Description
TitleEssays on foreign reserves
Date Created2022
Other Date2022-05 (degree)
Extent169 pages : illustrations
DescriptionIn 1999, Martin Feldstein stated in an NBER working paper (No. 6907) that international economic crises in the 80s and 90s showed the extreme social and economic costs that sudden shifts in market perceptions can cause even in countries with solid fundamentals. Therefore, the most important lesson from these events was that "Liquidity is the key to self-protection''.
The worldwide build-up in official foreign reserves holdings, since the start of the 21st century, was interpreted as a reflection that governments had taken this lesson by heart. However, accumulating reserves as self-insurance is costly since it deviates resources to low-return assets, and the economy only reaps its benefits if the negative shock happens.
In the same paper, Feldstein argued that, when it comes to financing an acute liquidity shortage, holding reserves was as valuable as having the ability to borrow on relatively short notice. If this is true, why do countries choose to incur the cost of accumulating reserves instead of borrowing? In the following three chapters, I study the role of foreign reserves as a liquidity instrument. The premise of this thesis is that preemptively accumulating reserves for precautionary purposes reflects an economy that can't borrow during adverse external conditions. The first chapter, "Foreign Reserves, Fiscal Capacity, and Lender of Last Resort", is motivated by the stylized fact that emerging markets and, to a second degree, low-income countries led the build-up in foreign reserves holdings. This chapter develops a novel theoretical framework to study why some governments accumulate reserves for precautionary purposes while others do not. This model is part of the liquidity literature, pioneered by Holmstrom and Tirole, where the amount of return that it can credibly promise lenders determines the ability to borrow of any agent. What can be credibly promised is known as pledgeable return or output. More precisely, I devise a small open economy that borrows from international markets to finance production. Financial frictions characterize the relationship between private domestic agents and global markets, creating a gap between this economy's future output and expected pledgeable output. This gap implies that this small open economy is solvent but illiquid in some states of world markets where external funding costs are high. While total output, valued at those rates, is an attractive investment to foreign lenders, pledgeable output is not. As a result, foreign lending is interrupted, and the economy experiences what is known as a sudden stop. I introduce a government that acts exclusively as a lender of last resort to this setting. This government can potentially prevent sudden stops because it has the exclusive right to tax private agents. This right, in principle, allows governments to overcome financial frictions implying that sovereign borrowing is not limited by pledgeable output but by how much taxes it can collect. However, in reality, the ability of a government to increase taxes given an existing tax structure and enforcement abilities is finite, and it is a concept denoted as fiscal capacity. I introduce this concept to the model where it determines the fraction of the gap between pledgeable and total output that backs up sovereign borrowing. I show that fiscal capacity is the key determinant of an economy's ability to borrow in equilibria. Government intermediation between domestic agents and international markets can eliminate sudden stops if fiscal capacity is sufficient to increase aggregate pledgeable output to the point that it is attractive for foreign lenders. I find the minimum level of fiscal capacity that is sufficient to attract foreign borrowing and eliminate sudden stops. Governments that don't have sufficient fiscal capacity can still eliminate sudden stops at the cost of preemptively accumulating reserves. However, I show that reserves accumulation is not optimal for every government since the cost of accumulating increases as fiscal capacity is lower. Therefore, only governments with intermediate fiscal capacity hoard reserves and eliminate sudden stops. In contrast, governments with low fiscal capacity don't accumulate reserves and prefer that their economies remain subject to sudden stops. Additionally, I show that low fiscal capacity creates the potential demand for reserves through lower fiscal space and by creating a crowding-out effect on private liquidity. More importantly, I argue that the crowding-out channel by itself is sufficient to generate the demand for reserves while the fiscal space channel is not. Intuitively, the same financial frictions bound sovereign borrowing of a government with low fiscal capacity and private liquidity. As sovereign borrowing increases, the amount of pledgeable output, net of taxes, that can be promised by private agents is lower. This destruction of private liquidity is the reason why reserves are useful. Reserves compensate for the lack of fiscal capacity because they don't crowd out private liquidity in times of need. The results of this chapter imply two hypotheses that can be tested in the data. First, at any given year, countries with lower fiscal capacity should have a greater stock of reserves. And second, as a country improves fiscal capacity, its reserves accumulation should be non-linear and concave. In the second chapter, "Fiscal Capacity and Foreign Reserves - An empirical approach", I take these two hypotheses to the data. To do so, I construct an unbalanced panel for 67 countries with data ranging from 1991 to 2016. My econometric specification builds on the literature focused on finding empirically different motives behind foreign reserves accumulation. The typical specification has the stock of reserves as a share of GDP as a dependent variable and proxies for current account shocks, financial account shocks, and mercantilist motive as regressors. I extend this specification by including variables explaining the number of reserves that a country accumulates according to the theoretical framework. First, I include external public and private liabilities and domestic financial liabilities to proxy an economy's financial development level. Second, I add the government's gross debt and fiscal balance to control for the ability to borrow due to a country's fiscal space. Third, I include GDP per capita as a measure of an economy's level of development. This last variable is relevant to address concerns that the fiscal capacity proxy could be capturing the level of development of a country and not the ability of a government to tax. Fiscal capacity, which is the ability to raise taxes given the tax structure and enforcement capabilities following the work of Besley and Persson, depends on governments making previous investments in compliance and enforcement institutions and information systems. For the empirical exercise, I proxy fiscal capacity with the share of total revenue collected through income tax. Countries are more likely to have made these investments if sophisticated taxes collect more revenue. I show that there is a between-negative elasticity between foreign reserves and fiscal capacity that is estimated around -0.2 depending on the sample. I argue that this is supporting evidence for the first hypothesis. Moreover, this result is robust to several specifications, such as including an original sin index and running a principal component analysis. Moreover, as a robustness check, I take advantage of the Lasso selection model and find that fiscal capacity is chosen 68\% of the time, out of 1,000 replications, as a relevant variable explaining foreign reserves stocks. This variable is only below imports and broad money, selected in every replication. Similarly, when I include country-fixed effects in the specification, I find a consistent quadratic and concave relationship between reserves and fiscal capacity to support the second hypothesis. This result is robust to including original sin index and even when running the specification by regions. Overall, the main contribution of the first two chapters is to underscore a novel motive, both theoretically and empirically sound, behind foreign reserves accumulation. To the extent that governments provide liquidity ex-post, fiscal capacity determines an economy's ability to borrow during adverse conditions and the potential demand for reserves to compensate when necessary. In the third chapter, "Foreign Reserves: Stock versus Flow", I empirically evaluate why some countries do not tap into their reserves stock during global shocks. I use the Global Financial Crisis (GFC) as an event study. In plain view, reluctance to use reserves is inconsistent with the idea that reserves satisfy precautionary purposes. However, in this chapter, I argue that it is not. Like banks facing a run, only solvent economies survive a liquidity shock. Before being a liquidity instrument, reserves are external assets and, as such, provide solvency services. I argue that one unit of reserves is an instrument that provides two services: liquidity when it is sold (flow) and solvency when it is as part of the stock. Therefore, the decision to sell depends on the trade-off between the marginal liquidity and the marginal solvency value. In turn, I expect this trade-off to lean towards marginal liquidity value when there are few reserves and more than enough reserves in the balance sheet. At the extremes, marginal reserves don't make a difference for solvency. I test this hypothesis during the GFC for 37 countries. I proxy the solvency value of reserves with reserves balance sheet exposure relative to total external liabilities. I find a convex quadratic relationship between the magnitude of depletion of reserves and the balance sheet exposure as expected, even when controlling for other motives for using reserves during crises.
NotePh.D.
NoteIncludes bibliographical references
Genretheses
LanguageEnglish
CollectionSchool of Graduate Studies Electronic Theses and Dissertations
Organization NameRutgers, The State University of New Jersey
RightsThe author owns the copyright to this work.