For the last decade or so financial institutions have relied increasingly and excessively on short-term financing, risky business that could quickly go south when a souring economy made it impossible for firms to roll over their short-term financing.
Some researchers have argued that the financial sector's tilt toward short-term financing was merely the industry responding to a lack of available government debt instruments, such as short-term risk-free Treasuries, that were much in demand by investors during the global savings glut. So investors purchased short-term bonds from financial institutions instead, which the industry churned out en masse.
One result: many financial institutions placed not only themselves, but the broader financial system, at risk. When a financial institution can no longer repay its debts, it may have to sell assets quickly, with adverse consequences for other financial institutions.
“Most people agree that the financing of large financial intermediaries put the larger financial system at risk”
Can the government do anything to discourage short-term borrowing by the private sector? HBS Professor Robin Greenwood, Harvard University and Harvard Business School PhD candidate Samuel Hanson, and Harvard University Professor Jeremy C. Stein take on this question in a recent working paper titled A Comparative Advantage Approach to Government Debt Maturity. Their paper asks how the government should optimally finance its debt, and whether this choice has any bearing on private sector financing choices.
The authors first consider the government financing problem in isolation, arguing that the government may try to "borrow cheap" by issuing short-term Treasury bills, which embed a premium relative to long-term financing, because of Treasury bills' "money-like" properties. However, the government may be reluctant to finance itself entirely with short-term debt, because doing so would incur significant rollover risk-in the extreme case, the government would not want to roll over its entire debt daily.
How do private firms enter the picture? The authors argue that private firms also choose from a menu of long- and short-term securities, and thus compete with the government in creating money-like securities. If Treasury bills are particularly scarce, for example, the private sector may tilt towards short-term financing. The government can counteract some of this by issuing more short-term debt.
Here Robin Greenwood discusses the research findings in the paper, and the need to understand better the potential implications for policy.
HBS Working Knowledge: Why did you decide to study the maturity structure of government debt?
Robin Greenwood: We have some earlier research on the determinants of corporate borrowing. In that work, we noticed that the maturity structure of corporate debt responds strongly to changes in the maturity of government debt. So when the government issues longer-term debt, for example, corporations back off and issue short-term instead. We called this "gap-filling," in the sense that firms fill in the gaps created by government financing policy. In that paper, we had thought about government maturity policy as exogenous, largely in the background.
But we ultimately realized the implications of our findings, which is to say that the government could actively influence the corporate sector's borrowing decisions by shifting its own financing between T-bills and bonds.
Why should we care about how the private sector finances itself? It is becoming increasingly appreciated that financial institutions in the past ten years have relied excessively on short-term financing. The investment bank Lehman Brothers, for example, collapsed because it could not roll over its short-term financing. More broadly, if we think about the financial sector, most people agree that the financing of large financial intermediaries put the larger financial system at risk. Once you recognize this, it's easy to argue for more regulation, but regulation is difficult, particularly when banks have figured out ways to shift their financing off-balance sheet.
This is where our "gap-filling" idea comes in--the government can dissuade firms from issuing short-term debt by simply making it less attractive to do so. It can accomplish this by issuing more short-term debt (T-bills), raising the yields on short-term debt relative to long-term debt. Once we recognized the potential for government debt maturity policy to have an impact on private sector financial decisions, we became interested in government debt maturity more generally, and the end result is this paper.
Q: Where does the title of the paper come from? When I think of "comparative advantage" I usually think of principles of trade between countries.
A: The title of the paper refers to the idea that, in choosing the optimal maturity structure of its debt, the government balances the costs of rollover risk (in the extreme case, the government faces enormous rollover risk as it crowds out the private sector fully) with the system-wide benefits of crowding out private sector money creation. In other words, the government should keep issuing short-term bills as long as it has a comparative advantage over the private sector in the production of riskless money-like securities.
Q: One way to read your paper is that it suggests that budget deficits have a silver lining.
A: Yes, you could say that our paper is about the "silver lining" of short-term government debt. The more the government issues, the more likely the financial sector is to rely on more sensible long-term financing. We want to be a bit cautious, however. When the level of public debt is high, investors demand for short-term safe securities is most likely satiated, meaning that there is not much the government can do to affect the relative prices of long- and short-term debt. The good news is, when this happens, the government has done as much as it can do to dissuade private sector borrowing at extremely short maturities. Of course, the government may still use other means to regulate borrowing, such as quantity regulation.
Q: During the financial crisis, the government issued an enormous amount of new debt, and dramatically changed its maturity structure. Does your paper have anything to say about that?
A: Yes. During the early days of the financial crisis, the private sector's ability to create short-term, money-like claims became significantly impaired. This required a simple government response: expand the supply of riskless short-term bills. Such reasoning seems to have been borne out in Treasury policy during the height of the financial crisis, when the U.S. Treasury issued $350 billion of short-term bills within a week of Lehman Brothers' failure, as part of the "Supplementary Financing Program." The proceeds from this program were lent to the Federal Reserve, which in turn bought long-maturity assets. Following the abrupt shortening, government debt maturity quickly reverted to normal levels in 2009. Thus, when the financial sector's ability to produce money-like securities disappeared, Treasury quickly ramped up its issuance of money-like claims. However, as the anticipated debt burden grew, concerns about rollover risk eventually trumped the desire to cater to money demand, and maturity structure was once again lengthened. So we think our paper provides a useful framework for thinking about policy, both past and future.
Q: Do you think your suggestions are practical?
A: Jeremy and Sam spent most of 2009 working in Washington on financial policy, so I'll defer to them. But keep in mind that in an era of enormous deficits and ballooning ratio of debt-to-GDP, the question of how the government is to finance itself is of central importance.
One thing we learned while working on this paper was that there appears to be a very significant money premium in the shortest maturity T-bills, those with maturities of less than four weeks. And much short-term financing on the part of financial institutions is extremely short maturity, often overnight. So, it seems to us that the Treasury could both lower the cost of financing its debt, as well as having a potentially powerful crowding out effect on the private sector, by moving its issuance to two- and four-week bills, for example.
In the paper we worked out a simple example of how this might work. At the end of last year, there were about $1.8 trillion of T-bills outstanding, with an average maturity of 102 days. One simple option is to simply shuffle maturities within the category of T-bills. For example, suppose the government were to cut the maturity by half, to fifty-one days. To offset the change in duration, the Treasury would have to swap approximately $52 billion of ten-year bonds for twenty-year bonds. This represents only about one percent of bonds and notes outstanding--not a big deal.
Q: What are you working on now?
A: Over the past few years, the three of us have become much more interested in the credit markets, and we continue to be excited by the research potential here.