By Ron Rimkus, CFA
In the meantime, the Wall Street Journal this week published a piece suggesting that banks (through loans from their central banks) are providing more support for their respective sovereign bond markets, making sovereign defaults less likely to occur. In contrast, Bloomberg published an article on the same topic suggesting that the banks supporting the bond markets are making them more vulnerable to a collapse. According to the Bloomberg article, as more of the bonds of a particular country are owned by local institutions, there is less willingness to support the sovereign debts of other countries.
While this is an important discussion, it is secondary to the larger question: Why is more debt the answer to these countries’ problems? Either the world can survive with more debt, or it can’t. Engineering low interest rates for an extended period of time means that the world will take on more debt than it otherwise would have — thus causing overconsumption of debt. There’s no telling exactly where this excess debt will wind up and just what consequences it will have on the economy. What is clear is that we will wind up with gross misallocations of capital and massive debt burdens.
Speaking of cause and effect . . . a recent Wall Street Journal article about a significant new trend in pharmaceutical research got me thinking about the perils of observational studies in modern financial research. According to the article, pharma researchers are shifting from randomly controlled experimental studies to observational ones. So for example, in a controlled study, two or more groups are chosen: one is given a placebo and the other a real drug. In this way, comparisons of the two groups eliminate whatever bias might be present in the selection of patients for the study. (For example, is the sample more healthy than the overall population? Is it more sickly or more disposed to disease A vs. disease B?)
Using this approach, any biases that are present should be equivalent among the two cohorts in the study across a large enough sample size. Contrast that with observational studies which simply look through large volumes of data to find a correlation. As the Journal article points out, observational studies cannot distinguish spurious correlation from cause and effect. To be sure, any study can be executed well or poorly (including controlled studies), but observational studies add additional risks to drawing conclusions.
So what has this got to do with modern finance? Plenty it turns out. In fact, the parallels are breathtaking. The modern U.S. Federal Reserve and its chairman, Ben Bernanke, in particular, are proponents of inductive reasoning. Consequently, they perform large-scale macroeconomic analysis of time-series data in order to draw conclusions about how to manage interest rates and economic growth. However, their approach suffers from the same exact pitfalls as the observational studies in pharma research. Namely, they cannot draw conclusions about cause and effect. As any good investor knows, first you develop your thesis, then you look for evidence. Starting with data and then creating a theory leads to greater risk or misunderstanding of how the world works.
Today, central banks the world over appear to be following suit. Historically, more debt led to more growth because most households, governments, and institutions could afford to lever up from low debt levels to high debt levels. Maybe this is why more debt seems to be the solution to every problem. But debt has a limit and paying for it with inflation means it is a regressive tax on our poorest citizens.
Just last week, Bernanke was asked during a press conference if his behavior as Fed governor differs from the advice he dispensed to Japan in the mid-1990s, when he was a professor. Bernanke responded with two points: first, that a central bank should work to eliminate deflation; and second, that a central bank can and should use nontraditional policies when interest rates fall near zero.
While I take the Fed chairman at his word as to whether or not his advice for Japan is consistent with his current views, I take issue with his advice. As a result of using inductive reasoning, Bernanke and his Fed colleagues cannot distinguish between good inflation and bad deflation. In fact, the Fed Chairman has arbitrarily decided that all deflation is bad. I can see why he says that. However, consider the case of semiconductors: Over time, the cost per bit has declined markedly as the productivity of individual semiconductors has soared. Semiconductor manufacturers made more money as demand grew in response to the wave of productivity improvements. Demand proved to be highly elastic with industry revenue compounding growth at about 18% per year from 1958–2000.
Are there not a number of innovations that have sparked a wave of productivity improvements in recent years? Does the Internet itself not fit this bill? Isn’t it possible, just possible, that the productivity improvements gleaned from information technology could have unleashed a wave of good deflation on the market as whole — and that the Fed’s actions to “maintain price stability” are the causal variable in stimulating a credit bubble? We should not judge the economy just on the basis of what has been, but rather on the basis of what could have been, given its potential.
We have plenty of evidence of what Bernanke meant by his second point about pulling out nontraditional tools in a zero-interest rate environment. For instance: Operation Twist. Selling short-dated bonds and buying long-dated bonds works to reduce interest rates. In effect, it serves to pull future demand forward as the private sector looks at accelerating its debt consumption to take advantage of low rates.
A study performed many years ago showed that the housing market is largely inelastic when rates are between 5% and 8%, but highly elastic outside this range. So, with rates near zero, the Fed has enabled the economy to pull in possibly massive amounts of future demand, yet the economy remains anemic. What does that tell us? What happens to demand when these programs end?