Information about Western Hemisphere Hemisferio Occidental
Chapter

Comparing Menu Items: Methodological Considerations and Policy Issues

Editor(s):
Peter Wickham, Jacob Frenkel, and Michael Dooley
Published Date:
March 1989
Share
  • ShareShare
Information about Western Hemisphere Hemisferio Occidental
Show Summary Details
Author(s)
MICHAEL P. DOOLEY and STEVEN A. SYMANSKY* 

It is argued that interest payments and debt buy-backs can be viewed as alternative forms of debt reduction. It follows that the division of available resources between debt reduction and interest payments can be interpreted as a renegotiation of the original contracts. Once the terms of this renegotiation are fully incorporated into market prices, a simple procedure for valuing new money and other menu items is presented. A simulation model illustrates the results of buy-backs under alternative assumptions about expectations of creditors.

A SIGNIFICANT DEVELOPMENT in the implementation of the debt strategy has been the growing emphasis on “broadening the menu” of financing instruments. When employed in conjunction with firm pursuit of growth-oriented adjustment policies, a broader array of financial instruments can be seen as easing the problem of channeling needed finance to heavily indebted countries. The approach of broadening the menu seeks to elicit additional financial flows by appealing to the varying portfolio preferences of creditors and potential creditors. Portfolio preferences can vary for a number of reasons: differing assessments of the economic prospects of debtor countries, different degrees of risk aversion among creditors, different regulatory and tax environments, different perceived comparative advantage and long-term strategies among creditor institutions, and different initial balance sheet structures.

More recently, the recognition has grown that techniques to reduce debt—provided they are market based and voluntary—can help indebted countries return to eventual creditworthiness. Two recent schemes have been the Bolivian debt buy-back and the Mexican debt exchange. In the Bolivian case, additional financing from official sources was used to buy back commercial debt, while in Mexico, an unexpected accumulation of reserves provided collateral for the Mexican authorities to exchange on favorable terms new debt for existing claims. These operations have sparked interest in market-based debt reduction techniques, but they have not as yet led to major additional initiatives.

This paper attempts to deal with methodological issues associated with debt reduction. These issues are particularly significant in cases in which countries need new money to meet a prospective financing gap and in which all parties recognize that future debt service will be so onerous that efforts to reduce debt would be desirable in order to restore a country’s external creditworthiness. In such cases, simply adding to existing debt to cover a financing gap compounds the longer-term problems of indebtedness. Nevertheless, countries with a financing gap do not, by definition, have resources readily available to devote to some of the more obvious techniques of debt reduction (e.g., debt buy-backs).

The plan of the rest of the paper is as follows. Section I deals with how debt reduction techniques can be compared with new money. Section II analyzes a stylized financing package containing new money and debt reduction techniques. It focuses therefore on equivalency among financing options from the point of view of creditors once the resources to be used for debt reduction are identified and proposes a simple methodology for evaluating such options.

I. New Money Versus Debt Reduction

In this section, new money and debt relief techniques are evaluated from the points of view of debtors and creditors. It is assumed that the typical debtor faces a “financing gap” in that resources available for debt-service payments are expected to fall short of interest and amortization payments due during a given time period. Thus, a financing package is needed to allow the debtor to satisfy its contractual obligations.

Consider, for example, the case of a debtor who has $10 in interest payments due, but has only $5 in cash flow to finance the payments. (For simplicity, amortization payments are ignored.) Assume further that initial stock of debt is $100 (face value) and that the discount on the debt is 50 percent. With debt reduction techniques included in the menu, the debtor might reason as follows: if interest and amortization payments are limited to, say, $3, then $2 will be available for a debt buy-back. Given the discount at which the debt is being traded, $4 of outstanding debt can be extinguished. Of course, if only $3 is allocated to interest and amortization, a further $7 of new money will have to be borrowed to meet contractual payments. The stock of debt at the end of the period will be $103: the initial $100, plus the $7 in new money, less the $4 of debt extinguished through a buy-back.

The creditor may have less incentive to agree to the buy-back, essentially because funds used to finance a buy-back would otherwise have been available to meet contractual interest payments. The creditor might prefer to receive the full $5 available in the form of interest payments. He could then lend $5 in the form of new money and complete the period with a claim of $105.

An alternative way of viewing the comparison is to regard both buy-backs and contractual payments as debt reduction techniques. In the above example, if the debtor had no resources available for debt service, indebtedness would grow from $100 to $110. The $5 that is available can be applied either to interest and amortization payments or to buy-backs. The difference is that interest and amortization payments follow contractual terms and therefore retire less debt than buy-backs or asset exchanges that follow market prices. For this reason, an offer of a buy-back in a menu may be viewed as a renegotiation of the terms at which debt will be serviced.

It follows that credit items and debt reduction items cannot be made “equivalent,” but instead should be viewed as striking a new balance between the interests of debtors and creditors. It might be better to interpret a financing package that includes buy-backs as follows: debtors and creditors, recognizing their common interest in restoring a “sustainable” debt-servicing position, reach agreement on a flow of interest and amortization payments that can be considered appropriate to the circumstances faced by the debtor. Beyond this stream of payments, the debtor (perhaps with contributions from other interested parties) undertakes to make “extra” resources available, provided these “extra” resources reduce debt.

As shown in the technical appendix, the growth of debt relative to debt-service capacity, and thus the market price of debt, can be quite sensitive to this negotiated division of payments by the debtor country between interest payments and buy-backs.

At one extreme, if a debtor devotes all its resources to contractual debt payments, but still has a large financing gap, then new borrowing can cause debt to grow more rapidly than debt-servicing capacity. On the other hand, if a country allocates a relatively small amount to debt service, then the current value of the future stream of debt-service payments will be correspondingly reduced. While it would become easier in these circumstances to buy back the depreciated debt instruments, such a development might be seen as being inconsistent with the cooperative approach.

In the simple examples developed above only one kind of debt is issued by the debtor government. As a practical matter, debtor governments issue a variety of debt instruments in international and domestic markets, and it is important to distinguish between net and gross debt reduction. As is developed in greater detail in the next section, menu items that reduce the gross value of external syndicated debt may do so at the cost of obliging the debtor to increase the value of other internal and external liabilities. A full evaluation of “debt reduction” made possible by a financing plan requires the identification of funds available to finance the debt reduction that are not obtained by issuing alternative types of debt.

The successful incorporation of debt reduction techniques in Bolivia, Mexico, and Chile indicates that debtors and creditors have found debt reduction techniques useful. The experience of these cases, however, plus other countries’ efforts to incorporate explicit debt buy-back arrangements indicate that it is not easy to obtain creditor agreement, particularly when new money is also required from the creditors. In general, the attractiveness to creditors of debt reduction techniques can be enhanced in the context of a negotiated financing package. In such a package, debtors can provide assurances that improved economic performance will enhance the value of remaining debt. Commitment to a sound economic program, particularly if supported by the Fund and the World Bank, could provide strong incentives for debt reduction techniques. In addition, official creditors can also make clear their own commitment to the debtor’s adjustment effort. This might include both financial resources and regulatory initiatives designed to enhance the attractiveness of menu items to private creditors

II. Comparisons of Financing Options

The conclusion from the foregoing is that the negotiated blend of new money and debt reduction techniques will require a negotiated and cooperative agreement among interested parties. In this section, a framework is developed that allows for comparisons among financing options. This framework, it should be emphasized, is illustrative only and does not go into practical detail. In particular, no attempt is made to take into account the wide variety of regulatory and tax structures imposed on banks in different countries or the different attitudes of individual banks toward their long-run involvement with debtor countries. As an example, banks with a relatively small exposure in a given country might find an exit bond more desirable than would be suggested by the calculations presented below since it would reduce the administrative costs of being involved in future financing packages. Such banks would presumably pay a premium for an exit instrument. Moreover, an instrument that allowed creditors greater flexibility in accounting for losses, or greater discretion in establishing reserves against possible losses, might be more valuable to some creditors than the basic “new money” instrument. In this respect, the “terms factor” relevant to a menu item might be enhanced by regulatory authorities in creditor countries. Finally, as with any financial instrument, the tax treatment of earnings and capital gains or losses essentially determines how various potential holders evaluate the contractual terms of alternative menu items. As in any regulatory environment, the tax treatment of a menu item can be seen as an opportunity for creditor governments to enhance the value of menu items.

Characteristics of Menu Items

To keep the exercise simple, it is assumed that investors consider only the expected market value of various options. In practice, menu items with uncertain yields would be less valuable, and this could be considered an additional factor in more realistic exercises. It will be useful to identify at the outset three characteristics of a menu item.

Contractual Terms

Menu items can incorporate many contractual terms. As in conventional credit markets, a variety of debt and equity contracts will appeal to different creditors. Assume that the basic “new money” option is a security with contractual terms identical to existing debt. Consider now a similar credit that carries a contractual interest formula equal to one half that on the basic “new money” instrument. If a prime borrower issues such an instrument, we would predict that the additional discount at which it traded would precisely reflect the fact that the stream of interest receipts was half as valuable. The discount would be greater for a relatively long-lived instrument than for a shorter-dated instrument (since in the former case the share of interest payments relative to amortization payments in the present discounted value of the instrument is greater).

Country Risk

Another important attribute of a menu item is the extent to which interest and principal payments are subject to the risk of default. It is useful to decompose the yield into a part that carries “pure” debtor country risk, defined as having risk equivalent to that associated with existing syndicated credits, and a part that is “risk-free,” or is expected to be fully serviced. Since the basic new money instrument is defined as being identical to existing syndicated credits, it carries the same country risk.

A menu item can be differentiated from new money in terms of country risk through subordination, guarantees, or collateral. If investors believe that an exit bond will be serviced in circumstances in which new money instruments will not be serviced, the exit bond will carry less country risk. Subordination of this sort can be associated with any menu item, including debt-equity conversions, and domestic currency bonds. The difficulty in evaluating country risk is its dependence on investors’ perceptions that the subordination is credible.

Guarantees by a third party, or collateral held outside the control of the debtor, also reduce the country risk associated with a menu item.

Exchange Ratio

The final important characteristic of a menu item is the exchange ratio offered between the menu item and existing credits. These exchange ratios are sometimes difficult to identify but are implicit in any menu offering. In every case, the menu instrument is offered to discharge a given interest obligation, or what is the same thing, to retire an existing credit. The basic new money instrument is offered at par to discharge an obligation that has accrued or will accrue. On the other hand, if the debtor offers $1 face value of an exit bond to retire $2 of interest due or in exchange for $2 of existing debt, this can be considered an exchange ratio of 0.5.

Illustrative Comparison of Menu Items

Classifying menu items in terms of the three characteristics identified above introduces a flexible way of evaluating financing options. Table 1 presents an illustrative comparison of typical menu items that might be relevant for a typical creditor. For purposes of illustration, it is assumed that the country’s existing debt consists of syndicated credits with a contractual yield of 8 percent and that these credits sell at a 50 percent discount in the secondary market.1 It is further assumed that the “new money” option involves the creditor accepting a new syndicated credit identical to existing credits. This is convenient because it means that if we compare a menu item to the “new money” option, the same comparison will generally apply to an exchange of existing debt for that menu item.

New Money Security

A typical new money option pays a market-related interest rate. For simplicity, it is assumed that the yield is fixed at 8 percent on both existing debt and new money securities.2 Thus, the new money security carries the same contractual yield as existing debt and its contractual-terms factor is defined as 1.0. Since payment of interest and principal is fully the obligation of the debtor country and because new money is indistinguishable from existing debt, the country-risk factor is equal to the full 50 percent market discount on existing debt, giving a country-risk factor of 0.5. Finally, since the face value of new money security is usually offered to settle an equivalent interest payment, the exchange ratio is 1:1 implying an exchange ratio factor of 1.0. In this case, the product of the three factors is 0.5 so that the standard new money instrument is worth 50 percent of a cash interest payment to the creditor.

Exit Bonds

Exit bonds typically carry a lower contractual interest rate, and in this example 4 percent is assumed. If any debtor issues a long-term security with an interest yield half that of similar obligation, the market price of the low interest rate security will be about one half that of the high rate instrument.3 Thus, the contractual factor is 0.5. The country-risk factor is more difficult to determine. Three cases are examined. In Case 1, it is assumed that exit bonds are believed by investors to carry exactly the same default risk as existing debt. Thus, the country-risk factor for Exit Bond 1 is 0.5. In Case 2, it is assumed that investors expect exit bonds to be repaid with certainty; thus, the country-risk factor is 1.0. A third case is one in which one half of the payments on the exit bond are guaranteed by a collateral, the remaining risk being pure country risk. In this case, the country-risk factor is 0.75. In every case, the exit bond is exchanged for an equivalent face value of existing debt so the exchange factor is 1.0.

Case 1:100 Percent Interest Payment
YearPriceDebtPresent valueBuy-Back
10.381,060.00399.990
20.361,126.00399.990
30.331,198.60399.990
40.311,278.46399.990
50.291,366.31399.990
100.201,956.24399.990
150.142,906.35399.990
200.094,436.49399.990
250.066,900.80399.990
300.0410,869.59399.990
Case 2:75 Percent Interest Payment—25 Percent Buy-Back
YearPriceDebtPresent valueBuy-Back
10.311,027.37314.8510
20.301,066.40316.3410
30.291,108.18317.9710
40.281,152.95319.7710
50.271,200.92321.7510
100.221,498.59335.0410
150.191,926.31356.4410
200.152,616.90390.9010
250.103,879.85399.990
300.076,004.32399.990

Exit Bond 1 carries a combined factor of 0.25 and from the creditor’s standpoint is clearly inferior to “new money” since it is simply a low interest variant of the new money security. Exit Bond 2, with a combined factor of 0.5, is equivalent to a new money security or existing debt because the assumed reduction in country risk offsets the low interest rate. Exit Bond 3 is also inferior to new money securities but less so because of the collateral.

Debt Equity Swaps

Although an equity security does not have a contractual rate of return, there will be an expected rate of return, adjusted for project risk. While there may be projects with relatively high risk-adjusted rates of return, it is assumed for simplicity that the expected rate is again 8 percent in terms of U.S. dollars. For Debt Equity Swap 1, it is also assumed that 100 percent of the payments to creditors are subject to country risk. Finally, it is assumed that the debtor government exchanges an equity with a market value of $1 for an equivalent interest payment. Under these conditions, the combined factor is 0.5, and this swap offer is therefore identical to the new money security.

It is possible, however, as in Debt Equity Swap 2, that investors believe that dividend payments on equity are not subject to country risk or that at some point in the near future the equity can be sold for cash without penalty. This implies a country-risk exposure of zero and a combined factor of 1.0—clearly superior to the new money security as far as the creditor is concerned. Debt-Equity Swap 3 reflects the fact that potential investors would be willing to swap $2 in existing debt for $1 in equity if the debtor authorized the conversion rights. In this case, the lower country risk is offset by an exchange factor, so that the cash value is equivalent to the new money security.

Domestic Currency Bonds

The contractual yield on domestic currency bonds must be translated into a dollar equivalent taking into account expected movements in the local currency’s exchange value. Again, we assume that the expected value is 8 percent so that the contractual-terms factor is 1.0. Domestic currency obligations of private sector debtors may carry a lower country risk. In particular, if the creditor can induce the private debtor to prepay its obligation, it might be possible to convert the domestic currency into dollars at a parallel market exchange rate. In this example it is assumed that the domestic currency obligation is prepared or sold at par for domestic currency. Thus, the country-risk factor is 1.0, although a 50 percent discount in the parallel exchange market reduces the exchange ratio to 0.5. It follows that the cash equivalent is the same as the new money security.4

Cash Buy-Back

A cash buy-back at market prices provides the creditor with the ability to purchase a safe financial asset. As with any cash transaction, the creditor can invest in an instrument with a market-related yield so that the contractual factor is 1.0. Moreover, since the asset carries no country risk, the country-risk factor is 1.0. Finally, the exchange ratio in the case of the cash buy-back is 1:2 since the debtor is buying its own debt at a 50 percent discount. Thus, the buy-back option priced at the market discount is equivalent to the new money security.

Overview

With this overall framework, there is no difficulty, in principle, in evaluating even very complicated menu options. With more carefully specified and more realistic menu items, the calculation will not produce round numbers—but the methodology will carry through. Still, the problems in making such equivalency calculations should not be underestimated.

Perhaps the most difficult problem is determining a country-risk factor for individual items. Suppose it is true that small issues of exit bonds will be serviced before other debt. Implicit seniority of one type of asset means, for a given total availability of resources, a reduced flow of resources to service other assets. This suggests that the discount on other debt will tend to increase and the equivalencies in the table will change. Moreover, since country risk has a large element of subjective judgment, it will be hard to estimate, ex ante, the risk factor attaching to individual assets. This risk factor will be reflected, ex post, in the differential discounts at which different assets trade.

Another difficulty in such exercises is that one cannot determine, by examining a financing package, the overall amount of debt reduction made possible. As discussed in Section I above, the net debt reduction depends on the initial blend of new money and debt reduction techniques in the financing package. Any external debt reduction over and above this amount that the debtor is obliged by a financing package to undertake would require offsetting increases in domestic government debt. This could, in turn, alter exchange rates and other factors assumed constant in the analysis. More generally, an important extension of the analysis would involve evaluating each menu item in terms of costs and benefits to the debtor.

III. Conclusions

There is now considerable support for exploring how debt reduction techniques can be included, on an agreed basis, in financing packages. There is also a measure of agreement that debt buy-backs and debt-equity exchanges are promising avenues for reducing external debt. The analysis developed above suggests that a blend of debt reduction techniques and new money can be in the medium-term interests of debtors and creditors. Moreover, once the resources available to support debt reduction are identified, evaluation of alternative menu items is a relatively straightforward exercise.

APPENDIX

This appendix offers some highly simplified examples of the possible effects of buy-backs financed by a portion of a country’s resources. Different time profiles for the stock and price of debt are derived that depend on the timing and rules imposed on the buy-back. No attempt is made here to provide realistic scenarios for any country or group of countries.

In every case the debtor country is assumed to devote a fixed share of export earnings to some blend of debt service and repurchases of debt. In the first set of simulations it is assumed that the debtor spreads his buy-backs over time and is not obliged to purchase debt at a price above what would have prevailed in the absence of future buy-backs. This assumption leads to a fall in the initial price of debt as the share of funds used for buy-backs increases. It follows that the stock of debt retired is substantial, and in one case sufficient debt is retired so that the price rises to par after a number of years.

The second set of simulations assumes that the debtor is obligated to utilize the funds that they have set aside for buy-backs. As compared with the first set of simulations, initial prices are higher because creditors fully incorporate future buy-backs into the market value of debt. The initial decline in the stock of debt is nevertheless substantial and, as in the first set of simulations, there are no feedbacks in the model between an improving situation and export performance. If such a feature was added to the model, the benefits of a buy-back would be larger and would occur more rapidly. These scenarios can be expressed more carefully as follows:

  • 1. A debtor country begins with a debt of $1 billion.
  • 2. The debt carries a 10 percent interest rate and infinite maturity.
  • 3. Exports of goods and services are assumed to be $10 billion.
  • 4. The country is expected to utilize 10 percent of exports for interest payments or debt buy-backs. The share of payments for interest, a, and buy-backs (1 − α) is a matter for negotiation. This rule is implemented for 20 years; thereafter a is set to 1. This share is established by negotiating new money equal to .10 Dt – αXt/10.
  • 5. The country’s debt is the ratio of the present value of payments divided by the contractual value of outstanding debt that remains following that time period’s buy-back.
  • 6. In the first set of simulations, it is assumed that future buy-backs are at the discretion of the debtor. In the second set of simulations, the buy-backs are included in the present value calculation. The initial price of debt is set by market participants with full knowledge about the size of the future buy-backs.

These assumptions suggest the following simple models:

Where

  • IPt = interest payment
  • Xt = exports
  • BB = cash used for buy-back
  • Dt = debt
  • Pt = market price of debt
  • PVP = present value of future payments including buy-backs (if known discounted at 10 percent).

The first simulation, Case 1, sets α = 1 so that all payments take the form of interest payments. In this case, debt grows more rapidly than the present value of expected payments, so the price of debt falls from $0.38 initially to $0.20 after 10 years and $0.09 after 20 years. (Given these assumptions, the price of debt will approach zero in the long run.)

Case 2 shows the same country, but now it uses 25 percent of its total payments for buy-backs and 75 percent for interest payments for the first 20 years. Notice that the initial price of the debt, $0.31, is lower than in Case 1 and that the price again declines over time, but less rapidly.

In Case 3, for the first 20 years, half of all payments are interest and half are buy-backs. Again, the initial price of debt is even lower at $0.24, but in this case, sufficient debt is retired so in the short run, the present value of interest payments grows more rapidly than debt, so that the price of debt actually increases to $0.32 after 20 years but again approaches zero in the long run.

Case 3:50 Percent Interest Payment—50 Percent Buy-Back
YearPriceDebtPresent valueBuy-Back
10.24975.11229.7320
20.24969.31232.7020
30.24964.49235.9720
40.25960.74239.5720
50.25958.12243.5320
100.28965.59270.0920
150.311,018.81312.8820
200.321,199.89381.8120
250.251,621.33399.990
300.172,366.96399.990

Finally, in Case 4, 75 percent of available funds are used for buy-backs and only 25 percent for interest payments. In this case the initial price is quite low, about $0.16, but the buy-back rapidly retires debt so that after 20 years contractual interest payments on remaining debt would be less than 10 percent of exports, the market price of debt would be $1.00, and the debt overhang would be eliminated.

Case 4:25 Percent Interest Payment—75 Percent Buy-Back
YearPriceDebtPresent valueBuy-Back
10.16877.86144.5930
20.19795.53149.0530
30.21724.00153.9630
40.24661.81159.3530
50.27607.70165.2930
100.48425.80205.1530
150.80337.35269.3330
201.00321.14321.1330
251.00310.25310.240
301.00310.25310.240

The second set of cases reflects the assumption that the future buy-backs are fully anticipated. In this case the initial price of debt is relatively unchanged since the present value includes future buy-backs.5 As can be seen in Case 5, however, the buy-back does succeed in limiting the growth of future debt so that prices fall less rapidly as compared with the baseline scenario.

Case 5:75 Percent Interest Payment—25 Percent Buy-Back Future Buy-Backs Known
YearPriceDebtPresent valueBuy-Back
10.391,034.05398.5010
20.371,080.34398.3510
30.351,129.99398.1910
40.341,183.26398.0110
50.321,240.41397.8110
100.251.544.79396.4810
150.192,097.59394.3410
200.142,809.67390.9010
250.094,295.44399.990
300.066,673.63399.990

The larger buy-backs in Cases 6 and 7 tell a similar story. The larger the buy-back, the more slowly prices decline. Notice that the price does not go to par in Case 7 even though the size of the buy-backs were identical to Case 4. Since future buy-backs are assumed known, the debtor country purchases debt at higher prices than the case where the future buy-backs were not reflected in the price.

Case 6:50 Percent Interest Payment—50 Percent Buy-Back Future Buy-Backs Known
YearPriceDebtPresent valueBuy-Back
10.391,009.16397.0120
20.381,037.76396.7220
30.371,067.67396.3920
40.361,098.94396.0320
50.351,131.62395.6420
100.301,318.42392.9820
150.251,549.79388.7020
200.211,833.27381.8120
250.152,737.57399.990
300.104,164.68399.990
Case 7:25 Percent Interest Payment—75 Percent Buy-Back Future Buy-Backs Known
YearPriceDebtPresent valueBuy-Back
10.40985.27395.5330
20.40998.01395.0830
30.391,010.95394.5930
40.381,024.08394.0530
50.381,037.39393.4630
100.351,106.21389.4730
150.331,176.89383.0530
200.301,244.72372.7230
250.221,804.35399.990
300.152,661.72399.990
*

Mr. Dooley, Chief of the External Adjustment Division in the Research Department of the IMF, is a graduate of Duquesne University, the University of Delaware, and the Pennsylvania State University.

Mr. Symansky, a Senior Economist in the Research Department of the IMF, is a graduate of the University of Wisconsin and formerly served with the World Bank and the Board of Governors of the Federal Reserve.

1

In reality the typical syndicated credit is a floating rate security that carries a spread over LIBOR of ½ to 2½ percent and has a maturity of about seven years. In actual financing packages, these additional details would be important, but the analytical framework is easier to work with if we assume that the benchmark security is a fixed interest perpetual obligation of the debtor country government.

2

A more realistic example would consider a floating-rate syndicated credit with a fixed spread over LIBOR. While this complicates the arithmetic, it does not alter the results in any important way.

3

This is strictly true only for consols, but it is approximately true for long maturity securities.

4

It should be recalled, however, that the buy-back might influence the market discount assumed to hold in this analysis.

5

The small rise in the price in the first period is due to the timing of buy-backs and interest payments in the simulation model. Buy-backs are assumed to take place at the beginning of the year, and interest payments are made at the end of the year.

    Other Resources Citing This Publication