Ford and Vlasenko’s Utter Failure in “The Downside of Monetary Easing”
Forbes’ John Tamny can be quite useful. Today he brings to my attention an astonishingly poor and misleading “research report” by William F. Ford and Polina Vlasenko of the American Institute for Economic Research titled “The Downside of Monetary Easing.”
The paper throws a few numbers into one of Tamny’s favorite arguments, which is that the Fed’s continuing zero interest rate policy (which is sometimes assigned the amusing acronym ZIRP) has reduced the incentives associated with saving, and therefore has caused a dearth of saving which would otherwise be available for businesses, so business investment is down and the economy sputters along.
[While this post isn’t about Tamny’s argument, I must point out that it holds no water. When the recession started, the 10-year Treasury was at about 4% and the personal savings rate was at about 2.5%. Now, the 10-year bond is about half what it was and the savings rate has more than doubled. I don’t know why Tamny continues to argue that there’s a dearth of savings when the contradictory proof is so easy to find.]
Ford and Vlasenko compare interest rates during this recovery to those during previous post-WWII recoveries, calculating that average recovery interest rates have been 4.93% higher than today’s, and assigning shockingly large negative effects on GDP and employment to low interest rates.
Unfortunately for the supply-siders like Tamny who blindly follow without question anything that appears to support their wrong-headed notions, the “analysis” of Ford and Vlasenko is a joke.
Ford and Vlasenko took absurd shortcuts in deriving their figures. They provide the following table in their article:
I wanted to replicate their data, but couldn’t find the 7.07% and 2.14% averages (last column in the table), which I thought strange. I assumed that these numbers would be the average yield of all outstanding Treasury securities or some other logically derived metric, but I was wrong: they’re the simple averages of the yields of the Treasury maturities selected for inclusion in the table above, for the current recovery (2.14%) and for the 9 prior recoveries dating to the 1953-4 recession (7.07%). The absurdity of this “method” is palpable; if they had chosen a different set of maturities in their “study,” they would have gotten different “average” yields.
Then I noticed the enormous disparity in the data: the gap between the 20-year and 30-year average yields for the prior 9 recoveries is 4.13%. This is 23 times the gap between the 20-year and30-year in the June 2010 data point. Nothing can make this error clearer than the yield curves:
Upon digging a little more, it became apparent what Ford and Vlasenko did: they only used the Treasury yield data available in the Federal Reserve’s FRED database, and didn’t bother to fill in the gaps where it was difficult to do so. There are two major gaps in these data; there’s no data for the 6-month series prior to 1982, and there’s no data for the 30-year series for the 2001 recovery or any prior to 1981. This is two-thirds of the recoveries they’re claiming to survey! Here’s the complete set of data, for the Treasury Constant Maturity Rate series that go back past 1982:
All of the average numbers match those of Ford/Vlasenko, except the 6-month, but we’ll come back to that. The most astonishing thing here is that they used the 30-year series with just 3 of 9 possible data points, skewing their “average yield” significantly higher. If we throw out the 30-year and the 6-month series because of insufficient data, the touted interest rate gap drops from 4.93% to 4.39%.
Now back to the 6-month data. Ford and Vlasenko filled out the data set by using the 6-month Secondary Market Rate (TB6MS) without adjustment:
Now all of the numbers match up perfectly.
So Ford and Vlasenko used simple, non-weighted averages to come up with their interest rate differential, and they had flawed data that made it appear higher than reality. What else is wrong?
Given the problems so far, it shouldn’t be a surprise that they made no attempt to incorporate inflation (expected or actual) into their model. Real returns are what matter here. When inflation is factored in, the gap between historical rates and the June 2010 rates shrivels, in my first run to less than 1% (using 3-year inflation data for maturities of 3 years or less, 5-year data for longer maturities).
Astonishing though it may sound, the errors in method and in choosing nominal data may not be the biggest, most obvious errors in the paper. Brace yourself.
In calculating the net effect of higher interest rates in the economy, Ford and Vlasenko only look at the impacts on the capital owners, ignoring completely the negative effects higher interest rates would have on borrowers. Ford and Vlasenko calculated that interest rates 4.93% higher on $14.35 trillion of rate-sensitive investments would have resulted in $707 billion more income and $371 billion more consumption. Period.
They don’t consider for a moment that that extra $707 billion has to come from somewhere. Every dollar in the pocket of a lender is a dollar not in the pocket of a borrower. The consumption, investment, and taxes assigned to lenders in Ford and Vlasenko’s model would replace consumption and investment by borrowers. There’s no effort to capture these negative impacts.
In reality, that $707 billion would come straight out of the domestic economy; businesses and consumers paying more interest on a host of loans funding everything from cars and boats to home purchases to seasonal retail purchasing to industrial expansion. Make those things cost more through higher interest rates, and there would be less borrowing. Increase interest payments, and businesses and consumers would have had less money to spend on consumption and other investment.
Ford and Vlasenko ignore these impacts. In their world, higher interest rates generate only positives effects. Which is true – if you’re only concerned (like John Tamny) with suppliers of capital. Ford and Vlasenko don’t recognize any negative impact on borrowers in their high interest rate model – they quite literally ignore half the story.
The real question is about the impacts on both sides; too bad Ford and Vlasenko didn’t try to answer it.
On second thought, maybe it’s better that they didn’t try that.