You are reading Economic Forces, a free weekly newsletter on economics, especially price theory, without the politics. Economic Forces arrives weekly in the inboxes of over 16,000 subscribers. You can support our newsletter by sharing this free post or becoming a paid subscriber:
If you have ever formally studied macroeconomics, you know that macroeconomists often seem to be obsessed with making sure that their macroeconomic models have microeconomic foundations. If you read this newsletter, you know that I have some broad sympathy for this view. I often write posts about macroeconomic topics, such as inflation, using the tools of price theory. However, if you are also familiar with how these microeconomic foundations have been implemented, you will know that I am much less sympathetic to this endeavor in practice. This is not universally true. There is much to like, but also much to dislike.
Not surprisingly, I think that there is a simple, key distinction that determines what I am sympathetic towards and what I am not. When this is done well, it is done with (surprise!) careful price theoretic foundations. It starts from first principles. When this is done poorly, microeconomic foundations just means optimization with a bunch of ad hoc assumptions.
Careful price theoretic foundations can come in a variety of forms. Although price theory is often framed as an approach that uses simple models to explain the world, this doesn’t necessarily mean that one cannot use structural, general equilibrium-type economic models. What is important isn’t so much the scale of the model, but whether or not one carefully models the incentives and costs faced by the agents of the model.
Nevertheless, a macroeconomics founded on price theoretic foundations also opens the door to analyses neglected by certain macroeconomists. There is a subset working on structural models to nowhere. In contrast, I think that a proper understanding of financial markets can often be used as a tool to assess macroeconomic predictions.
Thus, what I would like to do in this post is explain why I think a proper understanding of finance is crucial to a price theoretic foundation of macroeconomics. After doing so, I will provide an example to illustrate how basic price theoretic predictions and tools can be used to answer questions by observing asset prices and through reduced form empirical work.
The Importance of Financial Markets
Those who study financial markets have a useful conception of equilibrium: the absence of risk-free arbitrage. The basic idea is that if I can buy something in one market for a particular price and sell it in another market for a higher price, then I should not expect that price difference to persist — unless there is some cost or risk associated with arbitrage. This difference should dissipate as more people become aware of the profit opportunity. As more people buy the asset in the low-price market and sell the asset in the high-price market, the prices will converge and the profit opportunity will disappear.
This basic principle should not only apply for the same asset sold in different markets, but also for different assets with the same characteristics. For example, if there are two companies that have similar balance sheets, cash flows, and risk factors, the companies’ market capitalizations should reflect that similarity.
In theory, this seems quite simple. However, in practice it can be much more difficult. In fact, critics of this approach to finance will often claim to find anomalies, or what they perceive as persistent deviations from this no-arbitrage condition.
Whereas critics see deviations as violations of the basic principle, a price theorist sees an opportunity for explanation. What is going on to create such anomalies? What sort of costs are missing?
There are other important implications of this no-arbitrage approach. For example, maybe there isn’t a close comparison in the market for a particular asset. Nonetheless, suppose that I could create a portfolio of assets that have the same characteristics of another asset. If that is the case, the no-arbitrage condition says that the portfolio and the comparable asset should have the same expected rate of return. Why? Well, consider if this wasn’t true. Suppose that the portfolio has a higher expected rate of return. One could short the asset and use the proceeds to buy the portfolio of other assets. At least in expectation, this should earn a profit.
If this was the case, we would expect that other market participants would notice this profit opportunity as well. More people would short the asset and buy the portfolio. In the process, this would tend to increase the expected rate of return on the asset and reduce the expected rate of return on the portfolio until they had the same expected return.
A final implication of this approach is as follows. Suppose that we classify risk into two different categories. There is idiosyncratic risk, which is a risk unique to that particular asset (e.g., iPhone sales matter for Apple’s stock), and there is aggregate risk, which affects all assets (note that this does not imply it affects all assets equally). Again, portfolio construction can matter. If I create a big enough portfolio, then I can effectively eliminate all the idiosyncratic risk. Since these risks are uncorrelated, a big portfolio makes the risk of any one component pretty negligible. However, one cannot eliminate aggregate risk. This implies that asset holders should only be compensated with higher returns for these aggregate risk factors.
These straightforward implications of the no-arbitrage condition are important beyond the study of financial markets and asset pricing in and of itself. These implications can be used to test macroeconomic theories. All macroeconomic theories are going to have implications for what these no-arbitrage conditions will look like. It is therefore possible to see if the data is consistent with these conditions and whether movements in asset prices following some unexpected shock are consistent with the predictions of the theory.
Let’s consider an example.
The Market for Safe Assets
Following the financial crisis that began in 2008, there was considerable debate about what caused it and why the economic recovery in its aftermath was so slow. On one side of this debate, there were people who argued that the Federal Reserve held interest rates too low for too long in the early part of the decade. These low interest rates contributed to reckless lending practices and malinvestment. Others contended that the degree to which interest rates were low during that period could be explained by a shortage of safe assets. Such a shortage meant that people were willing to pay a premium for safe assets, which pushed yields lower.
In the midst of the slow recovery, interest rates remained low for a very long time. Critics of monetary policy argued that monetary policy was relatively impotent since the Federal Reserve was holding interest rates so low without evidence that this was helping speed up the slow recovery. Critics of these critics argued that this was wrong. Interest rates were low because the shortage of safe assets had been exacerbated by financial crisis thereby holding down yields.
More recently, I have written about the Treasury Standard, the term which I have used to describe the role of the dollar as the global reserve currency and the U.S. Treasury security as the global reserve asset. I argue that the yield on Treasury securities is determined by the supply and demand for those securities. I also argue that fluctuations in the exchange rate are determined by whether the flow of dollars out of the U.S. are greater than or less than the flow demand for dollars from the rest of the world. And these two things are related.
Suppose that you wanted to test these claims about safe asset shortages or the Treasury Standard. How would you do it? Price theory can be your guide.
First, according to each of these claims, there is something special about U.S. Treasury securities in comparison to other forms of government debt. As a result, we should expect there to be a premium on Treasury securities. Second, the premium on Treasury securities should explained by the fact that these securities are denominated in dollars. Third, according to each of these frameworks, a so-called “flight-to-safety’’ should result in lower yields and an immediate appreciation of the dollar. Fourth, an increase in the foreign demand for Treasury securities should be expected to result in a lower rate of return in terms of the relevant foreign currency than the rate of return in terms of dollars. But that implies that there must be a relationship between the yield on Treasury securities and the dollar exchange rate. In particular, an increase in foreign demand should result in an immediate appreciation of the dollar followed by an expected depreciation of the dollar such that the return in foreign currency is lower than that in terms of dollars. The only distinction between these two views is that the Treasury Standard implies that this should only apply to Treasury securities whereas those who take the safe asset view would argue that these arguments apply to safe assets more generally.
In a recent paper by Zhengyang Jiang, Arvind Krishnamurthy, and Hanno Lustig in the Journal of Finance, these authors use standard asset pricing models to both generate and confirm these predictions. In particular, they start by defining the Treasury Basis. By this, here is what they mean. Suppose that yield on Treasury securities is higher than the yield on some foreign government bond. A simple arbitrage opportunity would be to borrow in the foreign currency, lend in dollars and profit from arbitrage. People who take advantage of this arbitrage will then push these yields toward equality. Of course, things aren’t that simple. If you are borrowing in the foreign currency and lending in dollars, you have exchange rate risk. Thus, the correct arbitrage relationship would imply that any discrepancy between the yield on the U.S. Treasury security and the foreign security would be explained by the expected appreciation or depreciation of the dollar.
Yet, if there is something special about the U.S. Treasury security, then we actually wouldn’t expect this condition to hold because people are willing to pay a premium for this security. Thus, since we can observe these yields and we can observe the expected appreciation (or depreciation) of the dollar from the forward and current exchange rate, any remaining difference is referred to as the Treasury Basis. When this term is positive, it means that the yield on U.S. Treasury securities is higher than the foreign yield, even adjusting for the expected change in the exchange rate. If the Treasury Basis is negative, that implies that the yield is lower that the foreign yield, even accounting for expected changes in the exchange rate. Those who believe the safe asset hypothesis or take the view of the Treasury Standard would expect to find that the Treasury Basis is negative since it is expected that Treasury securities have a premium. This is indeed the case. The authors show that this is true for the overwhelming majority of the period since 1988.
These authors then use standard asset pricing models to derive a relationship between the Treasury Basis and other economic variables. From these models they show that the gap in the convenience yield provided by U.S. Treasury securities in comparison to foreign bonds is proportional to the Treasury basis. Furthermore, they show that a higher Treasury basis is associated with an immediate appreciation of the dollar and an expected depreciation of the dollar. They then find empirical support for each of these results.
This seems to provide some evidence in favor of the safe asset view or the Treasury Standard view. However, this doesn’t by itself resolve debates about monetary policy surrounding the financial crisis. Although one can’t settle the debate, one can at least test these views. If these views are correct, then tight monetary policy corresponds to an excess demand for safe assets. A shock that tightens monetary policy should therefore cause higher convenience yield on U.S. Treasury securities, a lower (more negative) Treasury Basis, and an appreciation in the dollar. In their paper, Jiang, et al. find that contractionary shocks to monetary policy cause a more negative Treasury Basis. In addition, the variation in the Treasury Basis explained by monetary policy is associated with an appreciation of the dollar.
Thus, in summary, what the safe asset view and the Treasury Standard suggest is that it is the critical role of the dollar and U.S. Treasury securities as the global reserve currency and global reserve asset, respectively, that is important. Because of these roles, one would expect that yields on U.S. Treasury securities are lower than comparable foreign alternatives even taking into account expected changes in the exchange rate. This premium on Treasury securities is precisely what we see in the Treasury Basis. In addition, these assets are uniquely related by these roles and therefore we should expect to see changes in the value of these securities and changes in the value of the dollar move in conjunction with one another. Again, the relationship between the Treasure Basis and the dollar exchange rate reflect this relationship.
Some More General Conclusions
What all of this hopefully illustrates is that one is capable of doing macroeconomic analysis using price theoretic tools. As we like to argue here at Economic Forces, you can often narrow things down to supply and demand, even in macroeconomics. Price theoretic foundations for macroeconomics are likely to be found in financial markets. This is not only because macroeconomic fluctuations are likely to be reflected in asset prices, but also because the predictions of the theory can often be combined with no-arbitrage conditions to derive predictions about how precisely asset prices would be expected to move. In addition, those price theoretic foundations can potentially provide a theoretical basis for explaining deviations from more simplified no-arbitrage conditions. Furthermore, fluctuations in those “deviations” might then serve as a valuable test of the theory. In short, a productive macroeconomics with price theoretic foundations has finance at its core.