevance of his calculation. After all, at face value a drop of two percentage points in the interest rate can have a huge effect on the price of a durable asset.BRAD DELONG, BERKELEY PROFESSOR: “If you believe that the Fed kept the fed funds rate 2% below its proper Taylor-rule value for 3 years, that has a 6% impact on the price of a long-duration asset like housing. Even with a lot of positive-feedback trading built in, that’s not enough to create a big bubble. And it wasn’t the bubble’s collapse that caused the current depression–2000-2001 saw a bigger bubble collapse, and no depression.”

After the dot-com crash, Greenspan steadily brought down the target for the “federal funds rate,” which is the interest rate banks charge each other for overnight loans of reserves. At its lowest point—from June 2003 through June 2004—the federal funds target was 1 percent, the lowest it had been in about 40 years. So let’s suppose that the rate should have instead been 3 percent. How much of a difference would this have on a very durable asset?

Well, if we simplistically plugged this number in to a standard formula for calculating the present value of an income stream, it would have an enormous impact. For example, if an investor were evaluating an asset that would yield $1,000 in income perpetually, then at a discount rate of 1% it would be valued at $100,000. Yet at the “modestly” higher discount rate of 3%, the asset’s price would collapse to $33,333. In this simplistic example, the artificially reduction of two percentage points would lead to a tripling in the asset value.

Now let’s be fair to DeLong, the guy’s got a Ph.D. in economics. He is aware of the simple formula relating asset prices to income streams and discount rates. Presumably what he had in mind was that the Fed pushed down very short term rates (namely overnight rates), and did this (at least according to the Taylor Rule) for three years. So DeLong could plausibly argue that our calculation above was grossly misleading, since no rational investor would have expected Greenspan (and his successors) to suppress rates indefinitely below their “proper” level.

Fair enough, let’s take a much more realistic approach. Let’s look at what happened to 30-year conventional mortgage rates during Greenspan’s easy-money period. In order to quell the dot-com bubble, Greenspan’s Fed had raised the target up to 6.5% by May 2000. But then starting in January 2001, the maestro steadily cut the target down to a low point of 1% by June 2003, where he held it for a full year.

Mortgage rates fell along with the Fed’s slashing of the overnight target. Now we can’t prove that the one event caused the other, but then again the whole point of the Fed cutting its short-term target is to bring down other interest rates as well. Today, if Bernanke suddenly decided to jack up the target to 10%, I’m sure DeLong would go nuts and warn that this would destroy the economy, including the housing market. Look at this link to see how low mortgage rates were by the time of the housing boom.

So what we’ll do is pick mortgage rates from the middle of the high target rate (“tight money”) period, and then from the middle of the low target rate (“easy money”) period. In August 2000, the average 30-year mortgage rate was 8.03%. In contrast, in December 2003, mortgage rates had fallen to 5.88%. Although it’s not a perfect test, I think plugging in these numbers gives a much better idea of the potency of Greenspan’s policy errors, than the calculation DeLong submitted to the Wall Street Journal survey.

Using a simple amortization calculator, we can see that the drop in mortgage rates makes a huge difference. At a rate of 8.03%, someone who can afford to make monthly payments of $1,000 will be able to buy a house worth a little less than $136,000. But if the rates fall to 5.88%, the person with the same willingness to make monthly mortgage payments can now afford a house that costs almost $169,000. So we see that the apparently modest drop of 2.15 percentage points in 30-year mortgage rates would cause perfectly rational homebuyers to spend 24 percent more on a house.

My point with the above calculation isn’t to make the definitive indictment against Greenspan. All I’m trying to show is that DeLong’s quick arithmetic is a complete non sequitur. The change in mortgage rates during Greenspan’s easy-money period certainly would have the power to kick start housing prices, at which point speculative self-fulfilling (yet unsustainable) prophecies could take over.

Finally, what of DeLong’s point that there wasn’t a depression following the dot-com crash that we now have? Well, as I have explained more fully in my latest book, the reason we are currently experiencing a Great Depression II is that the government has been implementing the New Deal II. It’s also true that there were huge financial crashes--and even worse price deflation--prior to the 1930s, yet with no decade-long depression. It was only when the Fed and government began doing the very things (though halfheartedly in his book) that DeLong currently suggests, that the U.S. economy was saddled with an unprecedented slump.

Comments:

Hayek takes on the net present value argument in PII and The Ricardo Effect. That's where everyone heads when they want to discuss the effects of changes in interest rates. But as Hayek points out, the NPV calculation is a tiny part of the story. The change in interest rates also changes the relative prices of consumer goods vs capital goods. As a result, the change in profits is amplified. Most businessmen make investment decisions based on relative profits, not NPV. And the effects of interest rate changes on relative profits is far greater than that on NPV.

I really like Hayek's use of turnover ratios, a pure accounting tool, to prove his point in Ricardo Effect.

Post a Comment
I really like Hayek's use of turnover ratios, a pure accounting tool, to prove his point in Ricardo Effect.

Subscribe to Post Comments [Atom]

<< Home

Subscribe to Posts [Atom]