Comments on Part IV

Adrián Armas, Eduardo Levy Yeyati, and Alain Ize
Published Date:
July 2006
  • ShareShare
Show Summary Details
claudio Irigoyen

The two chapters in this section tackle the very important topic of de-dollarization, certainly still a central issue for many Latin American economies. Chapter 10 addresses the question of how inflation-linked bonds can improve the composition of government debt in terms of maturity structure and rollover risk. The authors stress the role of indexed government debt as a key ingredient for the development of deeper markets that include privately issued indexed debt as well. Chapter 11 analyzes several de-dollarization experiences, highlighting the defining characteristics of both successful and unsuccessful ones.

I will mostly concentrate my comments on the first topic, related to the role of indexed debt (though not necessarily indexed to inflation) in a framework of optimal management of public debt. I will also discuss some issues related to the development of a market for indexed securities issued by the private sector.

The first question to be addressed is: what is the optimal composition of a public debt portfolio from the government’s perspective? The answer depends not only on the credibility of monetary and fiscal policy but also on the stochastic structure of the economy. Low credibility induces a higher proportion of indexed debt in an optimal portfolio (usually indexed to inflation) if the government does not want to confirm a relatively high risk premium for long-term nominal debt. Alternatively, governments may prefer to issue short-term debt as a solution to the same problem, at the expense of a higher rollover risk. Excessive short-term debt or long-term debt with floating rates may create a problem of fiscal dominance, constraining in this way the conduct of monetary policy. On the other hand, tax-smoothing considerations prescribe the issuance of nominal debt such that the real value of debt correlates negatively with shocks that might hit the economy (in bad times the real value of debt falls).

A more precise prescription depends on the particular stochastic structure of the economy, where indexed debt appears to be optimal to deal with aggregate spending shocks and nominal debt for supply shocks. For instance, a positive covariance between primary deficits and inflation favours the issuance of nominal debt, while a high variance of inflation increases the attractiveness of inflation-indexed bonds to avoid paying a high inflation-risk premium. In order to signal commitment to healthy public policies, many countries have indexed government debt to the exchange rate, or simply issued foreign currency-denominated debt instead of indexing it to inflation. The choice of indexation benchmark is also a function of the stochastic structure of the economy. If inflation and the real exchange rate correlate positively, nominal bonds and foreign currency-denominated bonds are complements. Nonetheless, foreign denominated debt might be too risky when the country is subjected to sudden stops, the risk being negatively related to the openness of the economy and the size of the tradable sector, and positively related to the degree of liability dollarization of the economy as a whole.

A corollary is that, depending on the structure of the covariance matrix of shocks, there might be a role for indexed debt even in a full credibility environment. Based on the same considerations, it might be optimal to issue some nominal debt even under severe credibility problems.

The analysis above should not be considered as purely static. It is a dynamic problem in the sense that optimal debt composition must be revised over time as credibility and reputation evolve and the stochastic structure of the economy changes. On the other hand, debt composition has an endogenous impact on the overall cost of debt, as credit risk (or default risk) is not independent of the incentives of the government to follow time-inconsistent policies, based on the actual composition of the debt portfolio.

Another recently popular product is debt indexed to the performance of the economy (GDP-linked bonds). The pay-off of these bonds is a positive function of the rate of growth of the economy, acting, therefore, as an automatic stabilizer of the debt to GDP ratio (although this ratio is not a good measure of the debt sustainability). As already underlined, the convenience of this type of bonds depends (among other things) on the covariance between the rate of growth of the economy (also a random variable) and the rest of the shocks to which the economy is subjected. Some observers suggest that in countries that depend heavily on the price of specific commodities public debt should be indexed to them based on tax-smoothing considerations.

Another question of interest is: what is the role of government debt in developing markets for privately issued indexed debt? What are the issues at stake in the indexation of private debt?

The development of a market for indexed government debt creates a positive externality for private borrowers. If there is a deep market for indexed government debt, it is much easier to price private indexed debt, even if indexation is related to different variables.

Nonetheless, the main consideration when analyzing the optimality of issuing indexed debt from the point of view of a corporation is the trade-off between hedging and liquidity. Hedging considerations recommend indexing to the relative price of the product commercialized by the company, or similarly, to the key relative price that influences its revenues. This argument rationalizes why governments index public debt to inflation. The problem with this prescription is that if all companies index debt to their key relative prices, a scale problem arises. On average, individual issuances are not large enough to ensure that those privately issued papers will be liquid enough, which will certainly have an impact on the cost of debt. One alternative is to index debt to inflation or the exchange rate (an imperfect hedge but a more liquid instrument). In that case, the role of the government in developing markets for indexed debt becomes crucial.

It is important to distinguish between causes and consequences of inflation in order to analyze the merits of inflation-indexed bonds. It has been argued that the issuance of inflation-protected bonds fosters indexation of the whole economy through fostering wage indexation and increasing inflationary inertia. This argument is flawed. Bad monetary or fiscal policy are instead to be blamed. The indexation of assets is a way to complete markets since it is definitely welfare improving.

Finally, a few words about the second topic: de-dollarization. Independently of the way de-dollarization took place in some emerging economies, the challenge is how to avoid re-dollarization. In that respect, much of the homework concentrates on the regulation of the financial system. One of the main features of dollarized financial systems is their exposure to real exchange rate fluctuations (that feed naturally into credit risk), reflecting the fact that they on lend denominated deposits to individuals or firms with an income in domestic currency.

Among other things, it is important to include in the policy menu: (i) high provisioning requirements for dollarized assets; (ii) high reserve requirements (with fairly low remunerations) for dollar-denominated deposits, coupled with restrictions on applications (to avoid massive currency mismatches); (iii) the development of markets for indexed assets as a substitute for dollarization (an intermediate step towards nominalization); and (iv) the development of derivative markets that help to deal with foreign exchange rate volatility. Needless to say, those initiatives are not substitutes for sound monetary and fiscal policies. The latter are absolutely necessary conditions for price stability and therefore the main ingredients of any de-dollarization strategy.

    Other Resources Citing This Publication